Abstract

Although blood hemoglobin (Hgb) testing is a routine procedure in a variety of clinical situations, noninvasive, continuous, and real-time blood Hgb measurements are still challenging. Optical spectroscopy can offer noninvasive blood Hgb quantification, but requires bulky optical components that intrinsically limit the development of mobile health (mHealth) technologies. Here, we report spectral super-resolution (SSR) spectroscopy that virtually transforms the built-in camera (RGB sensor) of a smartphone into a hyperspectral imager for accurate and precise blood Hgb analyses. Statistical learning of SSR enables us to reconstruct detailed spectra from three color RGB data. Peripheral tissue imaging with a mobile application is further combined to compute exact blood Hgb content without a priori personalized calibration. Measurements over a wide range of blood Hgb values show reliable performance of SSR blood Hgb quantification. Given that SSR does not require additional hardware accessories, the mobility, simplicity, and affordability of conventional smartphones support the idea that SSR blood Hgb measurements can be used as an mHealth method.

© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement

1. INTRODUCTION

As one of the most common clinical laboratory tests, blood hemoglobin (Hgb) tests are routinely ordered as an initial screening of reduced red blood cell production to examine the general health status before other specific examinations [1]. Blood Hgb tests are extensively performed for a variety of patient care needs, such as anemia detection as a cause of other underlying diseases, assessment of hematologic disorders, transfusion initiation, hemorrhage detection after traumatic injury, and acute kidney injury [2–6]. Several recent studies found that blood Hgb or hematocrit levels are highly correlated with the occurrence of acute hemorrhage [7–10]. Unfortunately, the current technologies for noninvasively measuring blood Hgb levels are significantly limited. Importantly, repeated blood Hgb measurements using invasive tests can cause iatrogenic blood loss.

Unlike measuring oxygen saturation with pulse oximetry, blood Hgb measurements are not straightforward. Using blood drawn by a needle-based method, there are several biological assays for measuring Hgb content in grams per deciliter (i.e., ${{\rm g}\,{\rm dL}^{- 1}}$) of the blood. Portable point-of-care blood analyzers using blood draws (e.g., Abbott i-STAT and HemoCue) are available [11,12], but heavily rely on environment-sensitive cartridges with short shelf lives [13]. Such blood analyzers remain suboptimal and unaffordable for both resource-limited and homecare settings. Noninvasive devices (from Masimo and OrSense) can be used to detect blood Hgb levels. However, these devices require specialized equipment with relatively high costs. Moreover, the wide limits of agreement (LOA) compared with central laboratory tests often pose skepticism [14–16]. Several smartphone-based anemia detection technologies have made progress, including HemoGlobe, Eyenaemia, Smartphone dongle [12], HemaApp [17], and fingernail mobile application [18]. The common agreement is that conventional mobile health (mHealth) applications still need more robust validations to be reliable digital health products [19].

For noninvasive blood Hgb measurements, it is critical to rely on an appropriate anatomical sensing site where the underlying microvasculature is exposed on the skin surface without being affected by confounding factors of skin pigmentation and light absorption of molecules (e.g.,  melanin) in tissue. Commonly used clinical examination sites of pallor or microcirculation, such as the conjunctiva, the nailbed, the palm, and the sublingual region, provide a clue for an examination site selection [20–24]. In particular, the palpebral conjunctiva (i.e., inner eyelid) can serve as an ideal site for peripheral access, because the microvasculature is easily visible, and melanocytes are absent in this area [25–28]. The easy accessibility and relatively uniform vasculature of the inner eyelid allow for optical reflectance spectroscopy [29–32] and digital photography [33,34] to be tested for anemia assessments.

For accurate and precise blood Hgb quantification, spectroscopic analyses of light absorption of Hgb in reflection spectra are extensively used to measure Hgb content in tissue, because Hgb has a distinct characteristic absorption spectrum in the visible and near-infrared range [35–37]. On the other hand, this method relies heavily on complex and costly optical instrumentation such as spectrometers, imaging spectrographs, mechanical filter wheels, or liquid crystal tunable filters. Such dispersive optical components also result in very slow data acquisition, hampering clinical translation. Fortunately, several different research communities have demonstrated that it is possible to mathematically reconstruct hyperspectral (with high spectral resolution) or multispectral (with several spectral measurements) data from RGB images taken by a conventional camera (i.e., three-color sensors) or a few wavelengths [32,38–53]. Although an alternative approach with a higher spectral resolution would be necessary for specific biomedical applications, such computational approaches lay the groundwork for spectral super-resolution (SSR) spectroscopy of blood Hgb analyses.

In this study, we introduce SSR spectroscopy that enables us to mathematically reconstruct high-resolution spectra of blood Hgb from three color values of the R, G, and B channels acquired by the built-in camera of a smartphone. To noninvasively quantify blood Hgb levels, SSR spectroscopy further combines imaging of the palpebral conjunctiva (i.e., the inner eyelid) and spectroscopic quantification of total blood Hgb content. A dual-channel hyperspectral imaging system allows us to evaluate the performance of spectroscopic blood Hgb measurements from the eyelids. Using high-resolution spectral data of the eyelids, we establish a statistical learning framework of SSR for an mHealth application using an affordable smartphone. When compared with a wide range of blood Hgb values of individuals, Bland–Altman analyses show reliable characteristics of SSR blood Hgb quantification from the eyelids.

2. METHODS

A. Spectral Super-Resolution Spectroscopy for Blood Hemoglobin Analysis

In Fig. 1(a), the combination of imaging of the inner eyelid, SSR, and spectroscopic quantification of blood Hgb content can serve as a spectrometer-free hematology analyzer only using a smartphone. In SSR, hyperspectral data in the visible range are mathematically reconstructed from an RGB image (i.e., three-color information from R, G, and B channels). In other words, it intends to solve an ill-posed problem as an inverse mapping from a subsampled space (three color values) to a dense space (multiple wavelengths). Several computational approaches, including compressive sensing and deep learning, have successfully demonstrated that the missing wavelength information from the RGB image can be inferred [32,38–52]. On the other hand, deep learning methods heavily rely on establishing correlations between spatial patterns and spectral representations, requiring an extremely large sample size. Compressive sensing (or sparsity algorithm) based on three (RGB) spectral response functions would also be limited in generating high-resolution spectra for analyzing detailed biological spectral features. In this respect, we developed a statistical learning framework to noninvasively measure blood Hgb levels from the eyelid (i.e., palpebral conjunctiva) in the same units (i.e., ${{\rm g}\,{\rm dL}^{- 1}}$) as clinical laboratory blood Hgb tests [Fig. 1(b)]. SSR can virtually transform the smartphone camera into a hyperspectral imager without any additional optical attachments to a conventional smartphone, and the spectroscopic computation of Hgb can return a value of blood Hgb content. In addition, we ensured that separate training and validation datasets strengthen the utilization of SSR blood Hgb measurements for rapid field adaptation.

 

Fig. 1. Spectral super-resolution (SSR) spectroscopy for mobile health (mHealth) hemoglobin (Hgb) analyses. (a) The inner eyelid (i.e., palpebral conjunctiva) is used as an accessible sensing site for noninvasive blood Hgb quantification. An RGB image of the inner eyelid is conveniently captured using the built-in camera of a smartphone. The subject simply pulls down the eyelid to expose the conjunctiva, and a healthcare professional takes a photograph of the eyelid. The mobile application collects red (R), green (G), and blue (B) color information from the eyelid image and applies SSR to mathematically reconstruct spectra in the visible wavelength range. The spectral intensity reflected from the inner eyelid is sensitive to changes in Hgb content in the blood. The reconstructed spectrum of the acquired eyelid image is then processed to accurately and precisely predict the amount of total blood Hgb content. The result displays the blood Hgb count in units of ${\rm g}\;{{\rm dL}^{- 1}}$ in the same manner of clinical laboratory Hgb tests. (b) Statistical learning in the mHematology for SSR blood Hgb computation developed using separate training and validation datasets. The first step is to apply SSR to the eyelid portion of the RGB image. The second step is to compute blood Hgb content in ${\rm g}\;{{\rm dL}^{- 1}}$ using the spectroscopic model of blood Hgb, which is also validated by the clinical laboratory blood Hgb tests (i.e., the gold standard).

Download Full Size | PPT Slide | PDF

B. Spectral Super-Resolution Algorithm

SSR is the key element to achieve spectrometer-free, yet spectroscopic, quantification of blood Hgb content from the inner eyelid. The mathematical relationship between the RGB and full spectral intensity is described as

$${{\boldsymbol x}_{3 \times 1}} = {{\boldsymbol S}_{3 \times N}}{{\boldsymbol y}_{N \times 1}} + {{\boldsymbol e}_{3 \times 1}},$$
where ${\boldsymbol x}$ is a $3 \times 1$ vector corresponding to three color values in the R, G, and B channels (${\boldsymbol x} = {[{\rm R,G,B}]^{\rm T}} = {[{ I({\rm R}),I({\rm G}),I({\rm B})}]^{\rm T}})$, ${\boldsymbol S}$ is a ${3} \times N$ matrix of the RGB spectral response of the three-color sensor (i.e., built-in camera of Samsung Galaxy J3), ${\boldsymbol y}$ is a $N \times {1}$ vector that has the spectral intensity (${\boldsymbol y} = {[{I({{\lambda _1}}),I({{\lambda _2}}), \ldots ,I({\lambda _N})}]^{\rm T}}$) where $\lambda$ is discretized from 450 nm to 679 nm with a spectral interval of 1 nm ($N = {230}$), and ${\boldsymbol e}$ is a ${3} \times {1}$ vector of the system noise with zero mean. ${{\boldsymbol x}_{3 \times 1}}$ and ${{\boldsymbol y}_{N \times 1}}$ of the eyelid on each participant were acquired by a smartphone camera and a customized hyperspectral imaging system [Figs. 2(a), 2(c), and Fig. S1 and Supplementary Methods in Supplement 1], respectively. Basically, the hyperspectral reconstruction for SSR from the RGB signal is to obtain ${[{{{\boldsymbol S}_{3 \times N}}}]^{- 1}}$. However, this inverse calculation is an underdetermined problem such that $N \gt 3$. Thus, we formulated a statistical learning framework to take advantage of robust computational methods of least-squares to solve ${[{{{\boldsymbol S}_{3 \times N}}}]^{- 1}}$ in Eq. (1).
 

Fig. 2. High-quality spectra acquired by the image-guided hyperspectral line-scanning system and the mHematology mobile application. (a) Photograph of the image-guided hyperspectral line-scanning system for imaging the exact portion of the inner eyelid. The participant sits in front of the system, facing the telecentric lens, places the chin on the chinrest, and pulls down the eyelid for imaging when instructed. (b) Location where the hyperspectral line-scanning is performed (translucent white rectangle). The hyperspectral line-scan dataset contains spatial ($y$) and wavelength ($\lambda$) information. The averaged spectrum corresponds to the average intensity along the spatial $y$ axis for each $\lambda$ value. The characteristic absorption spectrum of blood Hgb is clearly visible. (c) mHematology mobile application developed for data acquisition in a low-end Android smartphone (Samsung Galaxy J3). On the main application screen, it displays a circle and arc to serve as guidance for locating the eyeball and the inner eyelid at consistent distance and position within the image. To remove the background room light, the application automatically acquires two RGB photographs by controlling the built-in flashlight (i.e., white-light LED) to turn on and off. To compensate for the system response, two RGB images of a reflectance standard are taken (left). Similarly, the application automatically takes two RGB images with flash on and flash off for the individual’s exposed eyelid (right). (d) Spectral profiles of the white-light LED illumination sources in the image-guided hyperspectral line-scanning system and Samsung Galaxy J3. The data acquisition procedure incorporates reference measurements of white reflectance standards (99% reflectivity in the visible range) to compensate for the spectral responses of the light source and the camera in the system.

Download Full Size | PPT Slide | PDF

In particular, we implemented fixed-design linear regression with polynomial features to construct a highly stable transformation matrix that converts RGB data to high-resolution spectral data, maximizing the accuracy of blood Hgb quantification. Using the multiple collections of spectral data and RGB data, we changed the underdetermined problem into an overdetermined problem that can be stably solved by a least-squares method. Specifically, ${{\boldsymbol X}_{3 \times m}}$ and ${{\boldsymbol Y}_{N \times m}}$ were formed by adding ${{\boldsymbol x}_{3 \times 1}}$ and ${{\boldsymbol y}_{N \times 1}}$ from $m$ different measurements (i.e., training dataset). The relationship in Eq. (1) is rewritten as

$${{\boldsymbol X}_{3 \times m}} = {{\boldsymbol S}_{3 \times N}}{{\boldsymbol Y}_{N \times m}}.$$
Equation (2) can also be expressed explicitly:
$$\begin{split}\!\!\!\!\!\left[\begin{array}{*{20}{l}} {\rm R}_{1} & {\rm R}_{2} & \cdots & {\rm R}_{m} \\ {\rm G}_{1} & {\rm G}_{2} & \cdots & {\rm G}_{m} \\ {\rm B}_{1} & {\rm B}_{2} & \cdots & {\rm B}_{m} \end{array}\right]&=\left[\begin{array}{*{20}{l}} S_{1}\!\left(\lambda_{1}\right) S_{1}\!\left(\lambda_{2}\right) \cdots S_{1}\!\left(\lambda_{N}\right) \\ S_{2}\!\left(\lambda_{1}\right) S_{2}\!\left(\lambda_{2}\right) \cdots S_{2}\!\left(\lambda_{N}\right) \\ S_{3}\!\left(\lambda_{1}\right) S_{3}\!\left(\lambda_{2}\right) \cdots S_{3}\!\left(\lambda_{N}\right) \end{array}\right]\\&\quad\times\left[\begin{array}{*{20}{c}} I_{1}\!\left(\lambda_{1}\right) & I_{2}\!\left(\lambda_{1}\right) & \cdots & I_{m}\!\left(\lambda_{1}\right) \\ I_{1}\!\left(\lambda_{2}\right) & I_{2}\!\left(\lambda_{2}\right) & \cdots & I_{m}\!\left(\lambda_{2}\right) \\ \vdots & \vdots & \vdots & \vdots \\ I_{1}\!\left(\lambda_{N}\right) & I_{2}\!\left(\lambda_{N}\right) & \cdots & I_{m}\!\left(\lambda_{N}\right) \end{array}\right],\end{split}$$
where ${S_1}({{\lambda _1},{\lambda _2}, \ldots ,{\lambda _N}})$, ${S_2}({{\lambda _1},{\lambda _2}, \ldots ,{\lambda _N}})$, and ${S_3}({{\lambda _1},{\lambda _2}, \ldots ,{\lambda _N}})$ are the spectral response functions in the R, G, and B channels of the smartphone camera (also known as the sensitivity function of the camera), respectively. The relationship in Eq. (2) can be expressed in an inverse format:
$${{\boldsymbol Y}_{N \times m}} = {{\boldsymbol T}_{N \times 3}}{{\boldsymbol X}_{3 \times m}},$$
where ${{\boldsymbol T}_{N \times 3}} = {[{{{\boldsymbol S}_{3 \times N}}}]^{- 1}}$ is the transformation matrix. If Eq. (4) is solved for the unknown ${{\boldsymbol T}_{N \times 3}}$, then ${{\boldsymbol T}_{N \times 3}}$ can be used to transform the RGB dataset into the hyperspectral dataset. Second, we considered that each three-color sensor model in different cameras has unique RGB spectral responses with spectral overlaps among the R, G, and B channels. To effectively incorporate the RGB spectral responses of the camera, ${{\boldsymbol X}_{3 \times m}}$ was expanded to ${\hat{\boldsymbol X}_{p \times m}}$ for maximizing the accuracy of SSR such that
$${{\boldsymbol Y}_{N \times m}} = {\hat{\boldsymbol T}_{N \times p}}{\hat{\boldsymbol X}_{p \times m}},$$
where ${\hat{\boldsymbol X}_{p \times m}}$ can be expressed explicitly:
$${\hat{\boldsymbol X}_{p \times m}} = \left[\begin{array}{*{20}{c}} {\rm R}_{1} & {\rm G}_{1} & {\rm B}_{1} & \cdots & {\rm R}_{1}^{i} {\rm G}_{1}^{j} {\rm B}_{1}^{k} & \cdots \\ \vdots & \vdots & \vdots & \vdots & \vdots & \vdots \\ {\rm R}_{m} & {\rm G}_{m} & {\rm B}_{m} & \cdots & {\rm R}_{m}^{i} {\rm G}_{m}^{j} {\rm B}_{m}^{k} & \cdots \end{array}\right]^{\rm T},$$
where the exact powers of $i$, $j$, and $k$ of the cross terms were uniquely determined for the specific three-color sensor model (i.e., Samsung Galaxy J3), by checking the error between the reconstructed hyperspectral data and the original data. Third, the inverse of the expanded transformation matrix $\hat{\boldsymbol T}$ in Eq. (5) can be considered a minimum-norm-residual solution for ${\boldsymbol Y} = \hat{\boldsymbol T}\hat{\boldsymbol X}$. Typically, this inverse problem is to solve a least-squares problem using QR decomposition, in particular, the QR solver. After QR factorization was applied to $\hat{\boldsymbol X}$, $\hat{\boldsymbol T}$ was estimated by minimizing the sum of the squares of the elements of ${\boldsymbol Y} - \hat{\boldsymbol T}\hat{\boldsymbol X}$. Finally, when RGB data (${\boldsymbol x} = {[{\rm R,G,B}]^{\rm T}} = {[{I({\rm R}),I({\rm G}),I({\rm B})}]^{\rm T}})$ were substituted into Eq. (5), the corresponding spectral intensity values (${\boldsymbol y} = {[{I({{\lambda _1}}),I({{\lambda _2}}), \ldots ,I({\lambda _N})}]^{\rm T}}$) were generated. In other words, the transformation matrix $\hat{\boldsymbol T}$ converted RGB data to high-resolution spectral data. Before implementing SSR in our mobile application, the statistical learning method of SSR was initially validated using hyperspectral data and RGB images generated by the hyperspectral imaging system (see Supplementary Methods and Fig. S2 in Supplement 1.

C. Experimental Setups

It is essential for SSR to have high-quality hyperspectra from the exact portion of the eyelid as a part of the training dataset. To acquire high-resolution spectra reflected from the exact portion of the inner eyelid, we developed an image-guided hyperspectral line-scanning system (hereinafter referred to as the hyperspectral imaging system), custom-built from an astronomical spectroscopic system frame where a dual-channel spectrograph was mounted with a telescope [see Fig. 2(a) and Supplementary Methods and Figs. S1 and S3 in Supplement 1. In particular, the guiding camera of the hyperspectral imaging system allowed us to pinpoint the exact location of hyperspectral line-scanning in the eyelid [Fig. 2(b)]. In addition, the relatively large area (${0.8}\;{\rm mm} \times {7.2}\;{\rm mm}$) of hyperspectral line-scanning in the vertical direction results in a representative spectrum of the eyelid, independent of the exact horizontal location in the eyelid (Fig. S4). A white-light LED ring illuminator was mounted to a telecentric lens (${0.5}\; \times$, Edmund Optics) via a 3D printed ring holder to fit the lens circumference [Supplement 1 and Fig. 2(d)]. Typically, a telecentric lens is used in machine vision systems for inspecting circuit boards, surface defects, and packages, because of the ability to eliminate parallax error and to provide constant magnification. Telecentric imaging is also beneficial for biological tissue, including resolution enhancement by diffuse light suppression, large constant transverse field of view, consistent magnification over the axial direction, and long working distance [54–57]. Measurements of a white reference reflectance standard (SRT-99-050, Labsphere) were acquired to correct for the system response (both illumination and detection). While the patient was immobilized on the chinrest, two sequential data of the inner eyelid (and the white reflectance standard) were acquired by turning on and off the LED illuminator [Fig. 2(d)] with a time delay of 0.2 s to remove the ambient room light (Supplement 1). A series of tests using tissue-mimicking phantoms confirmed the basic performance and operation of the hyperspectral imaging system (Fig. S5).

In addition, we developed a mobile application for the purpose of acquiring RGB photographs of the eyelids using an Android smartphone. For widespread utilizations in resource-limited settings, such as low and middle-income countries as well as home settings, we implemented the mobile application in a low-end Android smartphone, Samsung Galaxy J3. Using Android Studio (open source integrated development environment) for Google’s Android operating system, we created a beta version of an mHematology mobile application. On the main application screen, it displayed a circle and arc to serve as guidance for locating the eyeball and the inner eyelid at consistent distance and position within the image [Fig. 2(c)]. As the microvasculature of the eyelid is not often uniform, we made use of RGB values averaged from the entire eyelid. Because we acquired a representative spectrum of the eyelid, independent of the horizontal location (Fig. S4 of Supplement 1), we avoided coregistration between hyperspectral line-scanning and smartphone imaging for the development of the SSR algorithm. To calibrate the system response, including the unique spectral profile of the built-in LED flashlight in the smartphone [Fig. 2(d)], reference measurements using a round-shaped white reflectance standard (SRS-99-010, Labsphere), which had a similar size to the eye, were also acquired in the mobile application [Fig. 2(c)]. To remove the ambient room light, the application automatically acquired two sequential RGB photographs with a time delay of 0.5 s for the individual’s exposed inner eyelid (and the white reflectance standard) by controlling the built-in LED flashlight (i.e., white-light LED) to turn on and off [Figs. 2(c) and 2(d)]. We validated the irradiance of both light sources used in the hyperspectral imaging system and the mHematology mobile application to ensure minimal light exposure to the eyes (see Supplementary Methods and Table S1 in Supplement 1), while never being directly aimed at the eyeball during imaging.

D. System Response Compensation and Ambient Light Removal

We implemented a data acquisition procedure not to be affected by variations in the illumination and detection of the imaging systems as well as the background ambient room light as follows: The measured spectral intensity ${I_m}(\lambda)$ (or RGB intensity ${I_m}({\rm RGB})$) reflected from the eyelid is expressed as a function of the wavelength $\lambda$ (or R, G, and B channels of the smartphone camera):

$$\begin{split}{I_m}({\lambda \,{\rm or \,RGB}} )& = L({\lambda \,{\rm or\, RGB}} )\cdot C ({\lambda \,{\rm or\,RGB}} )\\&\quad\cdot D({\lambda \,{\rm or\, RGB}} )\cdot y({\lambda \,{\rm or\, RGB}} ),\end{split}$$
where $L({\lambda \,{\rm or\, RGB}})$ is the spectral shape of the illumination light source, $C({\lambda \,{\rm or\, RGB}})$ is the spectral response of all optical components in the imaging system, $D({\lambda \,{\rm or\, RGB}})$ is the spectral response of the detector (e.g., mono sensor in the hyperspectral imaging system or three-color RGB sensor embedded in the smartphone), and $y({\lambda \,{\rm or\, RGB}})$ is the true spectral intensity (or RGB intensity) reflected from the eyelid. First, to compensate for the spectral response of the system including the LED light sources [Fig. 2(d)], we used the white reflectance standards that have a reflectivity of 99% in the visible range. The use of the reflectance standards could also be avoided by replacing them with the back of a white business card made of thick cardboard paper. $y({\lambda \,{\rm or\, RGB}})$ can be obtained by normalizing ${I_m}({\lambda \,{\rm or\, RGB}})$ with the reflectance measurement ${I_{\rm reference}}({\lambda \,{\rm or\, RGB}})$ ($= L({\lambda \,{\rm or\, RGB}})\cdot C({\lambda \,{\rm or\, RGB}})\cdot D({\lambda \,{\rm or\, RGB}})$):
$$y({\lambda \,{\rm or\, RGB}} ) = \frac{{{I_m}({\lambda \,{\rm or\, RGB}} )}}{{{I_{\rm reference}}({\lambda \,{\rm or\, RGB}} )}}.$$
Second, to remove the ambient stray and background room light ${I_{\rm background}}({\lambda \,{\rm or\, RGB}})$, two measurements were acquired with the illumination light source (i.e., white-light LED ring illuminator of the hyperspectral imaging system and built-in LED flashlight of the smartphone) on and off. Because the time delay of the sequential images was only 0.2–0.5 s, movement artifacts between the sequential images were minimized. Finally, Eq. (8) can be rewritten by subtracting ${I_{\rm background}}({\lambda \,{\rm or\, RGB}})$ from each measurement:
$$\!\!\!\!y\!\left({\lambda \,{\rm or\, RGB}} \right) = \frac{{{I_m}\!\left({\lambda \,{\rm or\, RGB}} \right) - {I_{\rm background}}\!\left({\lambda \,{\rm or\, RGB}} \right)}}{{{I_{\rm reference}}\!\left({\lambda \,{\rm or\, RGB}} \right) - {I_{\rm background}}\!\left({\lambda \,{\rm or\, RGB}} \right)}}.$$

E. Spectroscopic Model for Blood Hgb Quantification

We built a spectroscopic blood Hgb prediction model to compute blood Hgb content from spectra reflected from the eyelids using a training dataset only. Analytical model-based Hgb computation methods can be used, because Hgb has distinct spectral signatures (e.g.,  Soret and ${ Q}$ bands) in the visible range [35–37]. However, such analytical methods require a priori information on all possible light absorbers in tissue for accurate and precise blood Hgb quantification. When using a high-resolution spectrum $y(\lambda)$ with a large number of wavelengths $\lambda = {\lambda _1},{\lambda _2}, \ldots ,{\lambda _N}$, we can identify underlying key latent variables from the spectral intensity ($y(\lambda) = [{I({{\lambda _1}}),I({{\lambda _2}}), \ldots ,I({\lambda _N}})$]) that are responsible for capturing the most variations in the outcome variable (i.e., blood Hgb values). Thus, we made use of partial least-squares regression (PLSR), which has been extensively used to model a relationship among measured variables (i.e., predictors) and response variables (i.e., outcomes) in a variety of biological and medical applications [58,59]. PLSR transformed high-dimensional measured variables onto a reduced space of latent variables in a similar manner of principal component analysis. To avoid possible overfitting of the prediction model, we determined an optimal number of principal components and validated the spectroscopic blood Hgb prediction model with separate validation datasets that did not include any training hyperspectral data in a manner of cross-validation (Supplement 1 and Fig. S6).

F. Bland–Altman and Intra-Class Correlation Coefficient Analyses

We conducted Bland–Altman analyses to compare spectroscopic and SSR blood Hgb measurements as a non-parametric statistical method. Bland–Altman analyses are widely used for comparing a new diagnostic method with reference diagnostic tests [60]. In our case, the bias was defined by the mean of differences between spectroscopic (or SSR) and clinical laboratory blood Hgb measurements ($d = {{\rm Hgb}^{\rm spectroscopic}} - {{\rm Hgb}^{\rm clinical}})$:

$${\rm Bias }=\bar{d}=\frac{1}{n} \sum_{k=1}^{n} {\rm d}_{k}.$$
The 95% LOA was defined by a 95% prediction interval of the standard deviation:
$${\rm LOA}={\bar{d}} \pm 1.96 \sqrt{\frac{1}{n-1} \sum_{k=1}^{n}\left(d_{k}-{\bar{d}}\right)^{2}}.$$
We also computed intra-class correlation coefficients (ICCs) for evaluating repeatability (test–retest), intra-measurement, and inter-measurement reliability analyses [61], based on two-way mixed effects with absolute agreement.

G. Software for Signal Processing and Statistical Analyses

For data processing and algorithm developments, we computed the hyperspectral and RGB datasets of study participants and developed the spectroscopic blood Hgb model and the spectral super-resolution algorithm, using MATLAB (MATLAB R2018b, MathWorks). For statistical analyses, we evaluated linear regression, logistic regression, correlations, $t$ tests, and intra-class correlations, using STATA (STATA 14.2, StataCorp).

Tables Icon

Table 1. Participant Characteristics

View Table | View all tables in this article

3. RESULTS

A. Study Participants and Characteristics

Table 1 summarizes the characteristics of a total of 153 study participants in the clinical study within the facilities overseen by the Academic Model Providing Access to Healthcare (AMPATH) program. AMPATH is a consortium of academic health institutions in high-income countries to partner with health centers and universities in low-income countries. Moi University Teaching and Referral Hospital (MTRH) in Eldoret, Kenya, provides a clinical setting as one of Kenya’s two public hospitals, and Purdue University is one of the participating institutions in the United States. We utilize the data collected from patients who were referred for complete blood count tests at MTRH (Supplement 1). We use hyperspectral and RGB image data of the eyelids, acquired by the hyperspectral imaging system and the mHematology mobile application (Supplement 1). As the gold standard, we use blood Hgb levels measured from venous blood draws in the AMPATH Accredited Clinical Laboratory using a commercial hematology analyzer (Beckman Coulter AcT 5diff auto, Beckman Coulter, Inc.) immediately before or after imaging of the eyelids.

The clinical laboratory blood Hgb values cover a wide range from 3.3 to ${19.2}\;{\rm g}\;{{\rm dL}^{- 1}}$ (Fig. S7). In particular, the extremely low blood Hgb values (${\lt}{5}{-} {6}\;{\rm g}\;{{\rm dL}^{- 1}}$) are highly beneficial for validating the SSR spectroscopic blood Hgb quantification method, although the overall blood Hgb levels are relatively high with a median of ${13.1}\;{\rm g}\;{{\rm dL}^{- 1}}$, due to the high elevation of our study location (2700 m above sea level, Eldoret, Kenya). The acclimatization increases circulating blood Hgb concentration at a high altitude (${\gt}1500\;{\rm m}$ above sea level) [62,63]. We quantify the amount of Hgb in the blood in a similar manner of conventional blood testing, which is not adjusted by other pathological conditions. Although the sample size is relatively small, we ensure the reliability of modeling SSR spectroscopic blood Hgb prediction by employing separate training (90%) and validation datasets (10%). For developing the blood Hgb quantification model and the SSR algorithm, we randomly select 138 individuals as a training dataset ($n = {138}$). The average Hgb level is ${12.65}\;{\rm g}\;{{\rm dL}^{- 1}}$ with the standard deviation (SD) of ${3.11}\;{\rm g}\;{{\rm dL}^{- 1}}$, and the average age is 37.78 years with SD of 16.38 years. As a validation dataset, we blindly test the remaining 15 individuals ($n = {15}$) not included in the training dataset. The average Hgb level is ${11.06}\;{\rm g}\;{{\rm dL}^{- 1}}$ with SD of ${3.62}\;{\rm g}\;{{\rm dL}^{- 1}}$, and the average age is 39.13 years with SD of 17.30 years. The overall performance is also evaluated with additional cross-validations of 10 randomly selected different combinations of training and validation datasets, maintaining the ratio of the two datasets.

B. Spectroscopic Blood Hgb Quantification Using Hyperspectral Data

Using the eyelids’ spectra of the training dataset ($n = {138}$) acquired by the hyperspectral imaging system, we construct a blood Hgb computation model for predicting actual blood Hgb content (Supplement 1 and Fig. S6). In Fig. 3(a), the linear correlation between the computed and laboratory blood Hgb levels (i.e., the gold standard) shows a high coefficient of determination ${R^2}$ value of 0.953 for the training dataset. When we apply the validation dataset of 15 individuals not included in the training dataset to the identical spectroscopic blood Hgb prediction model, an excellent ${R^2}$ value of 0.954 and a narrow 95% LOA of [${-}{1.56},\;{1.58}\;{\rm g}\;{{\rm dL}^{- 1}}$] with bias of ${0.01}\;{\rm g}\;{{\rm dL}^{- 1}}$ in the validation dataset ($n = {15}$) underpin the fidelity of the model [Figs. 3(a) and 3(b)]. Within LOA, the data points are uniformly scattered; the bias of the validation dataset does not have any trend over the mean blood Hgb values (Table S2 in Supplement 1). We obtain the consistent $R^2$ and LOA values of validation datasets from different combinations of training and validation datasets [Figs. 3(c) and 3(d)]. These results support the idea that the hyperspectral measurements of the eyelids can be used for accurately and precisely assessing actual Hgb content in the blood, noninvasively.

 

Fig. 3. Performance of spectroscopic blood Hgb measurements of the left and right inner eyelids. (a) High correlations between the computed and clinical laboratory blood Hgb levels in both of the training ($n = {138}$ plotted in blue) and validation ($n = {15}$ plotted in red) datasets. (b) Bland–Altman analysis of comparing the computed blood Hgb levels with the clinical laboratory results, showing narrow 95% limits of agreement (LOA) of [${-}{1.56},\;{1.58}\;{\rm g}\;{{\rm dL}^{- 1}}$] with bias of ${0.01}\;{\rm g}\;{{\rm dL}^{- 1}}$ in the validation dataset. (c) ${R^2}$ values between the computed blood Hgb levels and the clinical laboratory results of validation datasets from 10 different combinations of training and validation datasets. (d) LOA (error bar) and bias (square point) values of validation datasets from the different training–validation combinations. (e) and (f) Systematic comparisons of the spectra measured from the left (green) and right (magenta) inner eyelids among a subset of 36 participants. (e) The average spectral differences between the left and right eyelid spectra (Fig. S8) are statistically insignificant, and the Pearson correlation coefficients are close to 1 in all of the participants. (f) High correlations of the left and right spectroscopic blood Hgb measurements, compared with the clinical laboratory blood Hgb test results.

Download Full Size | PPT Slide | PDF

Tables Icon

Table 2. Spectral Differences between the Left and Right Inner Eyelids

View Table | View all tables in this article

C. Comparison of Left and Right Inner Eyelids

We further test whether the spectroscopic blood Hgb levels measured from both left and right eyelids are statistically identical among a subset of 36 participants who agreed to image both eyelids (Supplement 1). We first check that the spectral profiles of both eyelids significantly resemble each other (Fig. S8). Using a response feature analysis [64], the intensity differences between the spectra of the left and right eyelids are reduced to a single response ($\mathop \sum \nolimits_{i = 1}^N [{{I^{\rm left}}({{\lambda _i}}) - {I^{\rm right}}({{\lambda _i}})}]/N$, where $I({{\lambda _i}})$ is the spectral intensity for the wavelength ${\lambda _i}$ and $N$ is the number of all wavelengths) on each individual participant. In Fig. 3(e) and Table 2, the spectral difference values of the left and right eyelids are statistically negligible ($p \hbox{-} {\rm value} = {0.722}$ with a two-tailed $t$ test). Pearson correlation coefficients between the spectra of both eyelids are also close to 1 for all of the participants [Fig. 3(e)]. Then, we compute blood Hgb levels with the spectra reflected from both eyelids on each participant, using the blood Hgb computation model developed above. Figure 3(f) depicts that the ${R^2}$ values of spectroscopic blood Hgb values computed from the left and right spectra, compared with the clinical laboratory blood Hgb levels, are 0.911 and 0.835, respectively. We also analyze ICC to quantify the degree of correlation and agreement between the left and right eyelids, based on two-way mixed effects with absolute agreement (Table 3). The ICC values for single and average measures are 0.87 with a 95% confidence interval of [0.77, 0.93] and 0.93 with a 95% confidence interval of [0.87, 0.97], respectively. These results support the underlying idea that reliable spectroscopic blood Hgb measurements are achieved regardless of the left or right eyelids.

D. SSR in mHematology for Blood Hgb Assessments

We first examine the quality of spectra reconstructed from RGB data acquired by the mHematology mobile application with Samsung Galaxy J3 [Fig. 2(c)]. The RGB values averaged from the entire eyelid are used to reconstruct a representative and reliable spectrum on each individual (Fig. S9). The delineation process of the eyelid [Fig. S9(b)] includes manual landmark selection, geometrical boundary extrapolation, specular reflection area removal, and the average of RGB values. The same training ($n = {138}$) dataset used in the spectroscopic Hgb quantification (Supplement 1 and Fig. S6) is employed to obtain the transformation matrix [Figs. 4(a)–4(d)]. The validation dataset ($n = {15}$) is never seen by the statistical learning of SSR. Figures 4(a) and 4(b) show the original spectra of the eyelids and the reconstructed spectra from the training dataset [Fig. S10(a)]. The differences between the original and reconstructed spectra are very small when plotted with 95% confidence intervals as a function of the wavelength $\lambda$ [Fig. 4(c)]. For the validation dataset [Fig. S10(b)], the reconstructed spectra over different blood Hgb levels are also in good agreement with the original spectra [Figs. 4(e) and 4(f)], while the differences in the validation dataset [Fig. 4(g)] are slightly higher than those in the training dataset [Fig. 4(c)]. In general, the wavelength range between 450 and 575 nm shows more variations than the longer wavelengths, because distinct Hgb absorption is present in this range. The average spectral difference values and the Pearson correlation coefficients between the original and reconstructed spectra in the training and validation datasets show the reliability of SSR in a broad range of blood Hgb levels [Figs. 4(d) and 4(h)].

Tables Icon

Table 3. Intra-Class Correlations (ICC) between the Left and Right Inner Eyelidsa

View Table | View all tables in this article
 

Fig. 4. Comparisons between the original spectra and the SSR-reconstructed spectra. (a) and (b) Inner eyelids’ spectral intensity (acquired by the image-guided hyperspectral line-scanning system and the mHematology application) plotted with the clinical laboratory blood Hgb values (vertical axis) from the training dataset ($n = {138}$). (c) Differences between the original and SSR-reconstructed spectra plotted with 95% confidence intervals at each wavelength. The transformation matrix that converts RGB data to spectral data is optimized by minimizing the differences in the training dataset. (d) Average spectral differences and Pearson correlation coefficients between the original and reconstructed spectra in the training dataset as a function of blood Hgb levels corresponding to each spectrum. (e) and (f) Inner eyelids’ spectral intensity (acquired by the image-guided hyperspectral line-scanning system and the mHematology application) visualized with the clinical laboratory blood Hgb values (vertical axis) from the validation dataset ($n = {15}$). (g) Differences between the original and SSR-reconstructed spectra plotted with 95% confidence intervals at each wavelength are still small, supporting the high fidelity of SSR. The differences in the wavelength range between 450 and 575 nm are generally higher, because distinct Hgb absorption is present in this range [Fig. S5(d)]. (h) Average spectral differences and Pearson correlation coefficients between the original and reconstructed spectra in the validation dataset as a function of blood Hgb levels.

Download Full Size | PPT Slide | PDF

We next compute blood Hgb content in units of [${\rm g}\;{{\rm dL}^{- 1}}$] from the reconstructed spectra of the eyelids using PLSR. The blood Hgb prediction model reliably estimates blood Hgb content for the training and validation datasets (Fig. 5). The linear correlation between the clinical laboratory blood Hgb levels and the SSR-based computed Hgb levels shows an excellent ${R^2}$ value of 0.932 for the training dataset [Fig. 5(a)]. More importantly, the blood Hgb levels computed from the validation dataset also result in an excellent ${R^2}$ value of 0.912 [Fig. 5(a)] with narrow LOA of [${-}{2.20},\;{2.29}\;{\rm g}\;{{\rm dL}^{- 1}}$] and bias of ${0.04}\;{\rm g}\;{{\rm dL}^{- 1}}$ [Fig. 5(b)]. The bias in the validation dataset is not associated with the mean blood Hgb values in the Bland–Altman plot; there is no significant slope between the independent and dependent variables with a $p$-value of 0.453 from linear regression (Table S3 in Supplement 1). The consistent ${R^2}$ and LOA values of validation datasets from 10 different combinations of training and validation datasets [Figs. 5(c) and 5(d)] support the overall performance that SSR blood Hgb measurements enable reliable assessments of actual Hgb content in the blood, noninvasively.

 

Fig. 5. Performance of SSR blood Hgb measurements with the mHematology mobile application. (a) High correlations between the SSR-computed and clinical laboratory blood Hgb levels in both training ($n = {138}$ plotted in blue) and validation ($n = {15}$ plotted in red) datasets. (b) Bland–Altman analyses of comparing the computed blood Hgb measurements with the clinical laboratory results, showing narrow 95% limits of agreement (LOA) of [${-}{2.20},\;{2.29}\;{\rm g}\;{{\rm dL}^{- 1}}$] and bias of ${0.04}\;{\rm g}\;{{\rm dL}^{- 1}}$ in the validation dataset. In particular, the bias is not associated with actual blood Hgb levels in the validation dataset (Table S3 in Supplement 1). (c) ${R^2}$ values between the computed blood Hgb levels and the clinical laboratory results of validation datasets from 10 different combinations of training and validation datasets. (d) LOA (error bar) and bias (square point) values of validation datasets from the different training–validation combinations. The mHematology application reliably predicts the actual blood Hgb levels without any hardware attachments to the smartphone.

Download Full Size | PPT Slide | PDF

4. DISCUSSION

Although clinical applications of the reported blood Hgb quantification are not limited to noninvasive anemia assessments, we analyze receiver operating characteristic (ROC) curves using blood Hgb thresholds for defining anemia, recommended by WHO: ${\rm Hgb}\; \lt \;{13}\;{\rm g}\;{{\rm dL}^{- 1}}$ for men (${\rm age}\; \ge \;{15}\;{\rm years}$) and ${\rm Hgb}\; \lt \;{12}\;{\rm g}\;{{\rm dL}^{- 1}}$ for women (${\rm age}\; \ge \;{15}\;{\rm years}$) [65]. The spectroscopic and SSR blood Hgb measurements show excellent performance of anemia assessments with areas under the ROC curves (${\gt}\;{0.98}$) for both men and women (Fig. S11 in Supplement 1). Only using mere RGB data (without SSR), conjunctival redness scoring or pallor examination for anemia assessments may not provide sufficient information to reliably assess blood Hgb levels. Indeed, a multiple linear regression analysis of the actual blood Hgb levels (continuous outcome variable) against the R, G, and B values (predictor variables) returns an underperforming ${R^2}$ value of 0.448 without any statistical significance (Table S4 and Fig. S12 in Supplement 1).

We analyze ICC to evaluate the practical reliability and performance of the reported blood Hgb measurements. The ICCs between two blood Hgb measurements (spectroscopic and SSR blood Hgb quantification methods) by the hyperspectral imaging system and the mHematology application are based on two-way mixed effects with absolute agreement (Table 4). The ICC values for single and average measures are 0.94 with a 95% confidence interval of [0.92, 0.96] and 0.97 with a 95% confidence interval of [0.96, 0.98], respectively. These ICC estimates clearly support the reliability of the spectroscopic and SSR blood Hgb measurements from the eyelids. In addition, we calculate percentage differences of the spectroscopic and SSR blood Hgb measurements from the clinical laboratory blood Hgb values, returning 4.61% and 6.01% on average for the hyperspectral imaging system and the mHematology application, respectively. These ranges of agreement limits are comparable to the analytical quality requirement by the Clinical Laboratory Improvement Amendments (CLIA) in the United States; the original criterion of acceptable performance for blood Hgb tests is $\pm \;{7}\%$ of target values [66], although the recently suggested one is $\pm \;{4}\%$ of target values [67].

Tables Icon

Table 4. ICC between Spectroscopic and SSR Blood Hgb Measurementsa

View Table | View all tables in this article

The reported mobile application has several advantages over medical device regulations in general. First, under the FDA guideline, the mHematology application would require an abbreviated investigational device exemption for clinical studies, mainly because it would not be considered as a custom device. Technically, this is the same as taking photographs using a camera under flashlight. The LED light illumination is not directly aimed at the human subjects’ eyeball during imaging of the eyelid. Indeed, the risk associated with the illumination intensity is minimal (Table S1 and Supplementary Methods in Supplement 1). Second, given no need for collecting any body fluids, the reported spectroscopic and SSR blood Hgb measurements can potentially be waivered by CLIA, if the analytical quality requirement is validated in external proficiency tests [66,67]. Third, as software-based health technologies including mobile applications (also known as “software as a medical device”) are increasingly used by providers and patients, FDA has developed an effective regulatory framework for low-risk mobile medical applications that supports innovation and commercialization of mHealth tools while protecting patient health [68].

5. CONCLUSION

We have developed SSR-based mHealth spectroscopy for noninvasively assessing blood Hgb content. Although the current version of the mHematology application warrants a larger clinical study with a full usability test including a systematic analysis of intra- and inter-blood Hgb measurements, the reported analyses support the feasibility of SSR to be translated into mHealth for noninvasive blood Hgb measurements. The reported SSR algorithm can easily be extended to different models of smartphones by incorporating the spectral response functions in the R, G, and B channels of the built-in camera in a specific model of smartphones. The spectral response functions of a three-color sensor can be obtained from the manufacturer or be measured in the laboratory [47,69]. The mHematology application also serves as an example that a data-science approach or a data-driven technology can minimize hardware complexity. We envision that the mHematology application can potentially be scalable for noninvasive, real-time, and continuous measurements of blood Hgb content on a personal level without relying on centralized clinical laboratories for healthcare services in resource-limited and homecare settings.

Funding

National Institutes of Health (R21TW010620, UL1TR001108); United States Agency for International Development (7200AA18CA00019); Purdue Shah Family Global Innovation Lab.

Acknowledgment

We thank Dr. Abraham Siika for clinical oversight, Cathrine Chiliswa and Kit Yee Yeung for assistance in the clinical study, Mhawila Mhawila and Simon Savai for assistance in mHealth application development, Dr. Sheikh Iqbal Ahamed for advising Kamrul Hasan, and Dr. Raymond Konger for insightful comments from a clinical laboratory perspective. In addition, we thank the AMPATH Kenya program and the clinical laboratory staff for their assistance during the study. The contents of this paper are the responsibility of the authors and do not necessarily reflect the views of the United States Agency for International Development or the United States Government.

Disclosures

SMP, MAV, MMH, and YLK are the inventors of provisional patent applications related to this work that have been filed to the U.S. Patents and Trademark Office by the Purdue Research Foundation (application Nos. 62945816 and 62945808 filed December 10, 2019). YLK is a founding member of HemaChrome, LLC. All other authors declare no conflicts of interest.

 

See Supplement 1 for supporting content.

REFERENCES

1. C. C. W. Hsia, “Mechanisms of disease – respiratory function of hemoglobin,” N. Engl. J. Med. 338, 239–247 (1998). [CrossRef]  

2. J. W. Cannon, “Hemorrhagic shock,” N. Engl. J. Med. 378, 370–379 (2018). [CrossRef]  

3. M. N. Hasan, A. Fraiwan, P. Thota, T. Oginni, G. M. Olanipekun, F. Hassan-Hanga, J. Little, S. K. Obaro, and U. A. Gurkan, “Clinical testing of hemechip in Nigeria for point-of-care screening of sickle cell disease,” Blood 132, 1095 (2018). [CrossRef]  

4. S. E. Juul, R. J. Derman, and M. Auerbach, “Perinatal iron deficiency: implications for mothers and infants,” Neonatology 115, 269–274 (2019). [CrossRef]  

5. L. R. Smart, E. E. Ambrose, K. C. Raphael, A. Hokororo, E. Kamugisha, E. A. Tyburski, W. A. Lam, R. E. Ware, and P. T. McGann, “Simultaneous point-of-care detection of anemia and sickle cell disease in Tanzania: the rapid study,” Ann. Hematol. 97, 239–246 (2018). [CrossRef]  

6. A. Dauvin, C. Donado, P. Bachtiger, K. C. Huang, C. M. Sauer, D. Ramazzotti, M. Bonvini, L. A. Celi, and M. J. Douglas, “Machine learning can accurately predict pre-admission baseline hemoglobin and creatinine in intensive care patients,” NPJ Digit. Med. 2, 116 (2019). [CrossRef]  

7. M. L. Ryan, C. M. Thorson, C. A. Otero, T. Vu, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Initial hematocrit in trauma: a paradigm shift?” J. Trauma Acute Care Surg. 72, 54–60 (2012). [CrossRef]  

8. C. M. Thorson, M. L. Ryan, R. M. Van Haren, R. Pereira, J. Olloqui, C. A. Otero, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Change in hematocrit during trauma assessment predicts bleeding even with ongoing fluid resuscitation,” Am. Surg. 79, 398–406 (2013).

9. K. Vishwanath, R. Gurjar, D. Wolf, S. Riccardi, M. Duggan, and D. King, “Diffuse optical monitoring of peripheral tissues during uncontrolled internal hemorrhage in a porcine model,” Biomed. Opt. Express 9, 569–580 (2018). [CrossRef]  

10. S. Figueiredo, C. Taconet, A. Harrois, S. Hamada, T. Gauss, M. Raux, and J. Duranteau, “How useful are hemoglobin concentration and its variations to predict significant hemorrhage in the early phase of trauma? A multicentric cohort study,” Ann. Intensive Care 8, 76 (2018). [CrossRef]  

11. C. Briggs, S. Kimber, and L. Green, “Where are we at with point-of-care testing in haematology?” Br. J. Haematol. 158, 679–690 (2012). [CrossRef]  

12. T. Guo, R. Patnaik, K. Kuhlmann, A. J. Rai, and S. K. Sia, “Smartphone dongle for simultaneous measurement of hemoglobin concentration and detection of HIV antibodies,” Lab Chip 15, 3514–3520 (2015). [CrossRef]  

13. M. Jaggernath, R. Naicker, S. Madurai, M. A. Brockman, T. Ndung’u, and H. C. Gelderblom, “Diagnostic accuracy of the HemoCue HB 301, STAT-Site MHgb and URIT-12 point-of-care hemoglobin meters in a central laboratory and a community based clinic in Durban, South Africa,” Plos One 11, e0152184 (2016). [CrossRef]  

14. L. J. Moore, C. E. Wade, L. Vincent, J. Podbielski, E. Camp, D. del Junco, H. Radhakrishnan, J. McCarthy, B. Gill, and J. B. Holcomb, “Evaluation of noninvasive hemoglobin measurements in trauma patients,” Am. J. Surg. 206, 1041–1047 (2013). [CrossRef]  

15. S. H. Kim, M. Lilot, L. S. L. Murphy, K. S. Sidhu, Z. X. Yu, J. Rinehart, and M. Cannesson, “Accuracy of continuous noninvasive hemoglobin monitoring: a systematic review and meta-analysis,” Anesth. Analg. 119, 332–346 (2014). [CrossRef]  

16. R. Hiscock, D. Kumar, and S. W. Simmons, “Systematic review and meta-analysis of method comparison studies of masimo pulse co-oximeters (Radical-7 or Pronto-7) and HemoCue absorption spectrometers (B-hemoglobin or 201+) with laboratory haemoglobin estimation,” Anaesth. Intensive Care 43, 341–350 (2015). [CrossRef]  

17. E. J. Wang, W. Li, J. Y. Zhu, R. Rana, and S. N. Patel, “Noninvasive hemoglobin measurement using unmodified smartphone camera and white flash,” in 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (2017), pp. 2333–2336.

18. R. G. Mannino, D. R. Myers, E. A. Tyburski, C. Caruso, J. Boudreaux, T. Leong, G. D. Clifford, and W. A. Lam, “Smartphone app for non-invasive detection of anemia using only patient-sourced photos,” Nat. Commun. 9, 4924 (2018). [CrossRef]  

19. S. C. Mathews, M. J. McShea, C. L. Hanley, A. Ravitz, A. B. Labrique, and A. B. Cohen, “Digital health: a path to validation,” NPJ Digit. Med. 2, 38 (2019). [CrossRef]  

20. T. N. Sheth, N. K. Choudhry, M. Bowes, and A. S. Detsky, “The relation of conjunctival pallor to the presence of anemia,” J. Gen. Intern. Med. 12, 102–106 (1997). [CrossRef]  

21. C. E. Shulman, M. Levene, L. Morison, E. Dorman, N. Peshu, and K. Marsh, “Screening for severe anaemia in pregnancy in Kenya, using pallor examination and self-reported morbidity,” Trans. R. Soc. Trop. Med. Hyg. 95, 250–255 (2001). [CrossRef]  

22. A. Kalantri, M. Karambelkar, R. Joshi, S. Kalantri, and U. Jajoo, “Accuracy and reliability of pallor for detecting anaemia: a hospital-based diagnostic accuracy study,” PLOS One 5, e8545 (2010). [CrossRef]  

23. A. Lima and J. Bakker, “Noninvasive monitoring of peripheral perfusion,” Intensive Care Med. 31, 1316–1326 (2005). [CrossRef]  

24. S. Hutchings, D. N. Naumann, T. Harris, J. Wendon, and M. J. Midwinter, “Observational study of the effects of traumatic injury, haemorrhagic shock and resuscitation on the microcirculation: a protocol for the microshock study,” BMJ Open 6, e010893 (2016). [CrossRef]  

25. E. Abraham, S. E. Fink, D. R. Markle, G. Pinholster, and M. Tsang, “Continuous monitoring of tissue pH with a fiberoptic conjunctival sensor,” Ann. Emerg. Med. 14, 840–844 (1985). [CrossRef]  

26. A. R. Kent, S. H. Elsing, and R. L. Hebert, “Conjunctival vasculature in the assessment of anemia,” Ophthalmology 107, 274–277 (2000). [CrossRef]  

27. J. Y. Lin and D. E. Fisher, “Melanocyte biology and skin pigmentation,” Nature 445, 843–850 (2007). [CrossRef]  

28. M. M. Khansari, M. Tan, P. Karamian, and M. Shahidi, “Inter-visit variability of conjunctival microvascular hemodynamic measurements in healthy and diabetic retinopathy subjects,” Microvasc. Res. 118, 7–11 (2018). [CrossRef]  

29. S. Suner, G. Crawford, J. McMurdy, and G. Jay, “Non-invasive determination of hemoglobin by digital photography of palpebral conjunctiva,” J. Emerg. Med. 33, 105–111 (2007). [CrossRef]  

30. J. McMurdy, G. Jay, S. Suner, and G. Crawford, “Photonics-based in vivo total hemoglobin monitoring and clinical relevance,” J. Biophoton. 2, 277–287 (2009). [CrossRef]  

31. O. Kim, J. McMurdy, G. Jay, C. Lines, G. Crawford, and M. Alber, “Combined reflectance spectroscopy and stochastic modeling approach for noninvasive hemoglobin determination via palpebral conjunctiva,” Physiol. Rep. 2, e00192 (2014). [CrossRef]  

32. T. Kim, S. H. Choi, N. Lambert-Cheatham, Z. Xu, J. E. Kritchevsky, F. R. Bertin, and Y. L. Kim, “Toward laboratory blood test-comparable photometric assessments for anemia in veterinary hematology,” J. Biomed. Opt. 21, 107001 (2016). [CrossRef]  

33. S. Collings, O. Thompson, E. Hirst, L. Goossens, A. George, and R. Weinkove, “Non-invasive detection of anaemia using digital photographs of the conjunctiva,” PLOS One 11, e0153286 (2016). [CrossRef]  

34. G. Dimauro, A. Guarini, D. Caivano, F. Girardi, C. Pasciolla, and A. Iacobazzi, “Detecting clinical signs of anaemia from digital images of the palpebral conjunctiva,” IEEE Access 7, 113488 (2019). [CrossRef]  

35. V. V. Tuchin, Tissue Optics: Light Scattering Methods and Instruments for Medical Diagnosis (SPIE, 2015).

36. I. J. Bigio and S. Fantini, Quantitative Biomedical Optics: Theory, Methods, and Applications (Cambridge University, 2016).

37. T. Vo-Dinh, Biomedical Photonics Handbook: Fundamentals, Devices, and Techniques (CRC Press, 2019).

38. S. Chen and Q. Liu, “Modified Wiener estimation of diffuse reflectance spectra from RGB values by the synthesis of new colors for tissue measurements,” J. Biomed. Opt. 17, 030501 (2012). [CrossRef]  

39. I. Nishidate, T. Maeda, K. Niizeki, and Y. Aizu, “Estimation of melanin and hemoglobin using spectral reflectance images reconstructed from a digital RGB image by the Wiener estimation method,” Sensors 13, 7902–7915 (2013). [CrossRef]  

40. K. Yoshida, I. Nishidate, T. Ishizuka, S. Kawauchi, S. Sato, and M. Satoc, “Multispectral imaging of absorption and scattering properties of in vivo exposed rat brain using a digital red-green-blue camera,” J. Biomed. Opt. 20, 051026 (2015). [CrossRef]  

41. S. C. Yoon, T. S. Shin, K. C. Lawrence, G. W. Heitschmidt, B. Park, and G. R. Gamble, “Hyperspectral imaging using RGB color for foodborne pathogen detection,” J. Electron. Imaging 24, 043008 (2015). [CrossRef]  

42. W. Saafin, M. Vega, R. Molina, and A. K. Katsaggelos, “Compressed sensing super resolution of color images,” in 24th European Signal Processing Conference (2016), pp. 1563–1567.

43. T. Kim, M. A. Visbal-Onufrak, R. L. Konger, and Y. L. Kim, “Data-driven imaging of tissue inflammation using RGB-based hyperspectral reconstruction toward personal monitoring of dermatologic health,” Biomed. Opt. Express 8, 5282–5296 (2017). [CrossRef]  

44. J. X. Liang and X. X. Wan, “Optimized method for spectral reflectance reconstruction from camera responses,” Opt. Express 25, 28273–28287 (2017). [CrossRef]  

45. S. Galliani, C. Lanaras, D. Marmanis, E. Baltsavias, and K. Schindler, “Learned spectral super-resolution,” arXiv:1703.09470 (2017).

46. S. Koundinya, H. Sharma, M. Sharma, A. Upadhyay, R. Manekar, R. Mukhopadhyay, A. Karmakar, and S. Chaudhury, “2D-3D CNN based architectures for spectral reconstruction from RGB images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (2018), pp. 957–964.

47. B. Kaya, Y. B. Can, and R. Timofte, “Towards spectral estimation from a single RGB image in the wild,” arXiv:1812.00805 (2018).

48. B. Arad, O. Ben-Shahar, R. Timofte, L. Van Gool, L. Zhang, and M. Yang, “Ntire 2018 challenge on spectral reconstruction from RGB images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshop (2018), pp. 1042–1051.

49. Y. Yan, L. Zhang, J. Li, W. Wei, and Y. Zhang, “Accurate spectral super-resolution from single RGB image using multi-scale CNN,” in Pattern Recognition and Computer Vision (Springer, 2018), pp. 206–217.

50. A. Signoroni, M. Savardi, A. Baronio, and S. Benini, “Deep learning meets hyperspectral image analysis: a multidisciplinary review,” J. Imaging 5, 52 (2019). [CrossRef]  

51. D. Gedalin, Y. Oiknine, and A. Stern, “DeepCubeNet: reconstruction of spectrally compressive sensed hyperspectral images with deep neural networks,” Opt. Express 27, 35811–35822 (2019). [CrossRef]  

52. K. Gwn Lore, K. K. Reddy, M. Giering, and E. A. Bernal, “Generative adversarial networks for spectral super-resolution and bidirectional RGB-to-multispectral mapping,” in IEEE Conference on Computer Vision and Pattern Recognition Workshops (2019).

53. T. R. Liu, Z. S. Wei, Y. Rivenson, K. de Haan, Y. B. Zhang, Y. C. Wu, and A. Ozcan, “Deep learning-based color holographic microscopy,” J. Biophoton. 12, e201900107 (2019). [CrossRef]  

54. A. Thomas, J. Newton, and M. Oldham, “A method to correct for stray light in telecentric optical-CT imaging of radiochromic dosimeters,” Phys. Med. Biol. 56, 4433–4451 (2011). [CrossRef]  

55. A. Doblas, E. Sanchez-Ortiga, M. Martinez-Corral, G. Saavedra, and J. Garcia-Sucerquia, “Accurate single-shot quantitative phase imaging of biological specimens with telecentric digital holographic microscopy,” J. Biomed. Opt. 19, 46022 (2014). [CrossRef]  

56. R. P. McNabb, P. Challa, A. N. Kuo, and J. A. Izatt, “Complete 360 degrees circumferential gonioscopic optical coherence tomography imaging of the iridocorneal angle,” Biomed. Opt. Express 6, 1376–1391 (2015). [CrossRef]  

57. M. A. Visbal-Onufrak, R. L. Konger, and Y. L. Kim, “Telecentric suppression of diffuse light in imaging of highly anisotropic scattering media,” Opt. Lett. 41, 143–146 (2016). [CrossRef]  

58. A. Krishnan, L. J. Williams, A. R. McIntosh, and H. Abdi, “Partial least squares (PLS) methods for neuroimaging: a tutorial and review,” Neuroimage 56, 455–475 (2011). [CrossRef]  

59. J. L. Martinez, H. Saulo, H. B. Escobar, and J. Leao, “A new model selection criterion for partial least squares regression,” Chemom. Intell. Lab. Syst. 169, 64–78 (2017). [CrossRef]  

60. D. Giavarina, “Understanding Bland Altman analysis,” Biochem. Med. 25, 141–151 (2015). [CrossRef]  

61. T. K. Koo and M. Y. Li, “A guideline of selecting and reporting intraclass correlation coefficients for reliability research,” J. Chiropr. Med. 15, 155–163 (2016). [CrossRef]  

62. J. F. Storz and H. Moriyama, “Mechanisms of hemoglobin adaptation to high altitude hypoxia,” High Alt. Med. Biol. 9, 148–157 (2008). [CrossRef]  

63. J. S. Windsor and G. W. Rodway, “Heights and haematology: the story of haemoglobin at altitude,” Postgrad. Med. J. 83, 148–151 (2007). [CrossRef]  

64. P. J. Diggle, P. Heagerty, K.-Y. Liang, and S. L. Zeger, Analysis of Longitudinal Data (Oxford University, 2002).

65. S. R. Pasricha, K. Colman, E. Centeno-Tablante, M. N. Garcia-Casal, and J. P. Pena-Rosas, “Revisiting WHO haemoglobin thresholds to define anaemia in clinical medicine and public health,” Lancet Haematol. 5, E60–E62 (2018). [CrossRef]  

66. U.S. Government, “Medicare, Medicaid and CLIA programs; regulations implementing the clinical laboratory improvement amendments of 1988 (CLIA)-HCFA. Final rule with comment period,” Fed. Regist. 57, 7002–7186 (1992).

67. U.S. Government, “Clinical laboratory improvement amendments of 1988 (CLIA) proficiency testing regulations related to analytes and acceptable performance,” Fed. Regist. 84, 1536–1567 (2019).

68. J. Shuren, B. Patel, and S. Gottlieb, “FDA regulation of mobile medical apps,” J. Am. Med. Assoc. 320, 337–338 (2018). [CrossRef]  

69. G. Finlayson, M. M. Darrodi, and M. Mackiewicz, “Rank-based camera spectral sensitivity estimation,” J. Opt. Soc. Am. A 33, 589–599 (2016). [CrossRef]  

References

  • View by:
  • Article Order
  • |
  • Year
  • |
  • Author
  • |
  • Publication
  1. C. C. W. Hsia, “Mechanisms of disease – respiratory function of hemoglobin,” N. Engl. J. Med. 338, 239–247 (1998).
    [Crossref]
  2. J. W. Cannon, “Hemorrhagic shock,” N. Engl. J. Med. 378, 370–379 (2018).
    [Crossref]
  3. M. N. Hasan, A. Fraiwan, P. Thota, T. Oginni, G. M. Olanipekun, F. Hassan-Hanga, J. Little, S. K. Obaro, and U. A. Gurkan, “Clinical testing of hemechip in Nigeria for point-of-care screening of sickle cell disease,” Blood 132, 1095 (2018).
    [Crossref]
  4. S. E. Juul, R. J. Derman, and M. Auerbach, “Perinatal iron deficiency: implications for mothers and infants,” Neonatology 115, 269–274 (2019).
    [Crossref]
  5. L. R. Smart, E. E. Ambrose, K. C. Raphael, A. Hokororo, E. Kamugisha, E. A. Tyburski, W. A. Lam, R. E. Ware, and P. T. McGann, “Simultaneous point-of-care detection of anemia and sickle cell disease in Tanzania: the rapid study,” Ann. Hematol. 97, 239–246 (2018).
    [Crossref]
  6. A. Dauvin, C. Donado, P. Bachtiger, K. C. Huang, C. M. Sauer, D. Ramazzotti, M. Bonvini, L. A. Celi, and M. J. Douglas, “Machine learning can accurately predict pre-admission baseline hemoglobin and creatinine in intensive care patients,” NPJ Digit. Med. 2, 116 (2019).
    [Crossref]
  7. M. L. Ryan, C. M. Thorson, C. A. Otero, T. Vu, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Initial hematocrit in trauma: a paradigm shift?” J. Trauma Acute Care Surg. 72, 54–60 (2012).
    [Crossref]
  8. C. M. Thorson, M. L. Ryan, R. M. Van Haren, R. Pereira, J. Olloqui, C. A. Otero, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Change in hematocrit during trauma assessment predicts bleeding even with ongoing fluid resuscitation,” Am. Surg. 79, 398–406 (2013).
  9. K. Vishwanath, R. Gurjar, D. Wolf, S. Riccardi, M. Duggan, and D. King, “Diffuse optical monitoring of peripheral tissues during uncontrolled internal hemorrhage in a porcine model,” Biomed. Opt. Express 9, 569–580 (2018).
    [Crossref]
  10. S. Figueiredo, C. Taconet, A. Harrois, S. Hamada, T. Gauss, M. Raux, and J. Duranteau, “How useful are hemoglobin concentration and its variations to predict significant hemorrhage in the early phase of trauma? A multicentric cohort study,” Ann. Intensive Care 8, 76 (2018).
    [Crossref]
  11. C. Briggs, S. Kimber, and L. Green, “Where are we at with point-of-care testing in haematology?” Br. J. Haematol. 158, 679–690 (2012).
    [Crossref]
  12. T. Guo, R. Patnaik, K. Kuhlmann, A. J. Rai, and S. K. Sia, “Smartphone dongle for simultaneous measurement of hemoglobin concentration and detection of HIV antibodies,” Lab Chip 15, 3514–3520 (2015).
    [Crossref]
  13. M. Jaggernath, R. Naicker, S. Madurai, M. A. Brockman, T. Ndung’u, and H. C. Gelderblom, “Diagnostic accuracy of the HemoCue HB 301, STAT-Site MHgb and URIT-12 point-of-care hemoglobin meters in a central laboratory and a community based clinic in Durban, South Africa,” Plos One 11, e0152184 (2016).
    [Crossref]
  14. L. J. Moore, C. E. Wade, L. Vincent, J. Podbielski, E. Camp, D. del Junco, H. Radhakrishnan, J. McCarthy, B. Gill, and J. B. Holcomb, “Evaluation of noninvasive hemoglobin measurements in trauma patients,” Am. J. Surg. 206, 1041–1047 (2013).
    [Crossref]
  15. S. H. Kim, M. Lilot, L. S. L. Murphy, K. S. Sidhu, Z. X. Yu, J. Rinehart, and M. Cannesson, “Accuracy of continuous noninvasive hemoglobin monitoring: a systematic review and meta-analysis,” Anesth. Analg. 119, 332–346 (2014).
    [Crossref]
  16. R. Hiscock, D. Kumar, and S. W. Simmons, “Systematic review and meta-analysis of method comparison studies of masimo pulse co-oximeters (Radical-7 or Pronto-7) and HemoCue absorption spectrometers (B-hemoglobin or 201+) with laboratory haemoglobin estimation,” Anaesth. Intensive Care 43, 341–350 (2015).
    [Crossref]
  17. E. J. Wang, W. Li, J. Y. Zhu, R. Rana, and S. N. Patel, “Noninvasive hemoglobin measurement using unmodified smartphone camera and white flash,” in 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (2017), pp. 2333–2336.
  18. R. G. Mannino, D. R. Myers, E. A. Tyburski, C. Caruso, J. Boudreaux, T. Leong, G. D. Clifford, and W. A. Lam, “Smartphone app for non-invasive detection of anemia using only patient-sourced photos,” Nat. Commun. 9, 4924 (2018).
    [Crossref]
  19. S. C. Mathews, M. J. McShea, C. L. Hanley, A. Ravitz, A. B. Labrique, and A. B. Cohen, “Digital health: a path to validation,” NPJ Digit. Med. 2, 38 (2019).
    [Crossref]
  20. T. N. Sheth, N. K. Choudhry, M. Bowes, and A. S. Detsky, “The relation of conjunctival pallor to the presence of anemia,” J. Gen. Intern. Med. 12, 102–106 (1997).
    [Crossref]
  21. C. E. Shulman, M. Levene, L. Morison, E. Dorman, N. Peshu, and K. Marsh, “Screening for severe anaemia in pregnancy in Kenya, using pallor examination and self-reported morbidity,” Trans. R. Soc. Trop. Med. Hyg. 95, 250–255 (2001).
    [Crossref]
  22. A. Kalantri, M. Karambelkar, R. Joshi, S. Kalantri, and U. Jajoo, “Accuracy and reliability of pallor for detecting anaemia: a hospital-based diagnostic accuracy study,” PLOS One 5, e8545 (2010).
    [Crossref]
  23. A. Lima and J. Bakker, “Noninvasive monitoring of peripheral perfusion,” Intensive Care Med. 31, 1316–1326 (2005).
    [Crossref]
  24. S. Hutchings, D. N. Naumann, T. Harris, J. Wendon, and M. J. Midwinter, “Observational study of the effects of traumatic injury, haemorrhagic shock and resuscitation on the microcirculation: a protocol for the microshock study,” BMJ Open 6, e010893 (2016).
    [Crossref]
  25. E. Abraham, S. E. Fink, D. R. Markle, G. Pinholster, and M. Tsang, “Continuous monitoring of tissue pH with a fiberoptic conjunctival sensor,” Ann. Emerg. Med. 14, 840–844 (1985).
    [Crossref]
  26. A. R. Kent, S. H. Elsing, and R. L. Hebert, “Conjunctival vasculature in the assessment of anemia,” Ophthalmology 107, 274–277 (2000).
    [Crossref]
  27. J. Y. Lin and D. E. Fisher, “Melanocyte biology and skin pigmentation,” Nature 445, 843–850 (2007).
    [Crossref]
  28. M. M. Khansari, M. Tan, P. Karamian, and M. Shahidi, “Inter-visit variability of conjunctival microvascular hemodynamic measurements in healthy and diabetic retinopathy subjects,” Microvasc. Res. 118, 7–11 (2018).
    [Crossref]
  29. S. Suner, G. Crawford, J. McMurdy, and G. Jay, “Non-invasive determination of hemoglobin by digital photography of palpebral conjunctiva,” J. Emerg. Med. 33, 105–111 (2007).
    [Crossref]
  30. J. McMurdy, G. Jay, S. Suner, and G. Crawford, “Photonics-based in vivo total hemoglobin monitoring and clinical relevance,” J. Biophoton. 2, 277–287 (2009).
    [Crossref]
  31. O. Kim, J. McMurdy, G. Jay, C. Lines, G. Crawford, and M. Alber, “Combined reflectance spectroscopy and stochastic modeling approach for noninvasive hemoglobin determination via palpebral conjunctiva,” Physiol. Rep. 2, e00192 (2014).
    [Crossref]
  32. T. Kim, S. H. Choi, N. Lambert-Cheatham, Z. Xu, J. E. Kritchevsky, F. R. Bertin, and Y. L. Kim, “Toward laboratory blood test-comparable photometric assessments for anemia in veterinary hematology,” J. Biomed. Opt. 21, 107001 (2016).
    [Crossref]
  33. S. Collings, O. Thompson, E. Hirst, L. Goossens, A. George, and R. Weinkove, “Non-invasive detection of anaemia using digital photographs of the conjunctiva,” PLOS One 11, e0153286 (2016).
    [Crossref]
  34. G. Dimauro, A. Guarini, D. Caivano, F. Girardi, C. Pasciolla, and A. Iacobazzi, “Detecting clinical signs of anaemia from digital images of the palpebral conjunctiva,” IEEE Access 7, 113488 (2019).
    [Crossref]
  35. V. V. Tuchin, Tissue Optics: Light Scattering Methods and Instruments for Medical Diagnosis (SPIE, 2015).
  36. I. J. Bigio and S. Fantini, Quantitative Biomedical Optics: Theory, Methods, and Applications (Cambridge University, 2016).
  37. T. Vo-Dinh, Biomedical Photonics Handbook: Fundamentals, Devices, and Techniques (CRC Press, 2019).
  38. S. Chen and Q. Liu, “Modified Wiener estimation of diffuse reflectance spectra from RGB values by the synthesis of new colors for tissue measurements,” J. Biomed. Opt. 17, 030501 (2012).
    [Crossref]
  39. I. Nishidate, T. Maeda, K. Niizeki, and Y. Aizu, “Estimation of melanin and hemoglobin using spectral reflectance images reconstructed from a digital RGB image by the Wiener estimation method,” Sensors 13, 7902–7915 (2013).
    [Crossref]
  40. K. Yoshida, I. Nishidate, T. Ishizuka, S. Kawauchi, S. Sato, and M. Satoc, “Multispectral imaging of absorption and scattering properties of in vivo exposed rat brain using a digital red-green-blue camera,” J. Biomed. Opt. 20, 051026 (2015).
    [Crossref]
  41. S. C. Yoon, T. S. Shin, K. C. Lawrence, G. W. Heitschmidt, B. Park, and G. R. Gamble, “Hyperspectral imaging using RGB color for foodborne pathogen detection,” J. Electron. Imaging 24, 043008 (2015).
    [Crossref]
  42. W. Saafin, M. Vega, R. Molina, and A. K. Katsaggelos, “Compressed sensing super resolution of color images,” in 24th European Signal Processing Conference (2016), pp. 1563–1567.
  43. T. Kim, M. A. Visbal-Onufrak, R. L. Konger, and Y. L. Kim, “Data-driven imaging of tissue inflammation using RGB-based hyperspectral reconstruction toward personal monitoring of dermatologic health,” Biomed. Opt. Express 8, 5282–5296 (2017).
    [Crossref]
  44. J. X. Liang and X. X. Wan, “Optimized method for spectral reflectance reconstruction from camera responses,” Opt. Express 25, 28273–28287 (2017).
    [Crossref]
  45. S. Galliani, C. Lanaras, D. Marmanis, E. Baltsavias, and K. Schindler, “Learned spectral super-resolution,” arXiv:1703.09470 (2017).
  46. S. Koundinya, H. Sharma, M. Sharma, A. Upadhyay, R. Manekar, R. Mukhopadhyay, A. Karmakar, and S. Chaudhury, “2D-3D CNN based architectures for spectral reconstruction from RGB images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (2018), pp. 957–964.
  47. B. Kaya, Y. B. Can, and R. Timofte, “Towards spectral estimation from a single RGB image in the wild,” arXiv:1812.00805 (2018).
  48. B. Arad, O. Ben-Shahar, R. Timofte, L. Van Gool, L. Zhang, and M. Yang, “Ntire 2018 challenge on spectral reconstruction from RGB images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshop (2018), pp. 1042–1051.
  49. Y. Yan, L. Zhang, J. Li, W. Wei, and Y. Zhang, “Accurate spectral super-resolution from single RGB image using multi-scale CNN,” in Pattern Recognition and Computer Vision (Springer, 2018), pp. 206–217.
  50. A. Signoroni, M. Savardi, A. Baronio, and S. Benini, “Deep learning meets hyperspectral image analysis: a multidisciplinary review,” J. Imaging 5, 52 (2019).
    [Crossref]
  51. D. Gedalin, Y. Oiknine, and A. Stern, “DeepCubeNet: reconstruction of spectrally compressive sensed hyperspectral images with deep neural networks,” Opt. Express 27, 35811–35822 (2019).
    [Crossref]
  52. K. Gwn Lore, K. K. Reddy, M. Giering, and E. A. Bernal, “Generative adversarial networks for spectral super-resolution and bidirectional RGB-to-multispectral mapping,” in IEEE Conference on Computer Vision and Pattern Recognition Workshops (2019).
  53. T. R. Liu, Z. S. Wei, Y. Rivenson, K. de Haan, Y. B. Zhang, Y. C. Wu, and A. Ozcan, “Deep learning-based color holographic microscopy,” J. Biophoton. 12, e201900107 (2019).
    [Crossref]
  54. A. Thomas, J. Newton, and M. Oldham, “A method to correct for stray light in telecentric optical-CT imaging of radiochromic dosimeters,” Phys. Med. Biol. 56, 4433–4451 (2011).
    [Crossref]
  55. A. Doblas, E. Sanchez-Ortiga, M. Martinez-Corral, G. Saavedra, and J. Garcia-Sucerquia, “Accurate single-shot quantitative phase imaging of biological specimens with telecentric digital holographic microscopy,” J. Biomed. Opt. 19, 46022 (2014).
    [Crossref]
  56. R. P. McNabb, P. Challa, A. N. Kuo, and J. A. Izatt, “Complete 360 degrees circumferential gonioscopic optical coherence tomography imaging of the iridocorneal angle,” Biomed. Opt. Express 6, 1376–1391 (2015).
    [Crossref]
  57. M. A. Visbal-Onufrak, R. L. Konger, and Y. L. Kim, “Telecentric suppression of diffuse light in imaging of highly anisotropic scattering media,” Opt. Lett. 41, 143–146 (2016).
    [Crossref]
  58. A. Krishnan, L. J. Williams, A. R. McIntosh, and H. Abdi, “Partial least squares (PLS) methods for neuroimaging: a tutorial and review,” Neuroimage 56, 455–475 (2011).
    [Crossref]
  59. J. L. Martinez, H. Saulo, H. B. Escobar, and J. Leao, “A new model selection criterion for partial least squares regression,” Chemom. Intell. Lab. Syst. 169, 64–78 (2017).
    [Crossref]
  60. D. Giavarina, “Understanding Bland Altman analysis,” Biochem. Med. 25, 141–151 (2015).
    [Crossref]
  61. T. K. Koo and M. Y. Li, “A guideline of selecting and reporting intraclass correlation coefficients for reliability research,” J. Chiropr. Med. 15, 155–163 (2016).
    [Crossref]
  62. J. F. Storz and H. Moriyama, “Mechanisms of hemoglobin adaptation to high altitude hypoxia,” High Alt. Med. Biol. 9, 148–157 (2008).
    [Crossref]
  63. J. S. Windsor and G. W. Rodway, “Heights and haematology: the story of haemoglobin at altitude,” Postgrad. Med. J. 83, 148–151 (2007).
    [Crossref]
  64. P. J. Diggle, P. Heagerty, K.-Y. Liang, and S. L. Zeger, Analysis of Longitudinal Data (Oxford University, 2002).
  65. S. R. Pasricha, K. Colman, E. Centeno-Tablante, M. N. Garcia-Casal, and J. P. Pena-Rosas, “Revisiting WHO haemoglobin thresholds to define anaemia in clinical medicine and public health,” Lancet Haematol. 5, E60–E62 (2018).
    [Crossref]
  66. U.S. Government, “Medicare, Medicaid and CLIA programs; regulations implementing the clinical laboratory improvement amendments of 1988 (CLIA)-HCFA. Final rule with comment period,” Fed. Regist. 57, 7002–7186 (1992).
  67. U.S. Government, “Clinical laboratory improvement amendments of 1988 (CLIA) proficiency testing regulations related to analytes and acceptable performance,” Fed. Regist. 84, 1536–1567 (2019).
  68. J. Shuren, B. Patel, and S. Gottlieb, “FDA regulation of mobile medical apps,” J. Am. Med. Assoc. 320, 337–338 (2018).
    [Crossref]
  69. G. Finlayson, M. M. Darrodi, and M. Mackiewicz, “Rank-based camera spectral sensitivity estimation,” J. Opt. Soc. Am. A 33, 589–599 (2016).
    [Crossref]

2019 (8)

S. E. Juul, R. J. Derman, and M. Auerbach, “Perinatal iron deficiency: implications for mothers and infants,” Neonatology 115, 269–274 (2019).
[Crossref]

A. Dauvin, C. Donado, P. Bachtiger, K. C. Huang, C. M. Sauer, D. Ramazzotti, M. Bonvini, L. A. Celi, and M. J. Douglas, “Machine learning can accurately predict pre-admission baseline hemoglobin and creatinine in intensive care patients,” NPJ Digit. Med. 2, 116 (2019).
[Crossref]

S. C. Mathews, M. J. McShea, C. L. Hanley, A. Ravitz, A. B. Labrique, and A. B. Cohen, “Digital health: a path to validation,” NPJ Digit. Med. 2, 38 (2019).
[Crossref]

G. Dimauro, A. Guarini, D. Caivano, F. Girardi, C. Pasciolla, and A. Iacobazzi, “Detecting clinical signs of anaemia from digital images of the palpebral conjunctiva,” IEEE Access 7, 113488 (2019).
[Crossref]

A. Signoroni, M. Savardi, A. Baronio, and S. Benini, “Deep learning meets hyperspectral image analysis: a multidisciplinary review,” J. Imaging 5, 52 (2019).
[Crossref]

D. Gedalin, Y. Oiknine, and A. Stern, “DeepCubeNet: reconstruction of spectrally compressive sensed hyperspectral images with deep neural networks,” Opt. Express 27, 35811–35822 (2019).
[Crossref]

T. R. Liu, Z. S. Wei, Y. Rivenson, K. de Haan, Y. B. Zhang, Y. C. Wu, and A. Ozcan, “Deep learning-based color holographic microscopy,” J. Biophoton. 12, e201900107 (2019).
[Crossref]

U.S. Government, “Clinical laboratory improvement amendments of 1988 (CLIA) proficiency testing regulations related to analytes and acceptable performance,” Fed. Regist. 84, 1536–1567 (2019).

2018 (9)

J. Shuren, B. Patel, and S. Gottlieb, “FDA regulation of mobile medical apps,” J. Am. Med. Assoc. 320, 337–338 (2018).
[Crossref]

S. R. Pasricha, K. Colman, E. Centeno-Tablante, M. N. Garcia-Casal, and J. P. Pena-Rosas, “Revisiting WHO haemoglobin thresholds to define anaemia in clinical medicine and public health,” Lancet Haematol. 5, E60–E62 (2018).
[Crossref]

M. M. Khansari, M. Tan, P. Karamian, and M. Shahidi, “Inter-visit variability of conjunctival microvascular hemodynamic measurements in healthy and diabetic retinopathy subjects,” Microvasc. Res. 118, 7–11 (2018).
[Crossref]

L. R. Smart, E. E. Ambrose, K. C. Raphael, A. Hokororo, E. Kamugisha, E. A. Tyburski, W. A. Lam, R. E. Ware, and P. T. McGann, “Simultaneous point-of-care detection of anemia and sickle cell disease in Tanzania: the rapid study,” Ann. Hematol. 97, 239–246 (2018).
[Crossref]

J. W. Cannon, “Hemorrhagic shock,” N. Engl. J. Med. 378, 370–379 (2018).
[Crossref]

M. N. Hasan, A. Fraiwan, P. Thota, T. Oginni, G. M. Olanipekun, F. Hassan-Hanga, J. Little, S. K. Obaro, and U. A. Gurkan, “Clinical testing of hemechip in Nigeria for point-of-care screening of sickle cell disease,” Blood 132, 1095 (2018).
[Crossref]

K. Vishwanath, R. Gurjar, D. Wolf, S. Riccardi, M. Duggan, and D. King, “Diffuse optical monitoring of peripheral tissues during uncontrolled internal hemorrhage in a porcine model,” Biomed. Opt. Express 9, 569–580 (2018).
[Crossref]

S. Figueiredo, C. Taconet, A. Harrois, S. Hamada, T. Gauss, M. Raux, and J. Duranteau, “How useful are hemoglobin concentration and its variations to predict significant hemorrhage in the early phase of trauma? A multicentric cohort study,” Ann. Intensive Care 8, 76 (2018).
[Crossref]

R. G. Mannino, D. R. Myers, E. A. Tyburski, C. Caruso, J. Boudreaux, T. Leong, G. D. Clifford, and W. A. Lam, “Smartphone app for non-invasive detection of anemia using only patient-sourced photos,” Nat. Commun. 9, 4924 (2018).
[Crossref]

2017 (3)

2016 (7)

M. A. Visbal-Onufrak, R. L. Konger, and Y. L. Kim, “Telecentric suppression of diffuse light in imaging of highly anisotropic scattering media,” Opt. Lett. 41, 143–146 (2016).
[Crossref]

T. K. Koo and M. Y. Li, “A guideline of selecting and reporting intraclass correlation coefficients for reliability research,” J. Chiropr. Med. 15, 155–163 (2016).
[Crossref]

G. Finlayson, M. M. Darrodi, and M. Mackiewicz, “Rank-based camera spectral sensitivity estimation,” J. Opt. Soc. Am. A 33, 589–599 (2016).
[Crossref]

M. Jaggernath, R. Naicker, S. Madurai, M. A. Brockman, T. Ndung’u, and H. C. Gelderblom, “Diagnostic accuracy of the HemoCue HB 301, STAT-Site MHgb and URIT-12 point-of-care hemoglobin meters in a central laboratory and a community based clinic in Durban, South Africa,” Plos One 11, e0152184 (2016).
[Crossref]

T. Kim, S. H. Choi, N. Lambert-Cheatham, Z. Xu, J. E. Kritchevsky, F. R. Bertin, and Y. L. Kim, “Toward laboratory blood test-comparable photometric assessments for anemia in veterinary hematology,” J. Biomed. Opt. 21, 107001 (2016).
[Crossref]

S. Collings, O. Thompson, E. Hirst, L. Goossens, A. George, and R. Weinkove, “Non-invasive detection of anaemia using digital photographs of the conjunctiva,” PLOS One 11, e0153286 (2016).
[Crossref]

S. Hutchings, D. N. Naumann, T. Harris, J. Wendon, and M. J. Midwinter, “Observational study of the effects of traumatic injury, haemorrhagic shock and resuscitation on the microcirculation: a protocol for the microshock study,” BMJ Open 6, e010893 (2016).
[Crossref]

2015 (6)

R. Hiscock, D. Kumar, and S. W. Simmons, “Systematic review and meta-analysis of method comparison studies of masimo pulse co-oximeters (Radical-7 or Pronto-7) and HemoCue absorption spectrometers (B-hemoglobin or 201+) with laboratory haemoglobin estimation,” Anaesth. Intensive Care 43, 341–350 (2015).
[Crossref]

T. Guo, R. Patnaik, K. Kuhlmann, A. J. Rai, and S. K. Sia, “Smartphone dongle for simultaneous measurement of hemoglobin concentration and detection of HIV antibodies,” Lab Chip 15, 3514–3520 (2015).
[Crossref]

D. Giavarina, “Understanding Bland Altman analysis,” Biochem. Med. 25, 141–151 (2015).
[Crossref]

R. P. McNabb, P. Challa, A. N. Kuo, and J. A. Izatt, “Complete 360 degrees circumferential gonioscopic optical coherence tomography imaging of the iridocorneal angle,” Biomed. Opt. Express 6, 1376–1391 (2015).
[Crossref]

K. Yoshida, I. Nishidate, T. Ishizuka, S. Kawauchi, S. Sato, and M. Satoc, “Multispectral imaging of absorption and scattering properties of in vivo exposed rat brain using a digital red-green-blue camera,” J. Biomed. Opt. 20, 051026 (2015).
[Crossref]

S. C. Yoon, T. S. Shin, K. C. Lawrence, G. W. Heitschmidt, B. Park, and G. R. Gamble, “Hyperspectral imaging using RGB color for foodborne pathogen detection,” J. Electron. Imaging 24, 043008 (2015).
[Crossref]

2014 (3)

O. Kim, J. McMurdy, G. Jay, C. Lines, G. Crawford, and M. Alber, “Combined reflectance spectroscopy and stochastic modeling approach for noninvasive hemoglobin determination via palpebral conjunctiva,” Physiol. Rep. 2, e00192 (2014).
[Crossref]

A. Doblas, E. Sanchez-Ortiga, M. Martinez-Corral, G. Saavedra, and J. Garcia-Sucerquia, “Accurate single-shot quantitative phase imaging of biological specimens with telecentric digital holographic microscopy,” J. Biomed. Opt. 19, 46022 (2014).
[Crossref]

S. H. Kim, M. Lilot, L. S. L. Murphy, K. S. Sidhu, Z. X. Yu, J. Rinehart, and M. Cannesson, “Accuracy of continuous noninvasive hemoglobin monitoring: a systematic review and meta-analysis,” Anesth. Analg. 119, 332–346 (2014).
[Crossref]

2013 (3)

L. J. Moore, C. E. Wade, L. Vincent, J. Podbielski, E. Camp, D. del Junco, H. Radhakrishnan, J. McCarthy, B. Gill, and J. B. Holcomb, “Evaluation of noninvasive hemoglobin measurements in trauma patients,” Am. J. Surg. 206, 1041–1047 (2013).
[Crossref]

C. M. Thorson, M. L. Ryan, R. M. Van Haren, R. Pereira, J. Olloqui, C. A. Otero, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Change in hematocrit during trauma assessment predicts bleeding even with ongoing fluid resuscitation,” Am. Surg. 79, 398–406 (2013).

I. Nishidate, T. Maeda, K. Niizeki, and Y. Aizu, “Estimation of melanin and hemoglobin using spectral reflectance images reconstructed from a digital RGB image by the Wiener estimation method,” Sensors 13, 7902–7915 (2013).
[Crossref]

2012 (3)

M. L. Ryan, C. M. Thorson, C. A. Otero, T. Vu, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Initial hematocrit in trauma: a paradigm shift?” J. Trauma Acute Care Surg. 72, 54–60 (2012).
[Crossref]

C. Briggs, S. Kimber, and L. Green, “Where are we at with point-of-care testing in haematology?” Br. J. Haematol. 158, 679–690 (2012).
[Crossref]

S. Chen and Q. Liu, “Modified Wiener estimation of diffuse reflectance spectra from RGB values by the synthesis of new colors for tissue measurements,” J. Biomed. Opt. 17, 030501 (2012).
[Crossref]

2011 (2)

A. Thomas, J. Newton, and M. Oldham, “A method to correct for stray light in telecentric optical-CT imaging of radiochromic dosimeters,” Phys. Med. Biol. 56, 4433–4451 (2011).
[Crossref]

A. Krishnan, L. J. Williams, A. R. McIntosh, and H. Abdi, “Partial least squares (PLS) methods for neuroimaging: a tutorial and review,” Neuroimage 56, 455–475 (2011).
[Crossref]

2010 (1)

A. Kalantri, M. Karambelkar, R. Joshi, S. Kalantri, and U. Jajoo, “Accuracy and reliability of pallor for detecting anaemia: a hospital-based diagnostic accuracy study,” PLOS One 5, e8545 (2010).
[Crossref]

2009 (1)

J. McMurdy, G. Jay, S. Suner, and G. Crawford, “Photonics-based in vivo total hemoglobin monitoring and clinical relevance,” J. Biophoton. 2, 277–287 (2009).
[Crossref]

2008 (1)

J. F. Storz and H. Moriyama, “Mechanisms of hemoglobin adaptation to high altitude hypoxia,” High Alt. Med. Biol. 9, 148–157 (2008).
[Crossref]

2007 (3)

J. S. Windsor and G. W. Rodway, “Heights and haematology: the story of haemoglobin at altitude,” Postgrad. Med. J. 83, 148–151 (2007).
[Crossref]

J. Y. Lin and D. E. Fisher, “Melanocyte biology and skin pigmentation,” Nature 445, 843–850 (2007).
[Crossref]

S. Suner, G. Crawford, J. McMurdy, and G. Jay, “Non-invasive determination of hemoglobin by digital photography of palpebral conjunctiva,” J. Emerg. Med. 33, 105–111 (2007).
[Crossref]

2005 (1)

A. Lima and J. Bakker, “Noninvasive monitoring of peripheral perfusion,” Intensive Care Med. 31, 1316–1326 (2005).
[Crossref]

2001 (1)

C. E. Shulman, M. Levene, L. Morison, E. Dorman, N. Peshu, and K. Marsh, “Screening for severe anaemia in pregnancy in Kenya, using pallor examination and self-reported morbidity,” Trans. R. Soc. Trop. Med. Hyg. 95, 250–255 (2001).
[Crossref]

2000 (1)

A. R. Kent, S. H. Elsing, and R. L. Hebert, “Conjunctival vasculature in the assessment of anemia,” Ophthalmology 107, 274–277 (2000).
[Crossref]

1998 (1)

C. C. W. Hsia, “Mechanisms of disease – respiratory function of hemoglobin,” N. Engl. J. Med. 338, 239–247 (1998).
[Crossref]

1997 (1)

T. N. Sheth, N. K. Choudhry, M. Bowes, and A. S. Detsky, “The relation of conjunctival pallor to the presence of anemia,” J. Gen. Intern. Med. 12, 102–106 (1997).
[Crossref]

1992 (1)

U.S. Government, “Medicare, Medicaid and CLIA programs; regulations implementing the clinical laboratory improvement amendments of 1988 (CLIA)-HCFA. Final rule with comment period,” Fed. Regist. 57, 7002–7186 (1992).

1985 (1)

E. Abraham, S. E. Fink, D. R. Markle, G. Pinholster, and M. Tsang, “Continuous monitoring of tissue pH with a fiberoptic conjunctival sensor,” Ann. Emerg. Med. 14, 840–844 (1985).
[Crossref]

Abdi, H.

A. Krishnan, L. J. Williams, A. R. McIntosh, and H. Abdi, “Partial least squares (PLS) methods for neuroimaging: a tutorial and review,” Neuroimage 56, 455–475 (2011).
[Crossref]

Abraham, E.

E. Abraham, S. E. Fink, D. R. Markle, G. Pinholster, and M. Tsang, “Continuous monitoring of tissue pH with a fiberoptic conjunctival sensor,” Ann. Emerg. Med. 14, 840–844 (1985).
[Crossref]

Aizu, Y.

I. Nishidate, T. Maeda, K. Niizeki, and Y. Aizu, “Estimation of melanin and hemoglobin using spectral reflectance images reconstructed from a digital RGB image by the Wiener estimation method,” Sensors 13, 7902–7915 (2013).
[Crossref]

Alber, M.

O. Kim, J. McMurdy, G. Jay, C. Lines, G. Crawford, and M. Alber, “Combined reflectance spectroscopy and stochastic modeling approach for noninvasive hemoglobin determination via palpebral conjunctiva,” Physiol. Rep. 2, e00192 (2014).
[Crossref]

Ambrose, E. E.

L. R. Smart, E. E. Ambrose, K. C. Raphael, A. Hokororo, E. Kamugisha, E. A. Tyburski, W. A. Lam, R. E. Ware, and P. T. McGann, “Simultaneous point-of-care detection of anemia and sickle cell disease in Tanzania: the rapid study,” Ann. Hematol. 97, 239–246 (2018).
[Crossref]

Arad, B.

B. Arad, O. Ben-Shahar, R. Timofte, L. Van Gool, L. Zhang, and M. Yang, “Ntire 2018 challenge on spectral reconstruction from RGB images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshop (2018), pp. 1042–1051.

Auerbach, M.

S. E. Juul, R. J. Derman, and M. Auerbach, “Perinatal iron deficiency: implications for mothers and infants,” Neonatology 115, 269–274 (2019).
[Crossref]

Bachtiger, P.

A. Dauvin, C. Donado, P. Bachtiger, K. C. Huang, C. M. Sauer, D. Ramazzotti, M. Bonvini, L. A. Celi, and M. J. Douglas, “Machine learning can accurately predict pre-admission baseline hemoglobin and creatinine in intensive care patients,” NPJ Digit. Med. 2, 116 (2019).
[Crossref]

Bakker, J.

A. Lima and J. Bakker, “Noninvasive monitoring of peripheral perfusion,” Intensive Care Med. 31, 1316–1326 (2005).
[Crossref]

Baltsavias, E.

S. Galliani, C. Lanaras, D. Marmanis, E. Baltsavias, and K. Schindler, “Learned spectral super-resolution,” arXiv:1703.09470 (2017).

Baronio, A.

A. Signoroni, M. Savardi, A. Baronio, and S. Benini, “Deep learning meets hyperspectral image analysis: a multidisciplinary review,” J. Imaging 5, 52 (2019).
[Crossref]

Benini, S.

A. Signoroni, M. Savardi, A. Baronio, and S. Benini, “Deep learning meets hyperspectral image analysis: a multidisciplinary review,” J. Imaging 5, 52 (2019).
[Crossref]

Ben-Shahar, O.

B. Arad, O. Ben-Shahar, R. Timofte, L. Van Gool, L. Zhang, and M. Yang, “Ntire 2018 challenge on spectral reconstruction from RGB images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshop (2018), pp. 1042–1051.

Bernal, E. A.

K. Gwn Lore, K. K. Reddy, M. Giering, and E. A. Bernal, “Generative adversarial networks for spectral super-resolution and bidirectional RGB-to-multispectral mapping,” in IEEE Conference on Computer Vision and Pattern Recognition Workshops (2019).

Bertin, F. R.

T. Kim, S. H. Choi, N. Lambert-Cheatham, Z. Xu, J. E. Kritchevsky, F. R. Bertin, and Y. L. Kim, “Toward laboratory blood test-comparable photometric assessments for anemia in veterinary hematology,” J. Biomed. Opt. 21, 107001 (2016).
[Crossref]

Bigio, I. J.

I. J. Bigio and S. Fantini, Quantitative Biomedical Optics: Theory, Methods, and Applications (Cambridge University, 2016).

Bonvini, M.

A. Dauvin, C. Donado, P. Bachtiger, K. C. Huang, C. M. Sauer, D. Ramazzotti, M. Bonvini, L. A. Celi, and M. J. Douglas, “Machine learning can accurately predict pre-admission baseline hemoglobin and creatinine in intensive care patients,” NPJ Digit. Med. 2, 116 (2019).
[Crossref]

Boudreaux, J.

R. G. Mannino, D. R. Myers, E. A. Tyburski, C. Caruso, J. Boudreaux, T. Leong, G. D. Clifford, and W. A. Lam, “Smartphone app for non-invasive detection of anemia using only patient-sourced photos,” Nat. Commun. 9, 4924 (2018).
[Crossref]

Bowes, M.

T. N. Sheth, N. K. Choudhry, M. Bowes, and A. S. Detsky, “The relation of conjunctival pallor to the presence of anemia,” J. Gen. Intern. Med. 12, 102–106 (1997).
[Crossref]

Briggs, C.

C. Briggs, S. Kimber, and L. Green, “Where are we at with point-of-care testing in haematology?” Br. J. Haematol. 158, 679–690 (2012).
[Crossref]

Brockman, M. A.

M. Jaggernath, R. Naicker, S. Madurai, M. A. Brockman, T. Ndung’u, and H. C. Gelderblom, “Diagnostic accuracy of the HemoCue HB 301, STAT-Site MHgb and URIT-12 point-of-care hemoglobin meters in a central laboratory and a community based clinic in Durban, South Africa,” Plos One 11, e0152184 (2016).
[Crossref]

Caivano, D.

G. Dimauro, A. Guarini, D. Caivano, F. Girardi, C. Pasciolla, and A. Iacobazzi, “Detecting clinical signs of anaemia from digital images of the palpebral conjunctiva,” IEEE Access 7, 113488 (2019).
[Crossref]

Camp, E.

L. J. Moore, C. E. Wade, L. Vincent, J. Podbielski, E. Camp, D. del Junco, H. Radhakrishnan, J. McCarthy, B. Gill, and J. B. Holcomb, “Evaluation of noninvasive hemoglobin measurements in trauma patients,” Am. J. Surg. 206, 1041–1047 (2013).
[Crossref]

Can, Y. B.

B. Kaya, Y. B. Can, and R. Timofte, “Towards spectral estimation from a single RGB image in the wild,” arXiv:1812.00805 (2018).

Cannesson, M.

S. H. Kim, M. Lilot, L. S. L. Murphy, K. S. Sidhu, Z. X. Yu, J. Rinehart, and M. Cannesson, “Accuracy of continuous noninvasive hemoglobin monitoring: a systematic review and meta-analysis,” Anesth. Analg. 119, 332–346 (2014).
[Crossref]

Cannon, J. W.

J. W. Cannon, “Hemorrhagic shock,” N. Engl. J. Med. 378, 370–379 (2018).
[Crossref]

Caruso, C.

R. G. Mannino, D. R. Myers, E. A. Tyburski, C. Caruso, J. Boudreaux, T. Leong, G. D. Clifford, and W. A. Lam, “Smartphone app for non-invasive detection of anemia using only patient-sourced photos,” Nat. Commun. 9, 4924 (2018).
[Crossref]

Celi, L. A.

A. Dauvin, C. Donado, P. Bachtiger, K. C. Huang, C. M. Sauer, D. Ramazzotti, M. Bonvini, L. A. Celi, and M. J. Douglas, “Machine learning can accurately predict pre-admission baseline hemoglobin and creatinine in intensive care patients,” NPJ Digit. Med. 2, 116 (2019).
[Crossref]

Centeno-Tablante, E.

S. R. Pasricha, K. Colman, E. Centeno-Tablante, M. N. Garcia-Casal, and J. P. Pena-Rosas, “Revisiting WHO haemoglobin thresholds to define anaemia in clinical medicine and public health,” Lancet Haematol. 5, E60–E62 (2018).
[Crossref]

Challa, P.

Chaudhury, S.

S. Koundinya, H. Sharma, M. Sharma, A. Upadhyay, R. Manekar, R. Mukhopadhyay, A. Karmakar, and S. Chaudhury, “2D-3D CNN based architectures for spectral reconstruction from RGB images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (2018), pp. 957–964.

Chen, S.

S. Chen and Q. Liu, “Modified Wiener estimation of diffuse reflectance spectra from RGB values by the synthesis of new colors for tissue measurements,” J. Biomed. Opt. 17, 030501 (2012).
[Crossref]

Choi, S. H.

T. Kim, S. H. Choi, N. Lambert-Cheatham, Z. Xu, J. E. Kritchevsky, F. R. Bertin, and Y. L. Kim, “Toward laboratory blood test-comparable photometric assessments for anemia in veterinary hematology,” J. Biomed. Opt. 21, 107001 (2016).
[Crossref]

Choudhry, N. K.

T. N. Sheth, N. K. Choudhry, M. Bowes, and A. S. Detsky, “The relation of conjunctival pallor to the presence of anemia,” J. Gen. Intern. Med. 12, 102–106 (1997).
[Crossref]

Clifford, G. D.

R. G. Mannino, D. R. Myers, E. A. Tyburski, C. Caruso, J. Boudreaux, T. Leong, G. D. Clifford, and W. A. Lam, “Smartphone app for non-invasive detection of anemia using only patient-sourced photos,” Nat. Commun. 9, 4924 (2018).
[Crossref]

Cohen, A. B.

S. C. Mathews, M. J. McShea, C. L. Hanley, A. Ravitz, A. B. Labrique, and A. B. Cohen, “Digital health: a path to validation,” NPJ Digit. Med. 2, 38 (2019).
[Crossref]

Collings, S.

S. Collings, O. Thompson, E. Hirst, L. Goossens, A. George, and R. Weinkove, “Non-invasive detection of anaemia using digital photographs of the conjunctiva,” PLOS One 11, e0153286 (2016).
[Crossref]

Colman, K.

S. R. Pasricha, K. Colman, E. Centeno-Tablante, M. N. Garcia-Casal, and J. P. Pena-Rosas, “Revisiting WHO haemoglobin thresholds to define anaemia in clinical medicine and public health,” Lancet Haematol. 5, E60–E62 (2018).
[Crossref]

Crawford, G.

O. Kim, J. McMurdy, G. Jay, C. Lines, G. Crawford, and M. Alber, “Combined reflectance spectroscopy and stochastic modeling approach for noninvasive hemoglobin determination via palpebral conjunctiva,” Physiol. Rep. 2, e00192 (2014).
[Crossref]

J. McMurdy, G. Jay, S. Suner, and G. Crawford, “Photonics-based in vivo total hemoglobin monitoring and clinical relevance,” J. Biophoton. 2, 277–287 (2009).
[Crossref]

S. Suner, G. Crawford, J. McMurdy, and G. Jay, “Non-invasive determination of hemoglobin by digital photography of palpebral conjunctiva,” J. Emerg. Med. 33, 105–111 (2007).
[Crossref]

Darrodi, M. M.

Dauvin, A.

A. Dauvin, C. Donado, P. Bachtiger, K. C. Huang, C. M. Sauer, D. Ramazzotti, M. Bonvini, L. A. Celi, and M. J. Douglas, “Machine learning can accurately predict pre-admission baseline hemoglobin and creatinine in intensive care patients,” NPJ Digit. Med. 2, 116 (2019).
[Crossref]

de Haan, K.

T. R. Liu, Z. S. Wei, Y. Rivenson, K. de Haan, Y. B. Zhang, Y. C. Wu, and A. Ozcan, “Deep learning-based color holographic microscopy,” J. Biophoton. 12, e201900107 (2019).
[Crossref]

del Junco, D.

L. J. Moore, C. E. Wade, L. Vincent, J. Podbielski, E. Camp, D. del Junco, H. Radhakrishnan, J. McCarthy, B. Gill, and J. B. Holcomb, “Evaluation of noninvasive hemoglobin measurements in trauma patients,” Am. J. Surg. 206, 1041–1047 (2013).
[Crossref]

Derman, R. J.

S. E. Juul, R. J. Derman, and M. Auerbach, “Perinatal iron deficiency: implications for mothers and infants,” Neonatology 115, 269–274 (2019).
[Crossref]

Detsky, A. S.

T. N. Sheth, N. K. Choudhry, M. Bowes, and A. S. Detsky, “The relation of conjunctival pallor to the presence of anemia,” J. Gen. Intern. Med. 12, 102–106 (1997).
[Crossref]

Diggle, P. J.

P. J. Diggle, P. Heagerty, K.-Y. Liang, and S. L. Zeger, Analysis of Longitudinal Data (Oxford University, 2002).

Dimauro, G.

G. Dimauro, A. Guarini, D. Caivano, F. Girardi, C. Pasciolla, and A. Iacobazzi, “Detecting clinical signs of anaemia from digital images of the palpebral conjunctiva,” IEEE Access 7, 113488 (2019).
[Crossref]

Doblas, A.

A. Doblas, E. Sanchez-Ortiga, M. Martinez-Corral, G. Saavedra, and J. Garcia-Sucerquia, “Accurate single-shot quantitative phase imaging of biological specimens with telecentric digital holographic microscopy,” J. Biomed. Opt. 19, 46022 (2014).
[Crossref]

Donado, C.

A. Dauvin, C. Donado, P. Bachtiger, K. C. Huang, C. M. Sauer, D. Ramazzotti, M. Bonvini, L. A. Celi, and M. J. Douglas, “Machine learning can accurately predict pre-admission baseline hemoglobin and creatinine in intensive care patients,” NPJ Digit. Med. 2, 116 (2019).
[Crossref]

Dorman, E.

C. E. Shulman, M. Levene, L. Morison, E. Dorman, N. Peshu, and K. Marsh, “Screening for severe anaemia in pregnancy in Kenya, using pallor examination and self-reported morbidity,” Trans. R. Soc. Trop. Med. Hyg. 95, 250–255 (2001).
[Crossref]

Douglas, M. J.

A. Dauvin, C. Donado, P. Bachtiger, K. C. Huang, C. M. Sauer, D. Ramazzotti, M. Bonvini, L. A. Celi, and M. J. Douglas, “Machine learning can accurately predict pre-admission baseline hemoglobin and creatinine in intensive care patients,” NPJ Digit. Med. 2, 116 (2019).
[Crossref]

Duggan, M.

Duranteau, J.

S. Figueiredo, C. Taconet, A. Harrois, S. Hamada, T. Gauss, M. Raux, and J. Duranteau, “How useful are hemoglobin concentration and its variations to predict significant hemorrhage in the early phase of trauma? A multicentric cohort study,” Ann. Intensive Care 8, 76 (2018).
[Crossref]

Elsing, S. H.

A. R. Kent, S. H. Elsing, and R. L. Hebert, “Conjunctival vasculature in the assessment of anemia,” Ophthalmology 107, 274–277 (2000).
[Crossref]

Escobar, H. B.

J. L. Martinez, H. Saulo, H. B. Escobar, and J. Leao, “A new model selection criterion for partial least squares regression,” Chemom. Intell. Lab. Syst. 169, 64–78 (2017).
[Crossref]

Fantini, S.

I. J. Bigio and S. Fantini, Quantitative Biomedical Optics: Theory, Methods, and Applications (Cambridge University, 2016).

Figueiredo, S.

S. Figueiredo, C. Taconet, A. Harrois, S. Hamada, T. Gauss, M. Raux, and J. Duranteau, “How useful are hemoglobin concentration and its variations to predict significant hemorrhage in the early phase of trauma? A multicentric cohort study,” Ann. Intensive Care 8, 76 (2018).
[Crossref]

Fink, S. E.

E. Abraham, S. E. Fink, D. R. Markle, G. Pinholster, and M. Tsang, “Continuous monitoring of tissue pH with a fiberoptic conjunctival sensor,” Ann. Emerg. Med. 14, 840–844 (1985).
[Crossref]

Finlayson, G.

Fisher, D. E.

J. Y. Lin and D. E. Fisher, “Melanocyte biology and skin pigmentation,” Nature 445, 843–850 (2007).
[Crossref]

Fraiwan, A.

M. N. Hasan, A. Fraiwan, P. Thota, T. Oginni, G. M. Olanipekun, F. Hassan-Hanga, J. Little, S. K. Obaro, and U. A. Gurkan, “Clinical testing of hemechip in Nigeria for point-of-care screening of sickle cell disease,” Blood 132, 1095 (2018).
[Crossref]

Galliani, S.

S. Galliani, C. Lanaras, D. Marmanis, E. Baltsavias, and K. Schindler, “Learned spectral super-resolution,” arXiv:1703.09470 (2017).

Gamble, G. R.

S. C. Yoon, T. S. Shin, K. C. Lawrence, G. W. Heitschmidt, B. Park, and G. R. Gamble, “Hyperspectral imaging using RGB color for foodborne pathogen detection,” J. Electron. Imaging 24, 043008 (2015).
[Crossref]

Garcia-Casal, M. N.

S. R. Pasricha, K. Colman, E. Centeno-Tablante, M. N. Garcia-Casal, and J. P. Pena-Rosas, “Revisiting WHO haemoglobin thresholds to define anaemia in clinical medicine and public health,” Lancet Haematol. 5, E60–E62 (2018).
[Crossref]

Garcia-Sucerquia, J.

A. Doblas, E. Sanchez-Ortiga, M. Martinez-Corral, G. Saavedra, and J. Garcia-Sucerquia, “Accurate single-shot quantitative phase imaging of biological specimens with telecentric digital holographic microscopy,” J. Biomed. Opt. 19, 46022 (2014).
[Crossref]

Gauss, T.

S. Figueiredo, C. Taconet, A. Harrois, S. Hamada, T. Gauss, M. Raux, and J. Duranteau, “How useful are hemoglobin concentration and its variations to predict significant hemorrhage in the early phase of trauma? A multicentric cohort study,” Ann. Intensive Care 8, 76 (2018).
[Crossref]

Gedalin, D.

Gelderblom, H. C.

M. Jaggernath, R. Naicker, S. Madurai, M. A. Brockman, T. Ndung’u, and H. C. Gelderblom, “Diagnostic accuracy of the HemoCue HB 301, STAT-Site MHgb and URIT-12 point-of-care hemoglobin meters in a central laboratory and a community based clinic in Durban, South Africa,” Plos One 11, e0152184 (2016).
[Crossref]

George, A.

S. Collings, O. Thompson, E. Hirst, L. Goossens, A. George, and R. Weinkove, “Non-invasive detection of anaemia using digital photographs of the conjunctiva,” PLOS One 11, e0153286 (2016).
[Crossref]

Giavarina, D.

D. Giavarina, “Understanding Bland Altman analysis,” Biochem. Med. 25, 141–151 (2015).
[Crossref]

Giering, M.

K. Gwn Lore, K. K. Reddy, M. Giering, and E. A. Bernal, “Generative adversarial networks for spectral super-resolution and bidirectional RGB-to-multispectral mapping,” in IEEE Conference on Computer Vision and Pattern Recognition Workshops (2019).

Gill, B.

L. J. Moore, C. E. Wade, L. Vincent, J. Podbielski, E. Camp, D. del Junco, H. Radhakrishnan, J. McCarthy, B. Gill, and J. B. Holcomb, “Evaluation of noninvasive hemoglobin measurements in trauma patients,” Am. J. Surg. 206, 1041–1047 (2013).
[Crossref]

Girardi, F.

G. Dimauro, A. Guarini, D. Caivano, F. Girardi, C. Pasciolla, and A. Iacobazzi, “Detecting clinical signs of anaemia from digital images of the palpebral conjunctiva,” IEEE Access 7, 113488 (2019).
[Crossref]

Goossens, L.

S. Collings, O. Thompson, E. Hirst, L. Goossens, A. George, and R. Weinkove, “Non-invasive detection of anaemia using digital photographs of the conjunctiva,” PLOS One 11, e0153286 (2016).
[Crossref]

Gottlieb, S.

J. Shuren, B. Patel, and S. Gottlieb, “FDA regulation of mobile medical apps,” J. Am. Med. Assoc. 320, 337–338 (2018).
[Crossref]

Green, L.

C. Briggs, S. Kimber, and L. Green, “Where are we at with point-of-care testing in haematology?” Br. J. Haematol. 158, 679–690 (2012).
[Crossref]

Guarini, A.

G. Dimauro, A. Guarini, D. Caivano, F. Girardi, C. Pasciolla, and A. Iacobazzi, “Detecting clinical signs of anaemia from digital images of the palpebral conjunctiva,” IEEE Access 7, 113488 (2019).
[Crossref]

Guo, T.

T. Guo, R. Patnaik, K. Kuhlmann, A. J. Rai, and S. K. Sia, “Smartphone dongle for simultaneous measurement of hemoglobin concentration and detection of HIV antibodies,” Lab Chip 15, 3514–3520 (2015).
[Crossref]

Gurjar, R.

Gurkan, U. A.

M. N. Hasan, A. Fraiwan, P. Thota, T. Oginni, G. M. Olanipekun, F. Hassan-Hanga, J. Little, S. K. Obaro, and U. A. Gurkan, “Clinical testing of hemechip in Nigeria for point-of-care screening of sickle cell disease,” Blood 132, 1095 (2018).
[Crossref]

Gwn Lore, K.

K. Gwn Lore, K. K. Reddy, M. Giering, and E. A. Bernal, “Generative adversarial networks for spectral super-resolution and bidirectional RGB-to-multispectral mapping,” in IEEE Conference on Computer Vision and Pattern Recognition Workshops (2019).

Hamada, S.

S. Figueiredo, C. Taconet, A. Harrois, S. Hamada, T. Gauss, M. Raux, and J. Duranteau, “How useful are hemoglobin concentration and its variations to predict significant hemorrhage in the early phase of trauma? A multicentric cohort study,” Ann. Intensive Care 8, 76 (2018).
[Crossref]

Hanley, C. L.

S. C. Mathews, M. J. McShea, C. L. Hanley, A. Ravitz, A. B. Labrique, and A. B. Cohen, “Digital health: a path to validation,” NPJ Digit. Med. 2, 38 (2019).
[Crossref]

Harris, T.

S. Hutchings, D. N. Naumann, T. Harris, J. Wendon, and M. J. Midwinter, “Observational study of the effects of traumatic injury, haemorrhagic shock and resuscitation on the microcirculation: a protocol for the microshock study,” BMJ Open 6, e010893 (2016).
[Crossref]

Harrois, A.

S. Figueiredo, C. Taconet, A. Harrois, S. Hamada, T. Gauss, M. Raux, and J. Duranteau, “How useful are hemoglobin concentration and its variations to predict significant hemorrhage in the early phase of trauma? A multicentric cohort study,” Ann. Intensive Care 8, 76 (2018).
[Crossref]

Hasan, M. N.

M. N. Hasan, A. Fraiwan, P. Thota, T. Oginni, G. M. Olanipekun, F. Hassan-Hanga, J. Little, S. K. Obaro, and U. A. Gurkan, “Clinical testing of hemechip in Nigeria for point-of-care screening of sickle cell disease,” Blood 132, 1095 (2018).
[Crossref]

Hassan-Hanga, F.

M. N. Hasan, A. Fraiwan, P. Thota, T. Oginni, G. M. Olanipekun, F. Hassan-Hanga, J. Little, S. K. Obaro, and U. A. Gurkan, “Clinical testing of hemechip in Nigeria for point-of-care screening of sickle cell disease,” Blood 132, 1095 (2018).
[Crossref]

Heagerty, P.

P. J. Diggle, P. Heagerty, K.-Y. Liang, and S. L. Zeger, Analysis of Longitudinal Data (Oxford University, 2002).

Hebert, R. L.

A. R. Kent, S. H. Elsing, and R. L. Hebert, “Conjunctival vasculature in the assessment of anemia,” Ophthalmology 107, 274–277 (2000).
[Crossref]

Heitschmidt, G. W.

S. C. Yoon, T. S. Shin, K. C. Lawrence, G. W. Heitschmidt, B. Park, and G. R. Gamble, “Hyperspectral imaging using RGB color for foodborne pathogen detection,” J. Electron. Imaging 24, 043008 (2015).
[Crossref]

Hirst, E.

S. Collings, O. Thompson, E. Hirst, L. Goossens, A. George, and R. Weinkove, “Non-invasive detection of anaemia using digital photographs of the conjunctiva,” PLOS One 11, e0153286 (2016).
[Crossref]

Hiscock, R.

R. Hiscock, D. Kumar, and S. W. Simmons, “Systematic review and meta-analysis of method comparison studies of masimo pulse co-oximeters (Radical-7 or Pronto-7) and HemoCue absorption spectrometers (B-hemoglobin or 201+) with laboratory haemoglobin estimation,” Anaesth. Intensive Care 43, 341–350 (2015).
[Crossref]

Hokororo, A.

L. R. Smart, E. E. Ambrose, K. C. Raphael, A. Hokororo, E. Kamugisha, E. A. Tyburski, W. A. Lam, R. E. Ware, and P. T. McGann, “Simultaneous point-of-care detection of anemia and sickle cell disease in Tanzania: the rapid study,” Ann. Hematol. 97, 239–246 (2018).
[Crossref]

Holcomb, J. B.

L. J. Moore, C. E. Wade, L. Vincent, J. Podbielski, E. Camp, D. del Junco, H. Radhakrishnan, J. McCarthy, B. Gill, and J. B. Holcomb, “Evaluation of noninvasive hemoglobin measurements in trauma patients,” Am. J. Surg. 206, 1041–1047 (2013).
[Crossref]

Hsia, C. C. W.

C. C. W. Hsia, “Mechanisms of disease – respiratory function of hemoglobin,” N. Engl. J. Med. 338, 239–247 (1998).
[Crossref]

Huang, K. C.

A. Dauvin, C. Donado, P. Bachtiger, K. C. Huang, C. M. Sauer, D. Ramazzotti, M. Bonvini, L. A. Celi, and M. J. Douglas, “Machine learning can accurately predict pre-admission baseline hemoglobin and creatinine in intensive care patients,” NPJ Digit. Med. 2, 116 (2019).
[Crossref]

Hutchings, S.

S. Hutchings, D. N. Naumann, T. Harris, J. Wendon, and M. J. Midwinter, “Observational study of the effects of traumatic injury, haemorrhagic shock and resuscitation on the microcirculation: a protocol for the microshock study,” BMJ Open 6, e010893 (2016).
[Crossref]

Iacobazzi, A.

G. Dimauro, A. Guarini, D. Caivano, F. Girardi, C. Pasciolla, and A. Iacobazzi, “Detecting clinical signs of anaemia from digital images of the palpebral conjunctiva,” IEEE Access 7, 113488 (2019).
[Crossref]

Ishizuka, T.

K. Yoshida, I. Nishidate, T. Ishizuka, S. Kawauchi, S. Sato, and M. Satoc, “Multispectral imaging of absorption and scattering properties of in vivo exposed rat brain using a digital red-green-blue camera,” J. Biomed. Opt. 20, 051026 (2015).
[Crossref]

Izatt, J. A.

Jaggernath, M.

M. Jaggernath, R. Naicker, S. Madurai, M. A. Brockman, T. Ndung’u, and H. C. Gelderblom, “Diagnostic accuracy of the HemoCue HB 301, STAT-Site MHgb and URIT-12 point-of-care hemoglobin meters in a central laboratory and a community based clinic in Durban, South Africa,” Plos One 11, e0152184 (2016).
[Crossref]

Jajoo, U.

A. Kalantri, M. Karambelkar, R. Joshi, S. Kalantri, and U. Jajoo, “Accuracy and reliability of pallor for detecting anaemia: a hospital-based diagnostic accuracy study,” PLOS One 5, e8545 (2010).
[Crossref]

Jay, G.

O. Kim, J. McMurdy, G. Jay, C. Lines, G. Crawford, and M. Alber, “Combined reflectance spectroscopy and stochastic modeling approach for noninvasive hemoglobin determination via palpebral conjunctiva,” Physiol. Rep. 2, e00192 (2014).
[Crossref]

J. McMurdy, G. Jay, S. Suner, and G. Crawford, “Photonics-based in vivo total hemoglobin monitoring and clinical relevance,” J. Biophoton. 2, 277–287 (2009).
[Crossref]

S. Suner, G. Crawford, J. McMurdy, and G. Jay, “Non-invasive determination of hemoglobin by digital photography of palpebral conjunctiva,” J. Emerg. Med. 33, 105–111 (2007).
[Crossref]

Joshi, R.

A. Kalantri, M. Karambelkar, R. Joshi, S. Kalantri, and U. Jajoo, “Accuracy and reliability of pallor for detecting anaemia: a hospital-based diagnostic accuracy study,” PLOS One 5, e8545 (2010).
[Crossref]

Juul, S. E.

S. E. Juul, R. J. Derman, and M. Auerbach, “Perinatal iron deficiency: implications for mothers and infants,” Neonatology 115, 269–274 (2019).
[Crossref]

Kalantri, A.

A. Kalantri, M. Karambelkar, R. Joshi, S. Kalantri, and U. Jajoo, “Accuracy and reliability of pallor for detecting anaemia: a hospital-based diagnostic accuracy study,” PLOS One 5, e8545 (2010).
[Crossref]

Kalantri, S.

A. Kalantri, M. Karambelkar, R. Joshi, S. Kalantri, and U. Jajoo, “Accuracy and reliability of pallor for detecting anaemia: a hospital-based diagnostic accuracy study,” PLOS One 5, e8545 (2010).
[Crossref]

Kamugisha, E.

L. R. Smart, E. E. Ambrose, K. C. Raphael, A. Hokororo, E. Kamugisha, E. A. Tyburski, W. A. Lam, R. E. Ware, and P. T. McGann, “Simultaneous point-of-care detection of anemia and sickle cell disease in Tanzania: the rapid study,” Ann. Hematol. 97, 239–246 (2018).
[Crossref]

Karambelkar, M.

A. Kalantri, M. Karambelkar, R. Joshi, S. Kalantri, and U. Jajoo, “Accuracy and reliability of pallor for detecting anaemia: a hospital-based diagnostic accuracy study,” PLOS One 5, e8545 (2010).
[Crossref]

Karamian, P.

M. M. Khansari, M. Tan, P. Karamian, and M. Shahidi, “Inter-visit variability of conjunctival microvascular hemodynamic measurements in healthy and diabetic retinopathy subjects,” Microvasc. Res. 118, 7–11 (2018).
[Crossref]

Karmakar, A.

S. Koundinya, H. Sharma, M. Sharma, A. Upadhyay, R. Manekar, R. Mukhopadhyay, A. Karmakar, and S. Chaudhury, “2D-3D CNN based architectures for spectral reconstruction from RGB images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (2018), pp. 957–964.

Katsaggelos, A. K.

W. Saafin, M. Vega, R. Molina, and A. K. Katsaggelos, “Compressed sensing super resolution of color images,” in 24th European Signal Processing Conference (2016), pp. 1563–1567.

Kawauchi, S.

K. Yoshida, I. Nishidate, T. Ishizuka, S. Kawauchi, S. Sato, and M. Satoc, “Multispectral imaging of absorption and scattering properties of in vivo exposed rat brain using a digital red-green-blue camera,” J. Biomed. Opt. 20, 051026 (2015).
[Crossref]

Kaya, B.

B. Kaya, Y. B. Can, and R. Timofte, “Towards spectral estimation from a single RGB image in the wild,” arXiv:1812.00805 (2018).

Kent, A. R.

A. R. Kent, S. H. Elsing, and R. L. Hebert, “Conjunctival vasculature in the assessment of anemia,” Ophthalmology 107, 274–277 (2000).
[Crossref]

Khansari, M. M.

M. M. Khansari, M. Tan, P. Karamian, and M. Shahidi, “Inter-visit variability of conjunctival microvascular hemodynamic measurements in healthy and diabetic retinopathy subjects,” Microvasc. Res. 118, 7–11 (2018).
[Crossref]

Kim, O.

O. Kim, J. McMurdy, G. Jay, C. Lines, G. Crawford, and M. Alber, “Combined reflectance spectroscopy and stochastic modeling approach for noninvasive hemoglobin determination via palpebral conjunctiva,” Physiol. Rep. 2, e00192 (2014).
[Crossref]

Kim, S. H.

S. H. Kim, M. Lilot, L. S. L. Murphy, K. S. Sidhu, Z. X. Yu, J. Rinehart, and M. Cannesson, “Accuracy of continuous noninvasive hemoglobin monitoring: a systematic review and meta-analysis,” Anesth. Analg. 119, 332–346 (2014).
[Crossref]

Kim, T.

T. Kim, M. A. Visbal-Onufrak, R. L. Konger, and Y. L. Kim, “Data-driven imaging of tissue inflammation using RGB-based hyperspectral reconstruction toward personal monitoring of dermatologic health,” Biomed. Opt. Express 8, 5282–5296 (2017).
[Crossref]

T. Kim, S. H. Choi, N. Lambert-Cheatham, Z. Xu, J. E. Kritchevsky, F. R. Bertin, and Y. L. Kim, “Toward laboratory blood test-comparable photometric assessments for anemia in veterinary hematology,” J. Biomed. Opt. 21, 107001 (2016).
[Crossref]

Kim, Y. L.

Kimber, S.

C. Briggs, S. Kimber, and L. Green, “Where are we at with point-of-care testing in haematology?” Br. J. Haematol. 158, 679–690 (2012).
[Crossref]

King, D.

Konger, R. L.

Koo, T. K.

T. K. Koo and M. Y. Li, “A guideline of selecting and reporting intraclass correlation coefficients for reliability research,” J. Chiropr. Med. 15, 155–163 (2016).
[Crossref]

Koundinya, S.

S. Koundinya, H. Sharma, M. Sharma, A. Upadhyay, R. Manekar, R. Mukhopadhyay, A. Karmakar, and S. Chaudhury, “2D-3D CNN based architectures for spectral reconstruction from RGB images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (2018), pp. 957–964.

Krishnan, A.

A. Krishnan, L. J. Williams, A. R. McIntosh, and H. Abdi, “Partial least squares (PLS) methods for neuroimaging: a tutorial and review,” Neuroimage 56, 455–475 (2011).
[Crossref]

Kritchevsky, J. E.

T. Kim, S. H. Choi, N. Lambert-Cheatham, Z. Xu, J. E. Kritchevsky, F. R. Bertin, and Y. L. Kim, “Toward laboratory blood test-comparable photometric assessments for anemia in veterinary hematology,” J. Biomed. Opt. 21, 107001 (2016).
[Crossref]

Kuhlmann, K.

T. Guo, R. Patnaik, K. Kuhlmann, A. J. Rai, and S. K. Sia, “Smartphone dongle for simultaneous measurement of hemoglobin concentration and detection of HIV antibodies,” Lab Chip 15, 3514–3520 (2015).
[Crossref]

Kumar, D.

R. Hiscock, D. Kumar, and S. W. Simmons, “Systematic review and meta-analysis of method comparison studies of masimo pulse co-oximeters (Radical-7 or Pronto-7) and HemoCue absorption spectrometers (B-hemoglobin or 201+) with laboratory haemoglobin estimation,” Anaesth. Intensive Care 43, 341–350 (2015).
[Crossref]

Kuo, A. N.

Labrique, A. B.

S. C. Mathews, M. J. McShea, C. L. Hanley, A. Ravitz, A. B. Labrique, and A. B. Cohen, “Digital health: a path to validation,” NPJ Digit. Med. 2, 38 (2019).
[Crossref]

Lam, W. A.

R. G. Mannino, D. R. Myers, E. A. Tyburski, C. Caruso, J. Boudreaux, T. Leong, G. D. Clifford, and W. A. Lam, “Smartphone app for non-invasive detection of anemia using only patient-sourced photos,” Nat. Commun. 9, 4924 (2018).
[Crossref]

L. R. Smart, E. E. Ambrose, K. C. Raphael, A. Hokororo, E. Kamugisha, E. A. Tyburski, W. A. Lam, R. E. Ware, and P. T. McGann, “Simultaneous point-of-care detection of anemia and sickle cell disease in Tanzania: the rapid study,” Ann. Hematol. 97, 239–246 (2018).
[Crossref]

Lambert-Cheatham, N.

T. Kim, S. H. Choi, N. Lambert-Cheatham, Z. Xu, J. E. Kritchevsky, F. R. Bertin, and Y. L. Kim, “Toward laboratory blood test-comparable photometric assessments for anemia in veterinary hematology,” J. Biomed. Opt. 21, 107001 (2016).
[Crossref]

Lanaras, C.

S. Galliani, C. Lanaras, D. Marmanis, E. Baltsavias, and K. Schindler, “Learned spectral super-resolution,” arXiv:1703.09470 (2017).

Lawrence, K. C.

S. C. Yoon, T. S. Shin, K. C. Lawrence, G. W. Heitschmidt, B. Park, and G. R. Gamble, “Hyperspectral imaging using RGB color for foodborne pathogen detection,” J. Electron. Imaging 24, 043008 (2015).
[Crossref]

Leao, J.

J. L. Martinez, H. Saulo, H. B. Escobar, and J. Leao, “A new model selection criterion for partial least squares regression,” Chemom. Intell. Lab. Syst. 169, 64–78 (2017).
[Crossref]

Leong, T.

R. G. Mannino, D. R. Myers, E. A. Tyburski, C. Caruso, J. Boudreaux, T. Leong, G. D. Clifford, and W. A. Lam, “Smartphone app for non-invasive detection of anemia using only patient-sourced photos,” Nat. Commun. 9, 4924 (2018).
[Crossref]

Levene, M.

C. E. Shulman, M. Levene, L. Morison, E. Dorman, N. Peshu, and K. Marsh, “Screening for severe anaemia in pregnancy in Kenya, using pallor examination and self-reported morbidity,” Trans. R. Soc. Trop. Med. Hyg. 95, 250–255 (2001).
[Crossref]

Li, J.

Y. Yan, L. Zhang, J. Li, W. Wei, and Y. Zhang, “Accurate spectral super-resolution from single RGB image using multi-scale CNN,” in Pattern Recognition and Computer Vision (Springer, 2018), pp. 206–217.

Li, M. Y.

T. K. Koo and M. Y. Li, “A guideline of selecting and reporting intraclass correlation coefficients for reliability research,” J. Chiropr. Med. 15, 155–163 (2016).
[Crossref]

Li, W.

E. J. Wang, W. Li, J. Y. Zhu, R. Rana, and S. N. Patel, “Noninvasive hemoglobin measurement using unmodified smartphone camera and white flash,” in 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (2017), pp. 2333–2336.

Liang, J. X.

Liang, K.-Y.

P. J. Diggle, P. Heagerty, K.-Y. Liang, and S. L. Zeger, Analysis of Longitudinal Data (Oxford University, 2002).

Lilot, M.

S. H. Kim, M. Lilot, L. S. L. Murphy, K. S. Sidhu, Z. X. Yu, J. Rinehart, and M. Cannesson, “Accuracy of continuous noninvasive hemoglobin monitoring: a systematic review and meta-analysis,” Anesth. Analg. 119, 332–346 (2014).
[Crossref]

Lima, A.

A. Lima and J. Bakker, “Noninvasive monitoring of peripheral perfusion,” Intensive Care Med. 31, 1316–1326 (2005).
[Crossref]

Lin, J. Y.

J. Y. Lin and D. E. Fisher, “Melanocyte biology and skin pigmentation,” Nature 445, 843–850 (2007).
[Crossref]

Lines, C.

O. Kim, J. McMurdy, G. Jay, C. Lines, G. Crawford, and M. Alber, “Combined reflectance spectroscopy and stochastic modeling approach for noninvasive hemoglobin determination via palpebral conjunctiva,” Physiol. Rep. 2, e00192 (2014).
[Crossref]

Little, J.

M. N. Hasan, A. Fraiwan, P. Thota, T. Oginni, G. M. Olanipekun, F. Hassan-Hanga, J. Little, S. K. Obaro, and U. A. Gurkan, “Clinical testing of hemechip in Nigeria for point-of-care screening of sickle cell disease,” Blood 132, 1095 (2018).
[Crossref]

Liu, Q.

S. Chen and Q. Liu, “Modified Wiener estimation of diffuse reflectance spectra from RGB values by the synthesis of new colors for tissue measurements,” J. Biomed. Opt. 17, 030501 (2012).
[Crossref]

Liu, T. R.

T. R. Liu, Z. S. Wei, Y. Rivenson, K. de Haan, Y. B. Zhang, Y. C. Wu, and A. Ozcan, “Deep learning-based color holographic microscopy,” J. Biophoton. 12, e201900107 (2019).
[Crossref]

Livingstone, A. S.

C. M. Thorson, M. L. Ryan, R. M. Van Haren, R. Pereira, J. Olloqui, C. A. Otero, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Change in hematocrit during trauma assessment predicts bleeding even with ongoing fluid resuscitation,” Am. Surg. 79, 398–406 (2013).

M. L. Ryan, C. M. Thorson, C. A. Otero, T. Vu, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Initial hematocrit in trauma: a paradigm shift?” J. Trauma Acute Care Surg. 72, 54–60 (2012).
[Crossref]

Mackiewicz, M.

Madurai, S.

M. Jaggernath, R. Naicker, S. Madurai, M. A. Brockman, T. Ndung’u, and H. C. Gelderblom, “Diagnostic accuracy of the HemoCue HB 301, STAT-Site MHgb and URIT-12 point-of-care hemoglobin meters in a central laboratory and a community based clinic in Durban, South Africa,” Plos One 11, e0152184 (2016).
[Crossref]

Maeda, T.

I. Nishidate, T. Maeda, K. Niizeki, and Y. Aizu, “Estimation of melanin and hemoglobin using spectral reflectance images reconstructed from a digital RGB image by the Wiener estimation method,” Sensors 13, 7902–7915 (2013).
[Crossref]

Manekar, R.

S. Koundinya, H. Sharma, M. Sharma, A. Upadhyay, R. Manekar, R. Mukhopadhyay, A. Karmakar, and S. Chaudhury, “2D-3D CNN based architectures for spectral reconstruction from RGB images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (2018), pp. 957–964.

Mannino, R. G.

R. G. Mannino, D. R. Myers, E. A. Tyburski, C. Caruso, J. Boudreaux, T. Leong, G. D. Clifford, and W. A. Lam, “Smartphone app for non-invasive detection of anemia using only patient-sourced photos,” Nat. Commun. 9, 4924 (2018).
[Crossref]

Markle, D. R.

E. Abraham, S. E. Fink, D. R. Markle, G. Pinholster, and M. Tsang, “Continuous monitoring of tissue pH with a fiberoptic conjunctival sensor,” Ann. Emerg. Med. 14, 840–844 (1985).
[Crossref]

Marmanis, D.

S. Galliani, C. Lanaras, D. Marmanis, E. Baltsavias, and K. Schindler, “Learned spectral super-resolution,” arXiv:1703.09470 (2017).

Marsh, K.

C. E. Shulman, M. Levene, L. Morison, E. Dorman, N. Peshu, and K. Marsh, “Screening for severe anaemia in pregnancy in Kenya, using pallor examination and self-reported morbidity,” Trans. R. Soc. Trop. Med. Hyg. 95, 250–255 (2001).
[Crossref]

Martinez, J. L.

J. L. Martinez, H. Saulo, H. B. Escobar, and J. Leao, “A new model selection criterion for partial least squares regression,” Chemom. Intell. Lab. Syst. 169, 64–78 (2017).
[Crossref]

Martinez-Corral, M.

A. Doblas, E. Sanchez-Ortiga, M. Martinez-Corral, G. Saavedra, and J. Garcia-Sucerquia, “Accurate single-shot quantitative phase imaging of biological specimens with telecentric digital holographic microscopy,” J. Biomed. Opt. 19, 46022 (2014).
[Crossref]

Mathews, S. C.

S. C. Mathews, M. J. McShea, C. L. Hanley, A. Ravitz, A. B. Labrique, and A. B. Cohen, “Digital health: a path to validation,” NPJ Digit. Med. 2, 38 (2019).
[Crossref]

McCarthy, J.

L. J. Moore, C. E. Wade, L. Vincent, J. Podbielski, E. Camp, D. del Junco, H. Radhakrishnan, J. McCarthy, B. Gill, and J. B. Holcomb, “Evaluation of noninvasive hemoglobin measurements in trauma patients,” Am. J. Surg. 206, 1041–1047 (2013).
[Crossref]

McGann, P. T.

L. R. Smart, E. E. Ambrose, K. C. Raphael, A. Hokororo, E. Kamugisha, E. A. Tyburski, W. A. Lam, R. E. Ware, and P. T. McGann, “Simultaneous point-of-care detection of anemia and sickle cell disease in Tanzania: the rapid study,” Ann. Hematol. 97, 239–246 (2018).
[Crossref]

McIntosh, A. R.

A. Krishnan, L. J. Williams, A. R. McIntosh, and H. Abdi, “Partial least squares (PLS) methods for neuroimaging: a tutorial and review,” Neuroimage 56, 455–475 (2011).
[Crossref]

McMurdy, J.

O. Kim, J. McMurdy, G. Jay, C. Lines, G. Crawford, and M. Alber, “Combined reflectance spectroscopy and stochastic modeling approach for noninvasive hemoglobin determination via palpebral conjunctiva,” Physiol. Rep. 2, e00192 (2014).
[Crossref]

J. McMurdy, G. Jay, S. Suner, and G. Crawford, “Photonics-based in vivo total hemoglobin monitoring and clinical relevance,” J. Biophoton. 2, 277–287 (2009).
[Crossref]

S. Suner, G. Crawford, J. McMurdy, and G. Jay, “Non-invasive determination of hemoglobin by digital photography of palpebral conjunctiva,” J. Emerg. Med. 33, 105–111 (2007).
[Crossref]

McNabb, R. P.

McShea, M. J.

S. C. Mathews, M. J. McShea, C. L. Hanley, A. Ravitz, A. B. Labrique, and A. B. Cohen, “Digital health: a path to validation,” NPJ Digit. Med. 2, 38 (2019).
[Crossref]

Midwinter, M. J.

S. Hutchings, D. N. Naumann, T. Harris, J. Wendon, and M. J. Midwinter, “Observational study of the effects of traumatic injury, haemorrhagic shock and resuscitation on the microcirculation: a protocol for the microshock study,” BMJ Open 6, e010893 (2016).
[Crossref]

Molina, R.

W. Saafin, M. Vega, R. Molina, and A. K. Katsaggelos, “Compressed sensing super resolution of color images,” in 24th European Signal Processing Conference (2016), pp. 1563–1567.

Moore, L. J.

L. J. Moore, C. E. Wade, L. Vincent, J. Podbielski, E. Camp, D. del Junco, H. Radhakrishnan, J. McCarthy, B. Gill, and J. B. Holcomb, “Evaluation of noninvasive hemoglobin measurements in trauma patients,” Am. J. Surg. 206, 1041–1047 (2013).
[Crossref]

Morison, L.

C. E. Shulman, M. Levene, L. Morison, E. Dorman, N. Peshu, and K. Marsh, “Screening for severe anaemia in pregnancy in Kenya, using pallor examination and self-reported morbidity,” Trans. R. Soc. Trop. Med. Hyg. 95, 250–255 (2001).
[Crossref]

Moriyama, H.

J. F. Storz and H. Moriyama, “Mechanisms of hemoglobin adaptation to high altitude hypoxia,” High Alt. Med. Biol. 9, 148–157 (2008).
[Crossref]

Mukhopadhyay, R.

S. Koundinya, H. Sharma, M. Sharma, A. Upadhyay, R. Manekar, R. Mukhopadhyay, A. Karmakar, and S. Chaudhury, “2D-3D CNN based architectures for spectral reconstruction from RGB images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (2018), pp. 957–964.

Murphy, L. S. L.

S. H. Kim, M. Lilot, L. S. L. Murphy, K. S. Sidhu, Z. X. Yu, J. Rinehart, and M. Cannesson, “Accuracy of continuous noninvasive hemoglobin monitoring: a systematic review and meta-analysis,” Anesth. Analg. 119, 332–346 (2014).
[Crossref]

Myers, D. R.

R. G. Mannino, D. R. Myers, E. A. Tyburski, C. Caruso, J. Boudreaux, T. Leong, G. D. Clifford, and W. A. Lam, “Smartphone app for non-invasive detection of anemia using only patient-sourced photos,” Nat. Commun. 9, 4924 (2018).
[Crossref]

Naicker, R.

M. Jaggernath, R. Naicker, S. Madurai, M. A. Brockman, T. Ndung’u, and H. C. Gelderblom, “Diagnostic accuracy of the HemoCue HB 301, STAT-Site MHgb and URIT-12 point-of-care hemoglobin meters in a central laboratory and a community based clinic in Durban, South Africa,” Plos One 11, e0152184 (2016).
[Crossref]

Naumann, D. N.

S. Hutchings, D. N. Naumann, T. Harris, J. Wendon, and M. J. Midwinter, “Observational study of the effects of traumatic injury, haemorrhagic shock and resuscitation on the microcirculation: a protocol for the microshock study,” BMJ Open 6, e010893 (2016).
[Crossref]

Ndung’u, T.

M. Jaggernath, R. Naicker, S. Madurai, M. A. Brockman, T. Ndung’u, and H. C. Gelderblom, “Diagnostic accuracy of the HemoCue HB 301, STAT-Site MHgb and URIT-12 point-of-care hemoglobin meters in a central laboratory and a community based clinic in Durban, South Africa,” Plos One 11, e0152184 (2016).
[Crossref]

Newton, J.

A. Thomas, J. Newton, and M. Oldham, “A method to correct for stray light in telecentric optical-CT imaging of radiochromic dosimeters,” Phys. Med. Biol. 56, 4433–4451 (2011).
[Crossref]

Niizeki, K.

I. Nishidate, T. Maeda, K. Niizeki, and Y. Aizu, “Estimation of melanin and hemoglobin using spectral reflectance images reconstructed from a digital RGB image by the Wiener estimation method,” Sensors 13, 7902–7915 (2013).
[Crossref]

Nishidate, I.

K. Yoshida, I. Nishidate, T. Ishizuka, S. Kawauchi, S. Sato, and M. Satoc, “Multispectral imaging of absorption and scattering properties of in vivo exposed rat brain using a digital red-green-blue camera,” J. Biomed. Opt. 20, 051026 (2015).
[Crossref]

I. Nishidate, T. Maeda, K. Niizeki, and Y. Aizu, “Estimation of melanin and hemoglobin using spectral reflectance images reconstructed from a digital RGB image by the Wiener estimation method,” Sensors 13, 7902–7915 (2013).
[Crossref]

Obaro, S. K.

M. N. Hasan, A. Fraiwan, P. Thota, T. Oginni, G. M. Olanipekun, F. Hassan-Hanga, J. Little, S. K. Obaro, and U. A. Gurkan, “Clinical testing of hemechip in Nigeria for point-of-care screening of sickle cell disease,” Blood 132, 1095 (2018).
[Crossref]

Oginni, T.

M. N. Hasan, A. Fraiwan, P. Thota, T. Oginni, G. M. Olanipekun, F. Hassan-Hanga, J. Little, S. K. Obaro, and U. A. Gurkan, “Clinical testing of hemechip in Nigeria for point-of-care screening of sickle cell disease,” Blood 132, 1095 (2018).
[Crossref]

Oiknine, Y.

Olanipekun, G. M.

M. N. Hasan, A. Fraiwan, P. Thota, T. Oginni, G. M. Olanipekun, F. Hassan-Hanga, J. Little, S. K. Obaro, and U. A. Gurkan, “Clinical testing of hemechip in Nigeria for point-of-care screening of sickle cell disease,” Blood 132, 1095 (2018).
[Crossref]

Oldham, M.

A. Thomas, J. Newton, and M. Oldham, “A method to correct for stray light in telecentric optical-CT imaging of radiochromic dosimeters,” Phys. Med. Biol. 56, 4433–4451 (2011).
[Crossref]

Olloqui, J.

C. M. Thorson, M. L. Ryan, R. M. Van Haren, R. Pereira, J. Olloqui, C. A. Otero, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Change in hematocrit during trauma assessment predicts bleeding even with ongoing fluid resuscitation,” Am. Surg. 79, 398–406 (2013).

Otero, C. A.

C. M. Thorson, M. L. Ryan, R. M. Van Haren, R. Pereira, J. Olloqui, C. A. Otero, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Change in hematocrit during trauma assessment predicts bleeding even with ongoing fluid resuscitation,” Am. Surg. 79, 398–406 (2013).

M. L. Ryan, C. M. Thorson, C. A. Otero, T. Vu, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Initial hematocrit in trauma: a paradigm shift?” J. Trauma Acute Care Surg. 72, 54–60 (2012).
[Crossref]

Ozcan, A.

T. R. Liu, Z. S. Wei, Y. Rivenson, K. de Haan, Y. B. Zhang, Y. C. Wu, and A. Ozcan, “Deep learning-based color holographic microscopy,” J. Biophoton. 12, e201900107 (2019).
[Crossref]

Park, B.

S. C. Yoon, T. S. Shin, K. C. Lawrence, G. W. Heitschmidt, B. Park, and G. R. Gamble, “Hyperspectral imaging using RGB color for foodborne pathogen detection,” J. Electron. Imaging 24, 043008 (2015).
[Crossref]

Pasciolla, C.

G. Dimauro, A. Guarini, D. Caivano, F. Girardi, C. Pasciolla, and A. Iacobazzi, “Detecting clinical signs of anaemia from digital images of the palpebral conjunctiva,” IEEE Access 7, 113488 (2019).
[Crossref]

Pasricha, S. R.

S. R. Pasricha, K. Colman, E. Centeno-Tablante, M. N. Garcia-Casal, and J. P. Pena-Rosas, “Revisiting WHO haemoglobin thresholds to define anaemia in clinical medicine and public health,” Lancet Haematol. 5, E60–E62 (2018).
[Crossref]

Patel, B.

J. Shuren, B. Patel, and S. Gottlieb, “FDA regulation of mobile medical apps,” J. Am. Med. Assoc. 320, 337–338 (2018).
[Crossref]

Patel, S. N.

E. J. Wang, W. Li, J. Y. Zhu, R. Rana, and S. N. Patel, “Noninvasive hemoglobin measurement using unmodified smartphone camera and white flash,” in 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (2017), pp. 2333–2336.

Patnaik, R.

T. Guo, R. Patnaik, K. Kuhlmann, A. J. Rai, and S. K. Sia, “Smartphone dongle for simultaneous measurement of hemoglobin concentration and detection of HIV antibodies,” Lab Chip 15, 3514–3520 (2015).
[Crossref]

Pena-Rosas, J. P.

S. R. Pasricha, K. Colman, E. Centeno-Tablante, M. N. Garcia-Casal, and J. P. Pena-Rosas, “Revisiting WHO haemoglobin thresholds to define anaemia in clinical medicine and public health,” Lancet Haematol. 5, E60–E62 (2018).
[Crossref]

Pereira, R.

C. M. Thorson, M. L. Ryan, R. M. Van Haren, R. Pereira, J. Olloqui, C. A. Otero, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Change in hematocrit during trauma assessment predicts bleeding even with ongoing fluid resuscitation,” Am. Surg. 79, 398–406 (2013).

Peshu, N.

C. E. Shulman, M. Levene, L. Morison, E. Dorman, N. Peshu, and K. Marsh, “Screening for severe anaemia in pregnancy in Kenya, using pallor examination and self-reported morbidity,” Trans. R. Soc. Trop. Med. Hyg. 95, 250–255 (2001).
[Crossref]

Pinholster, G.

E. Abraham, S. E. Fink, D. R. Markle, G. Pinholster, and M. Tsang, “Continuous monitoring of tissue pH with a fiberoptic conjunctival sensor,” Ann. Emerg. Med. 14, 840–844 (1985).
[Crossref]

Podbielski, J.

L. J. Moore, C. E. Wade, L. Vincent, J. Podbielski, E. Camp, D. del Junco, H. Radhakrishnan, J. McCarthy, B. Gill, and J. B. Holcomb, “Evaluation of noninvasive hemoglobin measurements in trauma patients,” Am. J. Surg. 206, 1041–1047 (2013).
[Crossref]

Proctor, K. G.

C. M. Thorson, M. L. Ryan, R. M. Van Haren, R. Pereira, J. Olloqui, C. A. Otero, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Change in hematocrit during trauma assessment predicts bleeding even with ongoing fluid resuscitation,” Am. Surg. 79, 398–406 (2013).

M. L. Ryan, C. M. Thorson, C. A. Otero, T. Vu, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Initial hematocrit in trauma: a paradigm shift?” J. Trauma Acute Care Surg. 72, 54–60 (2012).
[Crossref]

Radhakrishnan, H.

L. J. Moore, C. E. Wade, L. Vincent, J. Podbielski, E. Camp, D. del Junco, H. Radhakrishnan, J. McCarthy, B. Gill, and J. B. Holcomb, “Evaluation of noninvasive hemoglobin measurements in trauma patients,” Am. J. Surg. 206, 1041–1047 (2013).
[Crossref]

Rai, A. J.

T. Guo, R. Patnaik, K. Kuhlmann, A. J. Rai, and S. K. Sia, “Smartphone dongle for simultaneous measurement of hemoglobin concentration and detection of HIV antibodies,” Lab Chip 15, 3514–3520 (2015).
[Crossref]

Ramazzotti, D.

A. Dauvin, C. Donado, P. Bachtiger, K. C. Huang, C. M. Sauer, D. Ramazzotti, M. Bonvini, L. A. Celi, and M. J. Douglas, “Machine learning can accurately predict pre-admission baseline hemoglobin and creatinine in intensive care patients,” NPJ Digit. Med. 2, 116 (2019).
[Crossref]

Rana, R.

E. J. Wang, W. Li, J. Y. Zhu, R. Rana, and S. N. Patel, “Noninvasive hemoglobin measurement using unmodified smartphone camera and white flash,” in 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (2017), pp. 2333–2336.

Raphael, K. C.

L. R. Smart, E. E. Ambrose, K. C. Raphael, A. Hokororo, E. Kamugisha, E. A. Tyburski, W. A. Lam, R. E. Ware, and P. T. McGann, “Simultaneous point-of-care detection of anemia and sickle cell disease in Tanzania: the rapid study,” Ann. Hematol. 97, 239–246 (2018).
[Crossref]

Raux, M.

S. Figueiredo, C. Taconet, A. Harrois, S. Hamada, T. Gauss, M. Raux, and J. Duranteau, “How useful are hemoglobin concentration and its variations to predict significant hemorrhage in the early phase of trauma? A multicentric cohort study,” Ann. Intensive Care 8, 76 (2018).
[Crossref]

Ravitz, A.

S. C. Mathews, M. J. McShea, C. L. Hanley, A. Ravitz, A. B. Labrique, and A. B. Cohen, “Digital health: a path to validation,” NPJ Digit. Med. 2, 38 (2019).
[Crossref]

Reddy, K. K.

K. Gwn Lore, K. K. Reddy, M. Giering, and E. A. Bernal, “Generative adversarial networks for spectral super-resolution and bidirectional RGB-to-multispectral mapping,” in IEEE Conference on Computer Vision and Pattern Recognition Workshops (2019).

Riccardi, S.

Rinehart, J.

S. H. Kim, M. Lilot, L. S. L. Murphy, K. S. Sidhu, Z. X. Yu, J. Rinehart, and M. Cannesson, “Accuracy of continuous noninvasive hemoglobin monitoring: a systematic review and meta-analysis,” Anesth. Analg. 119, 332–346 (2014).
[Crossref]

Rivenson, Y.

T. R. Liu, Z. S. Wei, Y. Rivenson, K. de Haan, Y. B. Zhang, Y. C. Wu, and A. Ozcan, “Deep learning-based color holographic microscopy,” J. Biophoton. 12, e201900107 (2019).
[Crossref]

Rodway, G. W.

J. S. Windsor and G. W. Rodway, “Heights and haematology: the story of haemoglobin at altitude,” Postgrad. Med. J. 83, 148–151 (2007).
[Crossref]

Ryan, M. L.

C. M. Thorson, M. L. Ryan, R. M. Van Haren, R. Pereira, J. Olloqui, C. A. Otero, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Change in hematocrit during trauma assessment predicts bleeding even with ongoing fluid resuscitation,” Am. Surg. 79, 398–406 (2013).

M. L. Ryan, C. M. Thorson, C. A. Otero, T. Vu, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Initial hematocrit in trauma: a paradigm shift?” J. Trauma Acute Care Surg. 72, 54–60 (2012).
[Crossref]

Saafin, W.

W. Saafin, M. Vega, R. Molina, and A. K. Katsaggelos, “Compressed sensing super resolution of color images,” in 24th European Signal Processing Conference (2016), pp. 1563–1567.

Saavedra, G.

A. Doblas, E. Sanchez-Ortiga, M. Martinez-Corral, G. Saavedra, and J. Garcia-Sucerquia, “Accurate single-shot quantitative phase imaging of biological specimens with telecentric digital holographic microscopy,” J. Biomed. Opt. 19, 46022 (2014).
[Crossref]

Sanchez-Ortiga, E.

A. Doblas, E. Sanchez-Ortiga, M. Martinez-Corral, G. Saavedra, and J. Garcia-Sucerquia, “Accurate single-shot quantitative phase imaging of biological specimens with telecentric digital holographic microscopy,” J. Biomed. Opt. 19, 46022 (2014).
[Crossref]

Sato, S.

K. Yoshida, I. Nishidate, T. Ishizuka, S. Kawauchi, S. Sato, and M. Satoc, “Multispectral imaging of absorption and scattering properties of in vivo exposed rat brain using a digital red-green-blue camera,” J. Biomed. Opt. 20, 051026 (2015).
[Crossref]

Satoc, M.

K. Yoshida, I. Nishidate, T. Ishizuka, S. Kawauchi, S. Sato, and M. Satoc, “Multispectral imaging of absorption and scattering properties of in vivo exposed rat brain using a digital red-green-blue camera,” J. Biomed. Opt. 20, 051026 (2015).
[Crossref]

Sauer, C. M.

A. Dauvin, C. Donado, P. Bachtiger, K. C. Huang, C. M. Sauer, D. Ramazzotti, M. Bonvini, L. A. Celi, and M. J. Douglas, “Machine learning can accurately predict pre-admission baseline hemoglobin and creatinine in intensive care patients,” NPJ Digit. Med. 2, 116 (2019).
[Crossref]

Saulo, H.

J. L. Martinez, H. Saulo, H. B. Escobar, and J. Leao, “A new model selection criterion for partial least squares regression,” Chemom. Intell. Lab. Syst. 169, 64–78 (2017).
[Crossref]

Savardi, M.

A. Signoroni, M. Savardi, A. Baronio, and S. Benini, “Deep learning meets hyperspectral image analysis: a multidisciplinary review,” J. Imaging 5, 52 (2019).
[Crossref]

Schindler, K.

S. Galliani, C. Lanaras, D. Marmanis, E. Baltsavias, and K. Schindler, “Learned spectral super-resolution,” arXiv:1703.09470 (2017).

Schulman, C. I.

C. M. Thorson, M. L. Ryan, R. M. Van Haren, R. Pereira, J. Olloqui, C. A. Otero, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Change in hematocrit during trauma assessment predicts bleeding even with ongoing fluid resuscitation,” Am. Surg. 79, 398–406 (2013).

M. L. Ryan, C. M. Thorson, C. A. Otero, T. Vu, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Initial hematocrit in trauma: a paradigm shift?” J. Trauma Acute Care Surg. 72, 54–60 (2012).
[Crossref]

Shahidi, M.

M. M. Khansari, M. Tan, P. Karamian, and M. Shahidi, “Inter-visit variability of conjunctival microvascular hemodynamic measurements in healthy and diabetic retinopathy subjects,” Microvasc. Res. 118, 7–11 (2018).
[Crossref]

Sharma, H.

S. Koundinya, H. Sharma, M. Sharma, A. Upadhyay, R. Manekar, R. Mukhopadhyay, A. Karmakar, and S. Chaudhury, “2D-3D CNN based architectures for spectral reconstruction from RGB images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (2018), pp. 957–964.

Sharma, M.

S. Koundinya, H. Sharma, M. Sharma, A. Upadhyay, R. Manekar, R. Mukhopadhyay, A. Karmakar, and S. Chaudhury, “2D-3D CNN based architectures for spectral reconstruction from RGB images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (2018), pp. 957–964.

Sheth, T. N.

T. N. Sheth, N. K. Choudhry, M. Bowes, and A. S. Detsky, “The relation of conjunctival pallor to the presence of anemia,” J. Gen. Intern. Med. 12, 102–106 (1997).
[Crossref]

Shin, T. S.

S. C. Yoon, T. S. Shin, K. C. Lawrence, G. W. Heitschmidt, B. Park, and G. R. Gamble, “Hyperspectral imaging using RGB color for foodborne pathogen detection,” J. Electron. Imaging 24, 043008 (2015).
[Crossref]

Shulman, C. E.

C. E. Shulman, M. Levene, L. Morison, E. Dorman, N. Peshu, and K. Marsh, “Screening for severe anaemia in pregnancy in Kenya, using pallor examination and self-reported morbidity,” Trans. R. Soc. Trop. Med. Hyg. 95, 250–255 (2001).
[Crossref]

Shuren, J.

J. Shuren, B. Patel, and S. Gottlieb, “FDA regulation of mobile medical apps,” J. Am. Med. Assoc. 320, 337–338 (2018).
[Crossref]

Sia, S. K.

T. Guo, R. Patnaik, K. Kuhlmann, A. J. Rai, and S. K. Sia, “Smartphone dongle for simultaneous measurement of hemoglobin concentration and detection of HIV antibodies,” Lab Chip 15, 3514–3520 (2015).
[Crossref]

Sidhu, K. S.

S. H. Kim, M. Lilot, L. S. L. Murphy, K. S. Sidhu, Z. X. Yu, J. Rinehart, and M. Cannesson, “Accuracy of continuous noninvasive hemoglobin monitoring: a systematic review and meta-analysis,” Anesth. Analg. 119, 332–346 (2014).
[Crossref]

Signoroni, A.

A. Signoroni, M. Savardi, A. Baronio, and S. Benini, “Deep learning meets hyperspectral image analysis: a multidisciplinary review,” J. Imaging 5, 52 (2019).
[Crossref]

Simmons, S. W.

R. Hiscock, D. Kumar, and S. W. Simmons, “Systematic review and meta-analysis of method comparison studies of masimo pulse co-oximeters (Radical-7 or Pronto-7) and HemoCue absorption spectrometers (B-hemoglobin or 201+) with laboratory haemoglobin estimation,” Anaesth. Intensive Care 43, 341–350 (2015).
[Crossref]

Smart, L. R.

L. R. Smart, E. E. Ambrose, K. C. Raphael, A. Hokororo, E. Kamugisha, E. A. Tyburski, W. A. Lam, R. E. Ware, and P. T. McGann, “Simultaneous point-of-care detection of anemia and sickle cell disease in Tanzania: the rapid study,” Ann. Hematol. 97, 239–246 (2018).
[Crossref]

Stern, A.

Storz, J. F.

J. F. Storz and H. Moriyama, “Mechanisms of hemoglobin adaptation to high altitude hypoxia,” High Alt. Med. Biol. 9, 148–157 (2008).
[Crossref]

Suner, S.

J. McMurdy, G. Jay, S. Suner, and G. Crawford, “Photonics-based in vivo total hemoglobin monitoring and clinical relevance,” J. Biophoton. 2, 277–287 (2009).
[Crossref]

S. Suner, G. Crawford, J. McMurdy, and G. Jay, “Non-invasive determination of hemoglobin by digital photography of palpebral conjunctiva,” J. Emerg. Med. 33, 105–111 (2007).
[Crossref]

Taconet, C.

S. Figueiredo, C. Taconet, A. Harrois, S. Hamada, T. Gauss, M. Raux, and J. Duranteau, “How useful are hemoglobin concentration and its variations to predict significant hemorrhage in the early phase of trauma? A multicentric cohort study,” Ann. Intensive Care 8, 76 (2018).
[Crossref]

Tan, M.

M. M. Khansari, M. Tan, P. Karamian, and M. Shahidi, “Inter-visit variability of conjunctival microvascular hemodynamic measurements in healthy and diabetic retinopathy subjects,” Microvasc. Res. 118, 7–11 (2018).
[Crossref]

Thomas, A.

A. Thomas, J. Newton, and M. Oldham, “A method to correct for stray light in telecentric optical-CT imaging of radiochromic dosimeters,” Phys. Med. Biol. 56, 4433–4451 (2011).
[Crossref]

Thompson, O.

S. Collings, O. Thompson, E. Hirst, L. Goossens, A. George, and R. Weinkove, “Non-invasive detection of anaemia using digital photographs of the conjunctiva,” PLOS One 11, e0153286 (2016).
[Crossref]

Thorson, C. M.

C. M. Thorson, M. L. Ryan, R. M. Van Haren, R. Pereira, J. Olloqui, C. A. Otero, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Change in hematocrit during trauma assessment predicts bleeding even with ongoing fluid resuscitation,” Am. Surg. 79, 398–406 (2013).

M. L. Ryan, C. M. Thorson, C. A. Otero, T. Vu, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Initial hematocrit in trauma: a paradigm shift?” J. Trauma Acute Care Surg. 72, 54–60 (2012).
[Crossref]

Thota, P.

M. N. Hasan, A. Fraiwan, P. Thota, T. Oginni, G. M. Olanipekun, F. Hassan-Hanga, J. Little, S. K. Obaro, and U. A. Gurkan, “Clinical testing of hemechip in Nigeria for point-of-care screening of sickle cell disease,” Blood 132, 1095 (2018).
[Crossref]

Timofte, R.

B. Arad, O. Ben-Shahar, R. Timofte, L. Van Gool, L. Zhang, and M. Yang, “Ntire 2018 challenge on spectral reconstruction from RGB images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshop (2018), pp. 1042–1051.

B. Kaya, Y. B. Can, and R. Timofte, “Towards spectral estimation from a single RGB image in the wild,” arXiv:1812.00805 (2018).

Tsang, M.

E. Abraham, S. E. Fink, D. R. Markle, G. Pinholster, and M. Tsang, “Continuous monitoring of tissue pH with a fiberoptic conjunctival sensor,” Ann. Emerg. Med. 14, 840–844 (1985).
[Crossref]

Tuchin, V. V.

V. V. Tuchin, Tissue Optics: Light Scattering Methods and Instruments for Medical Diagnosis (SPIE, 2015).

Tyburski, E. A.

R. G. Mannino, D. R. Myers, E. A. Tyburski, C. Caruso, J. Boudreaux, T. Leong, G. D. Clifford, and W. A. Lam, “Smartphone app for non-invasive detection of anemia using only patient-sourced photos,” Nat. Commun. 9, 4924 (2018).
[Crossref]

L. R. Smart, E. E. Ambrose, K. C. Raphael, A. Hokororo, E. Kamugisha, E. A. Tyburski, W. A. Lam, R. E. Ware, and P. T. McGann, “Simultaneous point-of-care detection of anemia and sickle cell disease in Tanzania: the rapid study,” Ann. Hematol. 97, 239–246 (2018).
[Crossref]

Upadhyay, A.

S. Koundinya, H. Sharma, M. Sharma, A. Upadhyay, R. Manekar, R. Mukhopadhyay, A. Karmakar, and S. Chaudhury, “2D-3D CNN based architectures for spectral reconstruction from RGB images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (2018), pp. 957–964.

Van Gool, L.

B. Arad, O. Ben-Shahar, R. Timofte, L. Van Gool, L. Zhang, and M. Yang, “Ntire 2018 challenge on spectral reconstruction from RGB images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshop (2018), pp. 1042–1051.

Van Haren, R. M.

C. M. Thorson, M. L. Ryan, R. M. Van Haren, R. Pereira, J. Olloqui, C. A. Otero, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Change in hematocrit during trauma assessment predicts bleeding even with ongoing fluid resuscitation,” Am. Surg. 79, 398–406 (2013).

Vega, M.

W. Saafin, M. Vega, R. Molina, and A. K. Katsaggelos, “Compressed sensing super resolution of color images,” in 24th European Signal Processing Conference (2016), pp. 1563–1567.

Vincent, L.

L. J. Moore, C. E. Wade, L. Vincent, J. Podbielski, E. Camp, D. del Junco, H. Radhakrishnan, J. McCarthy, B. Gill, and J. B. Holcomb, “Evaluation of noninvasive hemoglobin measurements in trauma patients,” Am. J. Surg. 206, 1041–1047 (2013).
[Crossref]

Visbal-Onufrak, M. A.

Vishwanath, K.

Vo-Dinh, T.

T. Vo-Dinh, Biomedical Photonics Handbook: Fundamentals, Devices, and Techniques (CRC Press, 2019).

Vu, T.

M. L. Ryan, C. M. Thorson, C. A. Otero, T. Vu, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Initial hematocrit in trauma: a paradigm shift?” J. Trauma Acute Care Surg. 72, 54–60 (2012).
[Crossref]

Wade, C. E.

L. J. Moore, C. E. Wade, L. Vincent, J. Podbielski, E. Camp, D. del Junco, H. Radhakrishnan, J. McCarthy, B. Gill, and J. B. Holcomb, “Evaluation of noninvasive hemoglobin measurements in trauma patients,” Am. J. Surg. 206, 1041–1047 (2013).
[Crossref]

Wan, X. X.

Wang, E. J.

E. J. Wang, W. Li, J. Y. Zhu, R. Rana, and S. N. Patel, “Noninvasive hemoglobin measurement using unmodified smartphone camera and white flash,” in 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (2017), pp. 2333–2336.

Ware, R. E.

L. R. Smart, E. E. Ambrose, K. C. Raphael, A. Hokororo, E. Kamugisha, E. A. Tyburski, W. A. Lam, R. E. Ware, and P. T. McGann, “Simultaneous point-of-care detection of anemia and sickle cell disease in Tanzania: the rapid study,” Ann. Hematol. 97, 239–246 (2018).
[Crossref]

Wei, W.

Y. Yan, L. Zhang, J. Li, W. Wei, and Y. Zhang, “Accurate spectral super-resolution from single RGB image using multi-scale CNN,” in Pattern Recognition and Computer Vision (Springer, 2018), pp. 206–217.

Wei, Z. S.

T. R. Liu, Z. S. Wei, Y. Rivenson, K. de Haan, Y. B. Zhang, Y. C. Wu, and A. Ozcan, “Deep learning-based color holographic microscopy,” J. Biophoton. 12, e201900107 (2019).
[Crossref]

Weinkove, R.

S. Collings, O. Thompson, E. Hirst, L. Goossens, A. George, and R. Weinkove, “Non-invasive detection of anaemia using digital photographs of the conjunctiva,” PLOS One 11, e0153286 (2016).
[Crossref]

Wendon, J.

S. Hutchings, D. N. Naumann, T. Harris, J. Wendon, and M. J. Midwinter, “Observational study of the effects of traumatic injury, haemorrhagic shock and resuscitation on the microcirculation: a protocol for the microshock study,” BMJ Open 6, e010893 (2016).
[Crossref]

Williams, L. J.

A. Krishnan, L. J. Williams, A. R. McIntosh, and H. Abdi, “Partial least squares (PLS) methods for neuroimaging: a tutorial and review,” Neuroimage 56, 455–475 (2011).
[Crossref]

Windsor, J. S.

J. S. Windsor and G. W. Rodway, “Heights and haematology: the story of haemoglobin at altitude,” Postgrad. Med. J. 83, 148–151 (2007).
[Crossref]

Wolf, D.

Wu, Y. C.

T. R. Liu, Z. S. Wei, Y. Rivenson, K. de Haan, Y. B. Zhang, Y. C. Wu, and A. Ozcan, “Deep learning-based color holographic microscopy,” J. Biophoton. 12, e201900107 (2019).
[Crossref]

Xu, Z.

T. Kim, S. H. Choi, N. Lambert-Cheatham, Z. Xu, J. E. Kritchevsky, F. R. Bertin, and Y. L. Kim, “Toward laboratory blood test-comparable photometric assessments for anemia in veterinary hematology,” J. Biomed. Opt. 21, 107001 (2016).
[Crossref]

Yan, Y.

Y. Yan, L. Zhang, J. Li, W. Wei, and Y. Zhang, “Accurate spectral super-resolution from single RGB image using multi-scale CNN,” in Pattern Recognition and Computer Vision (Springer, 2018), pp. 206–217.

Yang, M.

B. Arad, O. Ben-Shahar, R. Timofte, L. Van Gool, L. Zhang, and M. Yang, “Ntire 2018 challenge on spectral reconstruction from RGB images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshop (2018), pp. 1042–1051.

Yoon, S. C.

S. C. Yoon, T. S. Shin, K. C. Lawrence, G. W. Heitschmidt, B. Park, and G. R. Gamble, “Hyperspectral imaging using RGB color for foodborne pathogen detection,” J. Electron. Imaging 24, 043008 (2015).
[Crossref]

Yoshida, K.

K. Yoshida, I. Nishidate, T. Ishizuka, S. Kawauchi, S. Sato, and M. Satoc, “Multispectral imaging of absorption and scattering properties of in vivo exposed rat brain using a digital red-green-blue camera,” J. Biomed. Opt. 20, 051026 (2015).
[Crossref]

Yu, Z. X.

S. H. Kim, M. Lilot, L. S. L. Murphy, K. S. Sidhu, Z. X. Yu, J. Rinehart, and M. Cannesson, “Accuracy of continuous noninvasive hemoglobin monitoring: a systematic review and meta-analysis,” Anesth. Analg. 119, 332–346 (2014).
[Crossref]

Zeger, S. L.

P. J. Diggle, P. Heagerty, K.-Y. Liang, and S. L. Zeger, Analysis of Longitudinal Data (Oxford University, 2002).

Zhang, L.

Y. Yan, L. Zhang, J. Li, W. Wei, and Y. Zhang, “Accurate spectral super-resolution from single RGB image using multi-scale CNN,” in Pattern Recognition and Computer Vision (Springer, 2018), pp. 206–217.

B. Arad, O. Ben-Shahar, R. Timofte, L. Van Gool, L. Zhang, and M. Yang, “Ntire 2018 challenge on spectral reconstruction from RGB images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshop (2018), pp. 1042–1051.

Zhang, Y.

Y. Yan, L. Zhang, J. Li, W. Wei, and Y. Zhang, “Accurate spectral super-resolution from single RGB image using multi-scale CNN,” in Pattern Recognition and Computer Vision (Springer, 2018), pp. 206–217.

Zhang, Y. B.

T. R. Liu, Z. S. Wei, Y. Rivenson, K. de Haan, Y. B. Zhang, Y. C. Wu, and A. Ozcan, “Deep learning-based color holographic microscopy,” J. Biophoton. 12, e201900107 (2019).
[Crossref]

Zhu, J. Y.

E. J. Wang, W. Li, J. Y. Zhu, R. Rana, and S. N. Patel, “Noninvasive hemoglobin measurement using unmodified smartphone camera and white flash,” in 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (2017), pp. 2333–2336.

Am. J. Surg. (1)

L. J. Moore, C. E. Wade, L. Vincent, J. Podbielski, E. Camp, D. del Junco, H. Radhakrishnan, J. McCarthy, B. Gill, and J. B. Holcomb, “Evaluation of noninvasive hemoglobin measurements in trauma patients,” Am. J. Surg. 206, 1041–1047 (2013).
[Crossref]

Am. Surg. (1)

C. M. Thorson, M. L. Ryan, R. M. Van Haren, R. Pereira, J. Olloqui, C. A. Otero, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Change in hematocrit during trauma assessment predicts bleeding even with ongoing fluid resuscitation,” Am. Surg. 79, 398–406 (2013).

Anaesth. Intensive Care (1)

R. Hiscock, D. Kumar, and S. W. Simmons, “Systematic review and meta-analysis of method comparison studies of masimo pulse co-oximeters (Radical-7 or Pronto-7) and HemoCue absorption spectrometers (B-hemoglobin or 201+) with laboratory haemoglobin estimation,” Anaesth. Intensive Care 43, 341–350 (2015).
[Crossref]

Anesth. Analg. (1)

S. H. Kim, M. Lilot, L. S. L. Murphy, K. S. Sidhu, Z. X. Yu, J. Rinehart, and M. Cannesson, “Accuracy of continuous noninvasive hemoglobin monitoring: a systematic review and meta-analysis,” Anesth. Analg. 119, 332–346 (2014).
[Crossref]

Ann. Emerg. Med. (1)

E. Abraham, S. E. Fink, D. R. Markle, G. Pinholster, and M. Tsang, “Continuous monitoring of tissue pH with a fiberoptic conjunctival sensor,” Ann. Emerg. Med. 14, 840–844 (1985).
[Crossref]

Ann. Hematol. (1)

L. R. Smart, E. E. Ambrose, K. C. Raphael, A. Hokororo, E. Kamugisha, E. A. Tyburski, W. A. Lam, R. E. Ware, and P. T. McGann, “Simultaneous point-of-care detection of anemia and sickle cell disease in Tanzania: the rapid study,” Ann. Hematol. 97, 239–246 (2018).
[Crossref]

Ann. Intensive Care (1)

S. Figueiredo, C. Taconet, A. Harrois, S. Hamada, T. Gauss, M. Raux, and J. Duranteau, “How useful are hemoglobin concentration and its variations to predict significant hemorrhage in the early phase of trauma? A multicentric cohort study,” Ann. Intensive Care 8, 76 (2018).
[Crossref]

Biochem. Med. (1)

D. Giavarina, “Understanding Bland Altman analysis,” Biochem. Med. 25, 141–151 (2015).
[Crossref]

Biomed. Opt. Express (3)

Blood (1)

M. N. Hasan, A. Fraiwan, P. Thota, T. Oginni, G. M. Olanipekun, F. Hassan-Hanga, J. Little, S. K. Obaro, and U. A. Gurkan, “Clinical testing of hemechip in Nigeria for point-of-care screening of sickle cell disease,” Blood 132, 1095 (2018).
[Crossref]

BMJ Open (1)

S. Hutchings, D. N. Naumann, T. Harris, J. Wendon, and M. J. Midwinter, “Observational study of the effects of traumatic injury, haemorrhagic shock and resuscitation on the microcirculation: a protocol for the microshock study,” BMJ Open 6, e010893 (2016).
[Crossref]

Br. J. Haematol. (1)

C. Briggs, S. Kimber, and L. Green, “Where are we at with point-of-care testing in haematology?” Br. J. Haematol. 158, 679–690 (2012).
[Crossref]

Chemom. Intell. Lab. Syst. (1)

J. L. Martinez, H. Saulo, H. B. Escobar, and J. Leao, “A new model selection criterion for partial least squares regression,” Chemom. Intell. Lab. Syst. 169, 64–78 (2017).
[Crossref]

Fed. Regist. (2)

U.S. Government, “Medicare, Medicaid and CLIA programs; regulations implementing the clinical laboratory improvement amendments of 1988 (CLIA)-HCFA. Final rule with comment period,” Fed. Regist. 57, 7002–7186 (1992).

U.S. Government, “Clinical laboratory improvement amendments of 1988 (CLIA) proficiency testing regulations related to analytes and acceptable performance,” Fed. Regist. 84, 1536–1567 (2019).

High Alt. Med. Biol. (1)

J. F. Storz and H. Moriyama, “Mechanisms of hemoglobin adaptation to high altitude hypoxia,” High Alt. Med. Biol. 9, 148–157 (2008).
[Crossref]

IEEE Access (1)

G. Dimauro, A. Guarini, D. Caivano, F. Girardi, C. Pasciolla, and A. Iacobazzi, “Detecting clinical signs of anaemia from digital images of the palpebral conjunctiva,” IEEE Access 7, 113488 (2019).
[Crossref]

Intensive Care Med. (1)

A. Lima and J. Bakker, “Noninvasive monitoring of peripheral perfusion,” Intensive Care Med. 31, 1316–1326 (2005).
[Crossref]

J. Am. Med. Assoc. (1)

J. Shuren, B. Patel, and S. Gottlieb, “FDA regulation of mobile medical apps,” J. Am. Med. Assoc. 320, 337–338 (2018).
[Crossref]

J. Biomed. Opt. (4)

A. Doblas, E. Sanchez-Ortiga, M. Martinez-Corral, G. Saavedra, and J. Garcia-Sucerquia, “Accurate single-shot quantitative phase imaging of biological specimens with telecentric digital holographic microscopy,” J. Biomed. Opt. 19, 46022 (2014).
[Crossref]

S. Chen and Q. Liu, “Modified Wiener estimation of diffuse reflectance spectra from RGB values by the synthesis of new colors for tissue measurements,” J. Biomed. Opt. 17, 030501 (2012).
[Crossref]

K. Yoshida, I. Nishidate, T. Ishizuka, S. Kawauchi, S. Sato, and M. Satoc, “Multispectral imaging of absorption and scattering properties of in vivo exposed rat brain using a digital red-green-blue camera,” J. Biomed. Opt. 20, 051026 (2015).
[Crossref]

T. Kim, S. H. Choi, N. Lambert-Cheatham, Z. Xu, J. E. Kritchevsky, F. R. Bertin, and Y. L. Kim, “Toward laboratory blood test-comparable photometric assessments for anemia in veterinary hematology,” J. Biomed. Opt. 21, 107001 (2016).
[Crossref]

J. Biophoton. (2)

J. McMurdy, G. Jay, S. Suner, and G. Crawford, “Photonics-based in vivo total hemoglobin monitoring and clinical relevance,” J. Biophoton. 2, 277–287 (2009).
[Crossref]

T. R. Liu, Z. S. Wei, Y. Rivenson, K. de Haan, Y. B. Zhang, Y. C. Wu, and A. Ozcan, “Deep learning-based color holographic microscopy,” J. Biophoton. 12, e201900107 (2019).
[Crossref]

J. Chiropr. Med. (1)

T. K. Koo and M. Y. Li, “A guideline of selecting and reporting intraclass correlation coefficients for reliability research,” J. Chiropr. Med. 15, 155–163 (2016).
[Crossref]

J. Electron. Imaging (1)

S. C. Yoon, T. S. Shin, K. C. Lawrence, G. W. Heitschmidt, B. Park, and G. R. Gamble, “Hyperspectral imaging using RGB color for foodborne pathogen detection,” J. Electron. Imaging 24, 043008 (2015).
[Crossref]

J. Emerg. Med. (1)

S. Suner, G. Crawford, J. McMurdy, and G. Jay, “Non-invasive determination of hemoglobin by digital photography of palpebral conjunctiva,” J. Emerg. Med. 33, 105–111 (2007).
[Crossref]

J. Gen. Intern. Med. (1)

T. N. Sheth, N. K. Choudhry, M. Bowes, and A. S. Detsky, “The relation of conjunctival pallor to the presence of anemia,” J. Gen. Intern. Med. 12, 102–106 (1997).
[Crossref]

J. Imaging (1)

A. Signoroni, M. Savardi, A. Baronio, and S. Benini, “Deep learning meets hyperspectral image analysis: a multidisciplinary review,” J. Imaging 5, 52 (2019).
[Crossref]

J. Opt. Soc. Am. A (1)

J. Trauma Acute Care Surg. (1)

M. L. Ryan, C. M. Thorson, C. A. Otero, T. Vu, C. I. Schulman, A. S. Livingstone, and K. G. Proctor, “Initial hematocrit in trauma: a paradigm shift?” J. Trauma Acute Care Surg. 72, 54–60 (2012).
[Crossref]

Lab Chip (1)

T. Guo, R. Patnaik, K. Kuhlmann, A. J. Rai, and S. K. Sia, “Smartphone dongle for simultaneous measurement of hemoglobin concentration and detection of HIV antibodies,” Lab Chip 15, 3514–3520 (2015).
[Crossref]

Lancet Haematol. (1)

S. R. Pasricha, K. Colman, E. Centeno-Tablante, M. N. Garcia-Casal, and J. P. Pena-Rosas, “Revisiting WHO haemoglobin thresholds to define anaemia in clinical medicine and public health,” Lancet Haematol. 5, E60–E62 (2018).
[Crossref]

Microvasc. Res. (1)

M. M. Khansari, M. Tan, P. Karamian, and M. Shahidi, “Inter-visit variability of conjunctival microvascular hemodynamic measurements in healthy and diabetic retinopathy subjects,” Microvasc. Res. 118, 7–11 (2018).
[Crossref]

N. Engl. J. Med. (2)

C. C. W. Hsia, “Mechanisms of disease – respiratory function of hemoglobin,” N. Engl. J. Med. 338, 239–247 (1998).
[Crossref]

J. W. Cannon, “Hemorrhagic shock,” N. Engl. J. Med. 378, 370–379 (2018).
[Crossref]

Nat. Commun. (1)

R. G. Mannino, D. R. Myers, E. A. Tyburski, C. Caruso, J. Boudreaux, T. Leong, G. D. Clifford, and W. A. Lam, “Smartphone app for non-invasive detection of anemia using only patient-sourced photos,” Nat. Commun. 9, 4924 (2018).
[Crossref]

Nature (1)

J. Y. Lin and D. E. Fisher, “Melanocyte biology and skin pigmentation,” Nature 445, 843–850 (2007).
[Crossref]

Neonatology (1)

S. E. Juul, R. J. Derman, and M. Auerbach, “Perinatal iron deficiency: implications for mothers and infants,” Neonatology 115, 269–274 (2019).
[Crossref]

Neuroimage (1)

A. Krishnan, L. J. Williams, A. R. McIntosh, and H. Abdi, “Partial least squares (PLS) methods for neuroimaging: a tutorial and review,” Neuroimage 56, 455–475 (2011).
[Crossref]

NPJ Digit. Med. (2)

A. Dauvin, C. Donado, P. Bachtiger, K. C. Huang, C. M. Sauer, D. Ramazzotti, M. Bonvini, L. A. Celi, and M. J. Douglas, “Machine learning can accurately predict pre-admission baseline hemoglobin and creatinine in intensive care patients,” NPJ Digit. Med. 2, 116 (2019).
[Crossref]

S. C. Mathews, M. J. McShea, C. L. Hanley, A. Ravitz, A. B. Labrique, and A. B. Cohen, “Digital health: a path to validation,” NPJ Digit. Med. 2, 38 (2019).
[Crossref]

Ophthalmology (1)

A. R. Kent, S. H. Elsing, and R. L. Hebert, “Conjunctival vasculature in the assessment of anemia,” Ophthalmology 107, 274–277 (2000).
[Crossref]

Opt. Express (2)

Opt. Lett. (1)

Phys. Med. Biol. (1)

A. Thomas, J. Newton, and M. Oldham, “A method to correct for stray light in telecentric optical-CT imaging of radiochromic dosimeters,” Phys. Med. Biol. 56, 4433–4451 (2011).
[Crossref]

Physiol. Rep. (1)

O. Kim, J. McMurdy, G. Jay, C. Lines, G. Crawford, and M. Alber, “Combined reflectance spectroscopy and stochastic modeling approach for noninvasive hemoglobin determination via palpebral conjunctiva,” Physiol. Rep. 2, e00192 (2014).
[Crossref]

PLOS One (2)

S. Collings, O. Thompson, E. Hirst, L. Goossens, A. George, and R. Weinkove, “Non-invasive detection of anaemia using digital photographs of the conjunctiva,” PLOS One 11, e0153286 (2016).
[Crossref]

A. Kalantri, M. Karambelkar, R. Joshi, S. Kalantri, and U. Jajoo, “Accuracy and reliability of pallor for detecting anaemia: a hospital-based diagnostic accuracy study,” PLOS One 5, e8545 (2010).
[Crossref]

M. Jaggernath, R. Naicker, S. Madurai, M. A. Brockman, T. Ndung’u, and H. C. Gelderblom, “Diagnostic accuracy of the HemoCue HB 301, STAT-Site MHgb and URIT-12 point-of-care hemoglobin meters in a central laboratory and a community based clinic in Durban, South Africa,” Plos One 11, e0152184 (2016).
[Crossref]

Postgrad. Med. J. (1)

J. S. Windsor and G. W. Rodway, “Heights and haematology: the story of haemoglobin at altitude,” Postgrad. Med. J. 83, 148–151 (2007).
[Crossref]

Sensors (1)

I. Nishidate, T. Maeda, K. Niizeki, and Y. Aizu, “Estimation of melanin and hemoglobin using spectral reflectance images reconstructed from a digital RGB image by the Wiener estimation method,” Sensors 13, 7902–7915 (2013).
[Crossref]

Trans. R. Soc. Trop. Med. Hyg. (1)

C. E. Shulman, M. Levene, L. Morison, E. Dorman, N. Peshu, and K. Marsh, “Screening for severe anaemia in pregnancy in Kenya, using pallor examination and self-reported morbidity,” Trans. R. Soc. Trop. Med. Hyg. 95, 250–255 (2001).
[Crossref]

Other (12)

E. J. Wang, W. Li, J. Y. Zhu, R. Rana, and S. N. Patel, “Noninvasive hemoglobin measurement using unmodified smartphone camera and white flash,” in 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (2017), pp. 2333–2336.

W. Saafin, M. Vega, R. Molina, and A. K. Katsaggelos, “Compressed sensing super resolution of color images,” in 24th European Signal Processing Conference (2016), pp. 1563–1567.

V. V. Tuchin, Tissue Optics: Light Scattering Methods and Instruments for Medical Diagnosis (SPIE, 2015).

I. J. Bigio and S. Fantini, Quantitative Biomedical Optics: Theory, Methods, and Applications (Cambridge University, 2016).

T. Vo-Dinh, Biomedical Photonics Handbook: Fundamentals, Devices, and Techniques (CRC Press, 2019).

K. Gwn Lore, K. K. Reddy, M. Giering, and E. A. Bernal, “Generative adversarial networks for spectral super-resolution and bidirectional RGB-to-multispectral mapping,” in IEEE Conference on Computer Vision and Pattern Recognition Workshops (2019).

S. Galliani, C. Lanaras, D. Marmanis, E. Baltsavias, and K. Schindler, “Learned spectral super-resolution,” arXiv:1703.09470 (2017).

S. Koundinya, H. Sharma, M. Sharma, A. Upadhyay, R. Manekar, R. Mukhopadhyay, A. Karmakar, and S. Chaudhury, “2D-3D CNN based architectures for spectral reconstruction from RGB images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (2018), pp. 957–964.

B. Kaya, Y. B. Can, and R. Timofte, “Towards spectral estimation from a single RGB image in the wild,” arXiv:1812.00805 (2018).

B. Arad, O. Ben-Shahar, R. Timofte, L. Van Gool, L. Zhang, and M. Yang, “Ntire 2018 challenge on spectral reconstruction from RGB images,” in IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshop (2018), pp. 1042–1051.

Y. Yan, L. Zhang, J. Li, W. Wei, and Y. Zhang, “Accurate spectral super-resolution from single RGB image using multi-scale CNN,” in Pattern Recognition and Computer Vision (Springer, 2018), pp. 206–217.

P. J. Diggle, P. Heagerty, K.-Y. Liang, and S. L. Zeger, Analysis of Longitudinal Data (Oxford University, 2002).

Supplementary Material (1)

NameDescription
» Supplement 1       Supplementary Figures, Supplementary Tables, Supplementary Methods

Cited By

OSA participates in Crossref's Cited-By Linking service. Citing articles from OSA journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (5)

Fig. 1.
Fig. 1. Spectral super-resolution (SSR) spectroscopy for mobile health (mHealth) hemoglobin (Hgb) analyses. (a) The inner eyelid (i.e., palpebral conjunctiva) is used as an accessible sensing site for noninvasive blood Hgb quantification. An RGB image of the inner eyelid is conveniently captured using the built-in camera of a smartphone. The subject simply pulls down the eyelid to expose the conjunctiva, and a healthcare professional takes a photograph of the eyelid. The mobile application collects red (R), green (G), and blue (B) color information from the eyelid image and applies SSR to mathematically reconstruct spectra in the visible wavelength range. The spectral intensity reflected from the inner eyelid is sensitive to changes in Hgb content in the blood. The reconstructed spectrum of the acquired eyelid image is then processed to accurately and precisely predict the amount of total blood Hgb content. The result displays the blood Hgb count in units of ${\rm g}\;{{\rm dL}^{- 1}}$ in the same manner of clinical laboratory Hgb tests. (b) Statistical learning in the mHematology for SSR blood Hgb computation developed using separate training and validation datasets. The first step is to apply SSR to the eyelid portion of the RGB image. The second step is to compute blood Hgb content in ${\rm g}\;{{\rm dL}^{- 1}}$ using the spectroscopic model of blood Hgb, which is also validated by the clinical laboratory blood Hgb tests (i.e., the gold standard).

View in Article | Download Full Size | PPT Slide | PDF

Fig. 2.
Fig. 2. High-quality spectra acquired by the image-guided hyperspectral line-scanning system and the mHematology mobile application. (a) Photograph of the image-guided hyperspectral line-scanning system for imaging the exact portion of the inner eyelid. The participant sits in front of the system, facing the telecentric lens, places the chin on the chinrest, and pulls down the eyelid for imaging when instructed. (b) Location where the hyperspectral line-scanning is performed (translucent white rectangle). The hyperspectral line-scan dataset contains spatial ($y$) and wavelength ($\lambda$) information. The averaged spectrum corresponds to the average intensity along the spatial $y$ axis for each $\lambda$ value. The characteristic absorption spectrum of blood Hgb is clearly visible. (c) mHematology mobile application developed for data acquisition in a low-end Android smartphone (Samsung Galaxy J3). On the main application screen, it displays a circle and arc to serve as guidance for locating the eyeball and the inner eyelid at consistent distance and position within the image. To remove the background room light, the application automatically acquires two RGB photographs by controlling the built-in flashlight (i.e., white-light LED) to turn on and off. To compensate for the system response, two RGB images of a reflectance standard are taken (left). Similarly, the application automatically takes two RGB images with flash on and flash off for the individual’s exposed eyelid (right). (d) Spectral profiles of the white-light LED illumination sources in the image-guided hyperspectral line-scanning system and Samsung Galaxy J3. The data acquisition procedure incorporates reference measurements of white reflectance standards (99% reflectivity in the visible range) to compensate for the spectral responses of the light source and the camera in the system.

View in Article | Download Full Size | PPT Slide | PDF

Fig. 3.
Fig. 3. Performance of spectroscopic blood Hgb measurements of the left and right inner eyelids. (a) High correlations between the computed and clinical laboratory blood Hgb levels in both of the training ($n = {138}$ plotted in blue) and validation ($n = {15}$ plotted in red) datasets. (b) Bland–Altman analysis of comparing the computed blood Hgb levels with the clinical laboratory results, showing narrow 95% limits of agreement (LOA) of [${-}{1.56},\;{1.58}\;{\rm g}\;{{\rm dL}^{- 1}}$] with bias of ${0.01}\;{\rm g}\;{{\rm dL}^{- 1}}$ in the validation dataset. (c) ${R^2}$ values between the computed blood Hgb levels and the clinical laboratory results of validation datasets from 10 different combinations of training and validation datasets. (d) LOA (error bar) and bias (square point) values of validation datasets from the different training–validation combinations. (e) and (f) Systematic comparisons of the spectra measured from the left (green) and right (magenta) inner eyelids among a subset of 36 participants. (e) The average spectral differences between the left and right eyelid spectra (Fig. S8) are statistically insignificant, and the Pearson correlation coefficients are close to 1 in all of the participants. (f) High correlations of the left and right spectroscopic blood Hgb measurements, compared with the clinical laboratory blood Hgb test results.

View in Article | Download Full Size | PPT Slide | PDF

Fig. 4.
Fig. 4. Comparisons between the original spectra and the SSR-reconstructed spectra. (a) and (b) Inner eyelids’ spectral intensity (acquired by the image-guided hyperspectral line-scanning system and the mHematology application) plotted with the clinical laboratory blood Hgb values (vertical axis) from the training dataset ($n = {138}$). (c) Differences between the original and SSR-reconstructed spectra plotted with 95% confidence intervals at each wavelength. The transformation matrix that converts RGB data to spectral data is optimized by minimizing the differences in the training dataset. (d) Average spectral differences and Pearson correlation coefficients between the original and reconstructed spectra in the training dataset as a function of blood Hgb levels corresponding to each spectrum. (e) and (f) Inner eyelids’ spectral intensity (acquired by the image-guided hyperspectral line-scanning system and the mHematology application) visualized with the clinical laboratory blood Hgb values (vertical axis) from the validation dataset ($n = {15}$). (g) Differences between the original and SSR-reconstructed spectra plotted with 95% confidence intervals at each wavelength are still small, supporting the high fidelity of SSR. The differences in the wavelength range between 450 and 575 nm are generally higher, because distinct Hgb absorption is present in this range [Fig. S5(d)]. (h) Average spectral differences and Pearson correlation coefficients between the original and reconstructed spectra in the validation dataset as a function of blood Hgb levels.

View in Article | Download Full Size | PPT Slide | PDF

Fig. 5.
Fig. 5. Performance of SSR blood Hgb measurements with the mHematology mobile application. (a) High correlations between the SSR-computed and clinical laboratory blood Hgb levels in both training ($n = {138}$ plotted in blue) and validation ($n = {15}$ plotted in red) datasets. (b) Bland–Altman analyses of comparing the computed blood Hgb measurements with the clinical laboratory results, showing narrow 95% limits of agreement (LOA) of [${-}{2.20},\;{2.29}\;{\rm g}\;{{\rm dL}^{- 1}}$] and bias of ${0.04}\;{\rm g}\;{{\rm dL}^{- 1}}$ in the validation dataset. In particular, the bias is not associated with actual blood Hgb levels in the validation dataset (Table S3 in Supplement 1). (c) ${R^2}$ values between the computed blood Hgb levels and the clinical laboratory results of validation datasets from 10 different combinations of training and validation datasets. (d) LOA (error bar) and bias (square point) values of validation datasets from the different training–validation combinations. The mHematology application reliably predicts the actual blood Hgb levels without any hardware attachments to the smartphone.

View in Article | Download Full Size | PPT Slide | PDF

Tables (4)

Tables Icon

Table 1. Participant Characteristics

View Table | View all tables in this article
Tables Icon

Table 2. Spectral Differences between the Left and Right Inner Eyelids

View Table | View all tables in this article
Tables Icon

Table 3. Intra-Class Correlations (ICC) between the Left and Right Inner Eyelidsa

View Table | View all tables in this article
Tables Icon

Table 4. ICC between Spectroscopic and SSR Blood Hgb Measurementsa

View Table | View all tables in this article

Equations (11)

Equations on this page are rendered with MathJax. Learn more.

x 3 × 1 = S 3 × N y N × 1 + e 3 × 1 ,
X 3 × m = S 3 × N Y N × m .
[ R 1 R 2 ⋯ R m G 1 G 2 ⋯ G m B 1 B 2 ⋯ B m ] = [ S 1 ( λ 1 ) S 1 ( λ 2 ) ⋯ S 1 ( λ N ) S 2 ( λ 1 ) S 2 ( λ 2 ) ⋯ S 2 ( λ N ) S 3 ( λ 1 ) S 3 ( λ 2 ) ⋯ S 3 ( λ N ) ] × [ I 1 ( λ 1 ) I 2 ( λ 1 ) ⋯ I m ( λ 1 ) I 1 ( λ 2 ) I 2 ( λ 2 ) ⋯ I m ( λ 2 ) ⋮ ⋮ ⋮ ⋮ I 1 ( λ N ) I 2 ( λ N ) ⋯ I m ( λ N ) ] ,
Y N × m = T N × 3 X 3 × m ,
Y N × m = T ^ N × p X ^ p × m ,
X ^ p × m = [ R 1 G 1 B 1 ⋯ R 1 i G 1 j B 1 k ⋯ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ R m G m B m ⋯ R m i G m j B m k ⋯ ] T ,
I m ( λ o r R G B ) = L ( λ o r R G B ) ⋅ C ( λ o r R G B ) ⋅ D ( λ o r R G B ) ⋅ y ( λ o r R G B ) ,
y ( λ o r R G B ) = I m ( λ o r R G B ) I r e f e r e n c e ( λ o r R G B ) .
y ( λ o r R G B ) = I m ( λ o r R G B ) − I b a c k g r o u n d ( λ o r R G B ) I r e f e r e n c e ( λ o r R G B ) − I b a c k g r o u n d ( λ o r R G B ) .
B i a s = d ¯ = 1 n ∑ k = 1 n d k .
L O A = d ¯ ± 1.96 1 n − 1 ∑ k = 1 n ( d k − d ¯ ) 2 .

Metrics