Open Access
Add to BibSonomy
Abstract
To facilitate the clinical applicability of the diffuse optical inspection device, a compact multi-wavelength diffuse optical tomography system for breast imaging (compact-DOTB) with a fiber-free parallel-plane structure was designed and fabricated for acquiring three-dimensional optical properties of the breast in continuous-wave mode. The source array consists of 56 surface-mounted micro light-emitting diodes (LEDs), each integrating three wavelengths (660, 750, and 840 nm). The detector array is arranged with 56 miniaturized surface-mounted optical sensors, each encapsulating a high-sensitivity photodiode (PD) and a low-noise current amplifier with a gain of 24×. The system provides 3,136 pairs of source-detector measurements at each wavelength, and the fiber-free design largely ensures consistency between source/detection channels while effectively reducing the complexity of system operation and maintenance. We have evaluated the compact-DOTB system’s characteristics and demonstrated its performance in terms of reconstruction positioning accuracy and recovery contrast with breast-sized phantom experiments. Furthermore, the breast cancer patient studies have been carried out, and the quantitative results indicate that the compact-DOTB system is able to observe the changes in the functional tissue components of the breast after receiving the neoadjuvant chemotherapy (NAC), demonstrating the great potential of the proposed compact system for clinical applications, while its cost and ease of operation are competitive with the existing breast-DOT devices.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Breast cancer remains a worldwide public health dilemma and is currently the most common tumor in the globe [1]. To date, multiple imaging modalities including ultrasound, x-ray mammography, digital breast tomosynthesis (DBT), magnetic resonance imaging (MRI) and positron emission tomography (PET) play an invaluable role in the diagnosis of breast cancer but still have limitations [1,2]. Since the biological characteristics of the tumor are more relevant to breast-cancer prognosis than the size of the tumor, the provision of richer functional information about the breast tissue is particularly important for the accurate diagnosis of breast cancer [3]. Diffuse optical tomography (DOT), a near-infrared (NIR) (600-1100 nm) optical imaging technique providing two-dimensional (2D) or three-dimensional (3D) maps of tissue optical absorption and scattering properties related to the physiological state of the tissue at depths up to several centimeters, is particularly attractive for breast cancer management, i.e., the detection and differentiation of breast cancer, and the monitoring of treatments [2]. Using multi-wavelength DOT measurements and unmixing reconstructions, the concentrations of oxyhemoglobin ([HbO2]), deoxyhemoglobin ([Hb]), water, and fat in breast tissue can be obtained, and then, the total hemoglobin concentration ([HbT]) and tissue oxygen saturation (StO2) are calculated. These functional information has been shown to be effective in distinguishing malignant from healthy tissue in the breast [4]. For the treatment of breast cancer, neoadjuvant chemotherapy (NAC), or preoperative systemic therapy, has become a well-established method to shrink the breast tumor before surgical intervention in recent years [5]. Effective assessment of the response to NAC at an earlier time point during chemotherapy is critical for physicians to dynamically optimize the treatment regimen and thereby improve patient outcomes. With the advantages of reasonable cost, the possibility of frequent monitoring, and being well tolerated by patients, DOT has shown promising applications in monitoring the response to NAC [4–9].
DOT systems have source-detector (SD) schemes that incorporate either frequency-domain (FD), continuous-wave (CW), or time-domain (TD) data acquisition. In addition to the data type of the measurement, these systems can differ by SD configuration or imaging geometry. These geometries include, for example, single-layer ring [10–12], multi-layer rings [13–15], semi ring [16,17], football-shaped or pentagonal interface [18–20], cup/cylindrical imaging chamber [21–23], parallel plane [24–36], handheld structures [37–40] and wearable structures [41,42]. Among them, the single-layer ring and football-shaped/pentagonal SD configurations focus on recovering the 2D image of the specific layer of breast tissue. The multi-layer ring structure needs to use the lifting stage to carry the SD ring moving up and down to achieve the 3D DOT imaging of the breast. For the cup-shaped or cylindrical imaging chamber structure, the optical matching liquid needs to be added between the breast tissue and the imaging chamber prior to the measurement. Flexman et al. have proposed a CW digital DOT system for dynamic measurements of the breast. The system with a multi-layer ring SD configuration uses 4 wavelengths of near-infrared light (765, 805, 827, and 905 nm) to illuminate the tissue and can acquire data from 32 sources and 64 detectors per breast with 4 wavelengths [17]. El-Ghussein et al. have developed an MRI-guided wideband (660–948 nm), hybrid FD and CW optical spectral tomography system. The optical detection part of the system is in a circular/pentagonal SD configuration [18]. Zhao et al. have proposed a portable 9-wavelength FD and CW system with a football-shaped SD configuration. Further, the system is modified to combine 12-wavelengths of FD (660–852 nm) and CW (850–1064 nm) for capturing tissue features and additional lipid and collagen features. They have compared the MRI T2 images with the DOT results, and the tumor pointed in the MRI image is quantified with higher hemoglobin and lower lipid concentration in DOT image [19,20]. For the handheld structure, multiple regions on the breast surface need to be manually selected for the measurement, resulting in a longer overall measurement time, however it has high flexibility in use. Vavadi et al. have proposed a handheld compact ultrasound-guided DOT system consisting of 9 sources of 4 optical wavelengths in the range of 730 to 830 nm and 14 parallel photomultiplier tube (PMT) detectors [40]. Compared with other imaging geometries, the parallel-plane configuration is similar to the mammography device, with the advantage of being able to obtaining high-quality imaging at the planes parallel to the SD plates, and it is also easily combined with other imaging modalities to achieve in-situ measurements [29,33–35]. Zimmermann et al. have developed a 3D co-registered DBT and dynamic DOT system with the parallel-plane configuration. The system integrates 96 CW and 24 FD source locations as well as 32 CW and 20 FD detection locations [33]. Moreover, the “squeezing-relaxing” measurement mode from the parallel-plane structure can also be adopted to estimate the hemodynamic parameters related to the diagnosis of breast tumors [36].
Currently, most of the above-mentioned DOT breast imaging devices adopt the structure of optical fiber-coupled light sources and detectors, and employ the avalanche photodiode (APD) or PMT module with a larger size as the detector, making the overall structure of the device complex. For the use of multiple fibers for light guidance, firstly, to ensure the effective contact between the source/detection fibers and the breast tissue, it is usually necessary for the operator to adjust most of the fibers prior to scanning, which increases the preparation time for DOT measurement. In addition, with a high-density SD configuration, the total number of the source/detection fibers exceeds one hundred. These fibers may be entangled in the imaging equipment, and the bending of hundreds of fibers may vary due to the mechanical movement of the scanning device, which may affect the consistency between the SD channels. Moreover, the damage to the fibers will also increase the equipment maintenance cost. For the detection module, the conventionally used APD detector requires high bias voltage, and the PMT in voltage output mode also needs to be equipped with a high-voltage power supply and an external trans-impedance amplifier (TIA), which takes up more space for the entire detection module, and the operator needs to carefully adjust the gain of each channel to avoid inconsistency. Based on the above analysis, the DOT systems for breast imaging should be further optimized to adapt to clinical applications where space and time are at a premium, while maintaining good performance of the equipment. For the physicians, the compactness of the device should be improved and the complexity should be reduced to make it easy to operate. For the patients, equipment preparation and adjustment time before each scanning should be shortened to relieve patient stress, especially during NAC, where multiple monitoring may be required over a long period of time.
In recent years, with the development of semiconductor and micro-electro-mechanical system (MEMS) technology, some miniaturized and highly integrated devices, such as light-emitting diode (LED) capable of encapsulating multiple excitation wavelengths, and optical sensors such as high-performance photodiode (PD) [42,43,44], silicon photomultiplier (SiPM) [45,46] and single photon avalanche diode (SPAD) [47], are widely used in biomedical optical application research. On the premise of safety, these devices can directly contact human tissues as surface-mounted structures and provide the potential for significantly reducing the complexity and cost of compact DOT systems. Therefore, in order to promote the clinical applicability of the diffuse optical inspection device, a compact multi-wavelength DOT system for breast imaging (compact-DOTB) with a fiber-free parallel-plane structure was designed and fabricated for acquiring 3D optical properties of the breast in CW mode. The system has 56 surface-mounted LEDs, each integrating three wavelengths (660, 750, and 840 nm), and 56 surface-mounted optical sensors. The two depth cameras mounted in the imaging bed are employed to obtain the contour of the breast and to generate the finite element mesh. We evaluated the compact-DOTB system’s characteristics and demonstrated the performance in terms of reconstruction positioning accuracy and recovery contrast with breast-sized phantom experiments. Furthermore, we present measurements on one breast cancer patient, which is one example from an ongoing clinical study involving many patients. The system’s capability to observe changes after receiving NAC demonstrates the great potential of the proposed developed compact system for clinical applications.
2. Description of the compact-DOTB system
2.1 Setup
The appearance photograph of the compact-DOTB system is shown in Fig. 1(a), and its schematic is shown in Fig. 1(b). Briefly, the system consists of a light source module, a detection module, a breast contour measurement module, a mechanical movement module and a control module.
Fig. 1. Compact-DOTB system. (a) Photograph of the imaging bed (dimensions of the imaging bed (L*W*H): 2056mm×916 mm×610 mm). (b) Schematic of the system. The dimensions of the source/detector plates are shown in Fig. 2.
The source module consists of the LED array plate with corresponding driving and switching circuits. The 56 three-wavelength LEDs are arranged in a 7×8 grid, with the LEDs spaced 14 mm (columns) in the y-direction and 13 mm (rows) in the z-direction, respectively. The first row of the LED array is 4 mm away from the top edge of the plate, and the first and last columns of LEDs are 31 mm away from the left and right edges of the plate respectively, as shown in Fig. 2(b). Each LED is a surface-mounted structure with three compactly arranged light-emitting points at 660 nm, 750 nm, and 840 nm as the central wavelengths and share one window lens (Fig. 1(b)). The excitation of each wavelength can be controlled independently, and the switching time of 56 LEDs for a single wavelength is less than 15 seconds. The gaps between the LEDs on the source plate are covered with a layer of silicone to ensure the comfort and safety of use while keeping the source plate closely fit with the breast tissue and avoiding the effects of ambient light on measurements and inter-channel crosstalk caused by light leakage, as illustrated in Fig. 2(b). The optical emitters and detectors protrude from the silicone surface by ∼0.5 mm and indent the soft tissue of the breast. The compact-DOTB system has a total of 56×3 = 168 light sources to be controlled, and four 3-8 decoders cascaded with multi-channel analog switches are employed to control the connection status between the LEDs and the driving circuit for time-sharing switching of the light sources. The LED driving circuit is based on the LT3092 constant current source chip to realize a stable current output. The programmable precision resistor in the driving circuit can be adjusted via the I2C protocol to enable rapid changes of the output current. In different experiments, the adjustable range of the driving current is 30 to 90 mA.
The system measures the optical signal transmitted through the breast tissue. The detection plate consists of a detector array of 56 miniaturized surface-mounted optical sensors, and the arrangement of the optical sensor array is identical to that of the LED array in the y-z plane, i.e., each optical sensor is in the mirror position of the opposite LED. The new optical sensor selected for the system is ADPD2211 (Analog Devices, INC), a 3 mm × 3 mm chip, which encapsulates a PD (diode active area: 1.8 mm2) and a low-noise current amplifier providing a 24× gain for low power consumption and near theoretical signal-to-noise ratio. Similar to the light source plate, the gaps between the optical sensors are wrapped with silicone on the side contacting the breast tissue. When designing and fabricating the printed circuit boards (PCBs) for the light source and detection (SD) plates, only the interfaces for power supply and control signals are reserved at the bottom of the PCBs and connected with flexible cables, and there are no external optical fibers or wires at other positions on the boards (source plate and detector plate in Fig. 2(a)). Under the excitation of each LED ($\textrm{S}_1^{\lambda 1}$), the current signals collected by the optical sensor array (${\textrm{D}_1} \cdots {\textrm{D}_{56}}$) are converted into the voltage signals and fed into a multi-channel DAQ card for synchronous acquisition (PCI8210, 64 channels, 16-bit conversion accuracy, ART Technology, China), with a sampling frequency of 8 KHz and a sampling time of 125 ms. At each wavelength, 56 LEDs ($\textrm{S}_1^{\lambda 1} \cdots \textrm{S}_{56}^{\lambda 1}$) are excited in sequence, and a total of 56×56 = 3136 usable SD pairs of data are measured in 15 seconds.
The source plate and the detection plate are respectively fixed on the two bases of the coaxial translation stage. The mechanical movement module drives the two plates to move close or away from each other to realize the compression and relaxation of the breast tissue. To ensure the safety of the participants, a limiter is installed on the translation stage to guarantee that the distance between the plates will not be less than the set value. An emergency switch is provided to the participant so that if the emergency switch is activated, the SD plates will immediately return to their initial positions to relax the breast.
2.2 Breast profilometry
When using the finite element method (FEM) framework for DOT reconstruction, the breast morphology needs to be obtained in advance for 3D mesh generation. The compact-DOTB system is equipped with two stereo vision depth cameras (RealSense D435, Intel), C1 and C2, respectively positioned below the space between the SD plates, as depicted in Fig. 3(a) and (b). Depending on the size of the breast, the distance between each camera and the surface of the breast tissue is about 17 cm∼19 cm. When acquiring the breast morphology, first, depth cameras C1 and C2 obtain the point cloud data of the breast surface contour from the left and right sides of the breast, respectively. The point cloud coordinates acquired by the C1 camera are then transformed to stitch together the point cloud acquired by the C2 camera (Fig. 3(c)). Next, the point cloud coordinates of the whole breast surface are transferred to the coordinate system used for DOT reconstruction by rigid transform. Finally, the boundary of the breast in the y-z plane is obtained by curve fitting using the outer contour points of the breast, and combined with the spacing of the double plates in the x-direction, a closed 3D geometry of the breast is formed. The 3D FEM mesh of the subject’s breast for DOT reconstruction is further generated using Gmsh software, which is an open-source 3D FEM mesh generator with built-in pre- and post-processing facilities [48] (Fig. 3(d)).
Fig. 3. Breast contour measurement module. (a) Photograph of two stereo vision depth cameras. (b) Schematic of silicone phantom contour acquisition by depth cameras. (c) Result of stitching the point cloud data of the phantom contour acquired by C1 with the point cloud data acquired by C2. (d) Generated 3D FEM mesh for DOT reconstruction.
2.3 Measurement procedure
We have customized a scanning control software for the compact-DOTB system named MammoView, and the procedure for measurement with the compact-DOTB system is as follows:
- (1) After applying the lotion evenly to the surface of the breast on the side to be measured, the subject lies in a prone position on the imaging bed with breast positioned inside the measurement space. The operator uses the scanning software to control the SD plates to apply a slight squeeze to the breast and stops squeezing at any time according to the feedback of the subject.
- (2) Before the measurement, the operator uses the [Fast Scan] function designed into the imaging system software to scan the breast in less than 2 seconds. The high symmetry of the SD arrangement of the compact-DOTB system makes it possible to quickly collect the measurement data at the “point-to-point projection position” at a single wavelength, i.e., ${\Gamma _{\textrm{S}_1^{\lambda 1}{\textrm{D}_1}}},\;{\Gamma _{\textrm{S}_2^{\lambda 1}{\textrm{D}_2}}}, \cdots ,{\Gamma _{\textrm{S}_{56}^{\lambda 1}{\textrm{D}_{56}}}}$, where $\Gamma $ is the optical signals acquired at different SD pairs and wavelengths. These 56 measured values form a topological image with a dimension of 7×8 reflecting the light intensity transmitted through the breast. As illustrated in Fig. 4, the black square indicates that one SD pair is covered by breast tissue and the white square indicates that the SD pair is not covered by breast tissue. The topological image obtained from the [Fast Scan] helps the operator to confirm whether the subject’s breast is centered relative to the SD plates, and if not, the operator can prompt the subject to adjust the position forwards or backwards in time.
- (3) Obtain the breast contour and the 3D FEM mesh is generated.
- (4) Collect a set of ambient light data without exciting the LEDs.
- (5) Automatic measurement: for three wavelengths, the source points are excited in turn and the detector array collects light intensity data. The total amount of data acquired is 56×56×3 = 9408 SD pairs.
- (6) After the measurement, the SD plates are moved to the initial positions respectively.
2.4 Data processing and image reconstruction
The selection of the measurement data set for image reconstruction at each wavelength (i.e., ${\Gamma _{{\textrm{S}^\lambda }\textrm{D}}}$) is based on the following two criteria: (1) The source number (e.g. i) that not covering the breast tissue is determined from the “point-to-point projection” measurement obtained by [Fast Scan], and then all measurements under the excitation of that source (${\Gamma _{{\textrm{S}_i}{\textrm{D}_m}}},m = (1, \cdots ,56)$) are excluded. (2) The separation distance between the SD plates varies for different subjects, and for each excitation source ${S_i}$, the detected data from SD pairs ${\Gamma _{{\textrm{S}_i}{\textrm{D}_m}}},m \in [1, \cdots ,56]$ with the Euclidean distance greater than 11 cm are excluded. This is based on the experimental results of the depth sensitivity of the system in section 4.2.
The DOT reconstruction is defined as an inverse problem for a given photon-propagation model. In this work, the diffusion equation (DE) with the Robin boundary condition is used to describe the light propagation in breast tissue, and the FEM framework is used for both the forward and inverse problems of the CW-DOT. We focus on reconstructing the heterogeneous distribution of the absorption coefficient ${\mu _a}$, which acts as the significant index in profiling the functional information (e.g., oxygen saturation). A Newton-type Levenberg-Marquardt (LM) algorithm is utilized to obtain the iterative update equation:
3. Methods
3.1 Phantom preparation
The background phantoms with homogeneous optical parameters used in the experiments were fabricated from polyformaldehyde (POM) with ${\mu _a} = 0.0038\;\textrm{m}{\textrm{m}^{ - 1}}$ and ${\mu ^{\prime}_s} = 0.938\;\textrm{m}{\textrm{m}^{ - 1}}$ at 670 nm [51]. To simulate different optical properties, the optical absorption targets were prepared using a combination of Intralipid (20%) (Sino-Swed Pharmaceutical Corp. Ltd., China), and India ink with different concentrations. The required concentration of diluted Intralipid (10%) corresponding to the set scattering can be determined according to the equations derived by Staveren et al. [52], while the required concentration of ink was determined according to the absorbance measured by a spectrometer. Two percentage agar powder was finally used to solidify the whole mixed solution. Before the absorber solidified, it was filled into cavities of different sizes drilled at specified locations in the POM phantom.
3.2 Linearity and stability
When measuring the breast of different sizes (thickness), the driving current of the light source module is adjusted by the control software of the compact-DOTB system, which allows the optical detection sensor to operate in the linear threshold. The linearity of the light intensity adjustment was tested. We utilized a homogeneous POM phantom with a size of $181\;\textrm{mm} \times 203\;\textrm{mm} \times 30\;\textrm{mm}$ clamped between the SD plates, as shown in Fig. 5(a), and excited any source point in the middle region of the LED array, such as No. 35 source point (at 660 nm). Taking 5 mA as the step, we set the driving current to gradually increase to 85 mA. At the same time, we recorded the voltage signal output by No. 35 detector located at the mirror position of the source point, and fitted the output signals under different driving currents. To assess the long-term stability of the system, we set 3 wavelengths for sequential excitation and measured the same POM phantom for more than 40 minutes at a sampling rate of 20 Hz at each wavelength. Then we calculated the relative error between all measured data (i.e. 50,000 output voltage data measured by the selected detector) and the mean value at each wavelength. Besides, the dynamic range of the system was also tested. Dynamic range was computed as the maximum measurable voltage of the optical sensor deducting the dark voltage divided by the standard deviation of the dark measurements.
Fig. 5. System linearity and stability testing. (a) Experimental scenario. (b) The linearity of detected voltage versus the driving current for the ${\textrm{S}_{35}}{\textrm{D}_{35}}$ pair. (c) Result of the measurement stability testing.
3.3 Depth sensitivity
To verify the effective measurement depth of the system, we used raw chicken breast tissue as the measurement sample to simulate more realistic biological tissue conditions. Marquez et al. [53] have measured the optical parameters of chicken breast tissue using the oblique incidence reflectometry method, and the calculated ${\mu _a}$ is close to that of human breast tissue in the wavelength range of 700 nm to 800 nm, but ${\mu ^{\prime}_s}$ is slightly lower than that of human breast tissue. A square mould was made to hold the chicken breast tissue in place. The height of the square frame is 45 mm and a layer of transparent polyethylene film is laid at the bottom, as shown in Fig. 6(a). The sliced chicken breast tissue was laid in the square mould layer by layer and pressed while laying to avoid interlayer gaps. After filling the mould, we also covered the surface with a layer of transparent polyethylene film, and wrapped the entire phantom around with film to achieve fixation. The sample was placed between the SD plates and the mechanical movement module was controlled to squeeze the sample slightly, resulting in a double plate spacing of 51 mm. Subsequent measurements of all SD signals were carried out under 3-wavelength excitation. In this experiment, the LED driving current is set to 45 mA.
Fig. 6. System detection depth testing. (a) Photograph of the measurement sample. Results of the linear fit of $\ln (r\Gamma )$ versus SD separations at (b) 660 nm, (c) 750 nm, and (d) 840 nm, respectively.
3.4 Positions of four cavities
The phantom is composed of three POM blocks with the same size stacked closely up and down, as illustrated in Fig. 7(a) and (b). The size of each block is $140\;\textrm{mm} \times 40\;\textrm{mm} \times 26\;\textrm{mm}$. Two cavities in each of the second and third layers of POM blocks are used to inject the optical absorber (i.e., a mixture of water and ink). The four cylindrical cavities are all 3 mm in diameter and 4 mm in depth. The optical properties of the targets are ${\mu _a} = 0.05\;\textrm{m}{\textrm{m}^{ - 1}}$ and ${\mu ^{\prime}_s} = 1\;\textrm{m}{\textrm{m}^{ - 1}}$, at 750 nm. For the final assembled 3D phantom, there are two targets in the y-z plane at x = 10 cm (position: upper left and lower right) and two targets in the y-z plane at x = 30 cm (position: lower left and upper right). We carried out DOT measurement and reconstruction at 750 nm and analyzed the positioning accuracy of the reconstructed targets. The FEM mesh for DOT reconstruction contains 65,921 linear tetrahedral elements that are joined at 15,917 nodes.
Fig. 7. Reconstruction positioning accuracy testing. (a) Geometric sketch and (b) Photograph of the phantom. Reconstruction results of the positioning phantom in y-z planes at (c) x = 10 mm, and (d) x = 30 mm, respectively. Reconstruction results of the positioning phantom in x-y planes at (e) z = 28 mm, and (f) z = 54 mm, respectively. (g) 3D DOT reconstruction result of the phantom.
3.5 Position and magnitude of an absorption change
In this experiment, a semi-cylindrical phantom similar to the morphology of the breast was designed, as shown in Fig. 8(a). The phantom consists of two POM blocks of identical dimensions, each with a diameter of 128.5 mm and a thickness of 22 mm. one of the POM blocks has two cylindrical cavities in the central area for injecting the optical absorber and the other with homogeneous optical properties. The diameter of the two cavities is 7.5 mm, the depth is 10 mm, and the center-to-center distance is 40 mm. The optical absorption coefficients of the two targets are ${\mu _a} = 0.01\;m{m^{ - 1}}$ and ${\mu _a} = 0.03\;m{m^{ - 1}}$ at 660 nm, respectively, and the reduced scattering coefficients are the same, i.e., ${\mu ^{\prime}_s} = 1\;m{m^{ - 1}}$. The two POM blocks were stacked closely front to back, and the targets were located near the middle layer between the SD plates. We have performed DOT measurements and reconstructions at 660 nm, 750 nm and 840 nm respectively, and analyzed the quantitative reconstruction results of ${\mu _a}$. In this experiment, the FEM mesh contains 114,603 linear tetrahedral elements that are joined at 22,134 nodes.
Fig. 8. Reconstruction contrast testing. (a) Geometric sketch of the phantom. Reconstruction results of the contrast phantom at (b) 660 nm, (c) 750 nm, and (d) 840 nm, respectively. ${\mu _a}$ profiles through the centers of the two targets at (e) 660 nm, (f) 750 nm, and (g) 840 nm, respectively.
3.6 Patient measurements before and after NAC
This is one example from an ongoing clinical study involving many patients with a planned enrollment of more than 20. The subject was a 35-years-old female with a body mass index (BMI) of 23 kg/m2 undergoing the NAC treatment for cancer of the left breast. This study was approved by the Ethics Committee of Xijing Hospital (Approval No. KY20212008-F-1). The patient was diagnosed by core biopsy revealing invasive carcinoma in the left breast. The molecular subtype of the tumor was HER2-positive. DOT and PET measurements were performed on the patient in the Nuclear Medicine Department of Xijing Hospital. For PET imaging, the patient was injected with 390 MBq 18F-FDG and scanned after 97 min post injection before the first cycle of the NAC treatment. After first cycle of the treatment, the patient was injected with 405.6 MBq 18F-FDG and scanned after 63 min post injection. For DOT scanning, the whole measurements were carried out in a dark room, and the spacing of the SD plates was 60 mm in both measurements. In DOT reconstruction, the FEM mesh contains 66,620 linear tetrahedral elements that are joined at 13,748 nodes.
4. Results
The presented results include: 1) the linearity of output voltage versus driving current, 2) the stability of output voltage during 40 minutes, 3) the depth sensitivity measured on chicken breast tissue, 4) the retrieved positioning accuracy of four cavities inside a layered phantom, 5) the accuracy of the retrieved location and magnitude of an absorption change that was introduced inside a contrast phantom, and 6) the results of measurements on a cancer patient before and after NAC treatment.
4.1 linearity and stability
Figure 5(b) is the outputs of the detection module for different light source driving currents. The linear fitting results show that the R2 value of the system is close to 1, indicating that the light intensity adjustment of the compact-DOTB system has ideal linearity. Figure 5(c) illustrates the relative error among all detected values and the mean value for a long period of measurement at each wavelength. Under the continuous excitation of each wavelength for more than 40 minutes, the fluctuations of the measured values are less than ±0.1%, demonstrating the excellent long-term stability of the system. Besides, we measured the dark noise of the system under the condition of avoiding ambient light and without excitation light, and its standard deviation was ${\textrm{V}_{\textrm{std}}} = 0.183\;\textrm{mV}$. According to the formula dynamic range (DR) $\textrm{DR} = \textrm{20} \times \textrm{lg(}{{{{\mathrm{\bar{V}}}_{\textrm{max}}}} / {{\textrm{V}_{\textrm{std}}}}}\textrm{)}$, the calculated DR of the system is 88.439 dB, where ${\mathrm{\bar{V}}_{\textrm{max}}} = 4755.838\;\textrm{mV}$ represents the maximum output value after deducting the dark voltage.
4.2 Depth sensitivity
The propagated light intensity falls with the distance travelled and for the case of biological tissue where ${\mu ^{\prime}_s} \gg {\mu _a}$, the following relationship can be obtained from the diffusion equation.
where r is the distance between the light source point and the detector point, D is the diffusion coefficient, defined as $1/3({\mu ^{\prime}_s} + {\mu _a})$. $\sqrt {{{{\mu _a}} / D}} $ is related to the tissue optical properties, and if the optical parameters inside the tissue are homogeneous, then $\ln (r\Gamma )$ and r are linearly related, with a slope of $- \sqrt {{{{\mu _a}} / D}} $. Since more SD distances can be formed between the source points located at the four corners and the detector array on the opposite side, we selected the light intensity datasets obtained under the excitation by the No. 1, No. 8, No. 49 and No. 56 LEDs respectively, i.e., ${\Gamma _{\textrm{S}_i^\lambda {\textrm{D}_m}}},\;({i \in [1,8,49,56],\;m = [1, \cdots ,56]} )$. $\ln (r\Gamma )$ versus SD separations ($r$) at the three excitation wavelengths are plotted in Figs. 6(b) to (d), respectively. It can be seen from the results that at 660 nm, the value of $\ln (r\Gamma )$ gradually deviates from the fitted line when the SD distance is greater than 100 mm, reflecting that the maximum effective detection distance of the imaging system at 660 nm is about 100 mm. At 750 nm and 840 nm, the data points still obey a linear distribution at SD distances less than 115 mm, indicating that the upper limit of the effective detection distance of the imaging system at the wavelengths of 750 nm and 840 nm is about 110 mm. The depth sensitivity is comparable to the modern presented DOT system [33].4.3 Positions of four cavities
The results of the DOT reconstruction of the positioning phantom are illustrated in Fig. 7, where the reconstructions of the two sagittal planes (i.e., the y-z planes at x = 10 mm and x = 30 mm) passing through the centers of the two groups of optical absorbers (targets: 1-1, 1-2 and targets: 2-1, 2-2) respectively are shown in Figs. 7(c) and (d), with the asterisks in the figure indicating the real positions of the centers of the four targets. In order to quantitatively evaluate the localization accuracy of the reconstruction, we calculated the centroid coordinates of the four targets in the DOT reconstruction results, and calculated the Euclidean distance between these coordinates and the real ones, respectively. The results are listed in Table 1 demonstrating that the system has satisfactory positioning accuracy for the optical absorbers, with the position errors of less than 1.5 mm in the y-z plane for the positioning phantom. The reconstructed images are comparable to the results of other DOT systems [32,54]. In Fig. 7(d), some faint artifacts appear around z = 60 mm and y = 40∼60 mm. This may be due to the small gap between the layers 2 and 3 in the three-layer phantom structure, which leads to the interlayer reflection during photon propagation. Figure 7(g) depicts the 3D DOT reconstruction result of the phantom, and the reconstructions of the two x-y planes where the two groups of absorbers (targets: 1-1, 2-1 and targets: 1-2, 2-2) are located are shown in Fig. 7(e) and (f), respectively. It can be noticed that in the reconstructed images in the x-y plane, the targets are slightly elongated in the x-direction, which is also a drawback of measuring only in the transmission mode.
Table 1. Centroid Coordinates of the Four Reconstructed Targets and the Euclidean Distance between the Coordinates and the Real Ones
4.4 Position and magnitude of an absorption change
At the wavelengths of 660 nm, 750 nm and 840 nm, the DOT results of the phantom with absorption contrast targets are shown in Figs. 8(b) to (d), respectively. The images exhibit the distribution of ${\mu _a}$ in the reconstructed sagittal plane (y-z plane) at x = 18 mm, and the results indicate that the absorption targets can be recovered at all three wavelengths. Meanwhile, since the India ink was used as an absorber when making the target, and the absorbance of its aqueous solution decreases as the wavelength increases, the ${\mu _a}$ value of the reconstructed targets at 660 nm are greater than those at the other two wavelengths, as indicated by the ${\mu _a}$ profiles through the centers of the two targets (Figs. 8(e) to (g)). The two targets are expected to be prepared with an optical absorption contrast of 3:1 at 660 nm, and at this wavelength the reconstructed regional ${\mu _a}$ contrast of the two targets is 0.0160/0.0084 = 1.8984. In addition, at 750 nm and 840 nm, the reconstructed contrasts are 0.0146/0.0075 = 1.9449 and 0.0132/0.0071 = 1.8659, respectively. The results demonstrate that a reasonable quantitative contrast can be obtained and the errors in the absolute quantification of the optical parameters may also stem from the differences between the prepared optical properties of the targets and the ideal ones.
4.5 Patient measurements before and after NAC
PET images are used as cross validation for evaluating the performance of the proposed compact-DOTB. Figure 9 shows the DOT results of the patient before the NAC treatment. According to the reconstructed PET result (Fig. 9(a)), the tumor is located in the right side of the left breast with maximum standard uptake value (SUV) of 14.09. Figures 9(b) to (e) illustrate the concentration images of [Hb], [HbO2], [HbT], and [StO2], respectively. The averaged tumor to background (T/B) contrast is calculated to be 2.30× for [Hb], 1.37× for [HbO2], 1.80× for [HbT], and 0.76× for [StO2]. Due to the enhanced metabolism of the tumor area, the lesion region exhibits low blood oxygen saturation and high hemoglobin concentration. Busch et al. have found that increases in [HbT] and ${\mu ^{\prime}_s}$ contrast showed correspondence with similar high-FDG regions in the PET images [55]. Our study is consistent with the literature, but the current analysis is incomplete due to the small number of the cases. Figure 10 shows the DOT results of the patient after first cycle of the NAC treatment. According to the PET result (Fig. 10(a)), the tumor volume is significantly reduced after chemotherapy with a maximum SUV of 2.41. For the DOT reconstruction results, as illustrated in Figs. 10(b) to (e), the averaged T/B contrast is calculated to be 1.34× for [Hb], 1.57× for [HbO2], 1.44× for [HbT], and 1.08× for [StO2]. The results demonstrate that the DOT measurements are sensitive to local metabolism of the tissue and have a high consistency with the imaging results of PET. Corresponding to the decrease in SUV values of PET images, it can be clearly observed that [HbT] decreases and [StO2] increases in the tumor region after the treatment. These results are consistent with the findings of the Refs. [5] and [7].
Fig. 9. Reconstruction results of the patient’s left breast in y-z plane at x = 30 mm before the treatment. (a) SUV distribution reconstructed from PET. (b) Reconstructed concentration distribution of deoxyhemoglobin ([Hb]). (c) Reconstructed concentration distribution of oxyhemoglobin ([HbO2]). (d) Reconstructed concentration distribution of total hemoglobin (HbT). (e) Reconstructed oxygen saturation distribution ([StO2]).
Fig. 10. Reconstruction results of the patient’s left breast in y-z plane at x = 30 mm after one cycle of the treatment. (a) SUV distribution reconstructed from PET. (b) Reconstructed concentration distribution of deoxyhemoglobin ([Hb]). (c) Reconstructed concentration distribution of oxyhemoglobin ([HbO2]). (d) Reconstructed concentration distribution of total hemoglobin (HbT). (e) Reconstructed oxygen saturation distribution ([StO2]).
5. Discussion
The system is designed with a highly stable constant-current driving circuit, which enables the system to have a linearity of light intensity adjustment with R2 values close to 1, and to keep the fluctuation of the measured values of all three wavelengths within 0.1% under the long-term measurement condition (∼40 minutes). For the detection module, due to the integration of an on-board precision current amplifier with a 24× gain, the phantom and in vivo experiments have shown that the system can obtain a measurement dynamic range of 88.439 dB to meet the needs of the breast measurement even without the addition of external trans-impedance amplifier. Biological tissue experiment has indicated that the compact-DOTB system is capable of detecting breast scales with a maximum effective SD distance greater than 10 cm at three excitation wavelengths. The phantom experiments have verified that the system has satisfactory reconstruction localization accuracy and reasonable recovery contrast for detecting and imaging large-scale targets, and is able to obtain high quantitative accuracy for optical targets even with small absorption coefficients.
The parallel-plane based measurement structure makes it easier to locate the region of interest from the reconstruction results in clinical use compared to the free-space measurement structure. For the transmission measurement mode, high reconstruction accuracy can be obtained in the plane parallel to the plates, i.e., the sagittal plane (y-z plane of the compact-DOTB system), as shown in Figs. 7(c) and (d). In the later system upgrade, we are also considering inserting encrypted arrangement of source and detector points on both plates to further improve the imaging quality. At the same time, we are considering using multiple imaging modalities to cooperate with DOT [56]. While providing richer functional information for diagnosis, the use of functional or structural a priori information provided by other modalities that can further constrain the DOT reconstruction to improve imaging resolution, e.g., photoacoustic tomography-guided DOT reconstruction that studied in our previous work [57], and the breast PET imaging that we are retrofitting in the compact-DOTB measuring bed.
In the in vivo experiment, thanks to the breast morphology acquisition module built by the depth cameras, the subject’s breast contour can be accurately extracted to generate 3D FEM mesh for DOT reconstruction. In this subject experiment, based on the proposed two guidelines for the data exclusion, a total of 441 measurements at each wavelength were involved in the DOT reconstruction. The patient reported here is undergoing the NAC treatment, and the DOT imaging performed prior to the treatment demonstrates an abundant blood supply in the tumor tissue region and a lower [StO2] than that in normal breast tissue. The DOT results provided by the compact-DOTB system also indicate that these parameters are changing with the NAC treatment, such as a reduced [HBT] and a slightly elevated [StO2]. [HbT] contrast has been always presented as the most important indicator in breast cancer diagnosis and tumor response monitoring to NAC [20]. Nevertheless, several publications illustrate that locally increased hemoglobin concentration is not only a feature of cancer but also of many benign tissue alterations [2]. In addition to [HBT] indicator, [StO2] also plays a very important role in the diagnosis of benign and malignant breast lesion. Besides, to enhance the ability of the compact-DOTB system to distinguish the benign and malignant lesions, we will employ the system to perform “squeezing-relaxing” measurements of the breast at 660 nm, where the optical absorption of [Hb] is dominant, and focus on the change in the optical absorption of the breast tissue in both states. This is an additional diagnostic criterion because that the increased vascularity associated with the growth of malignant lesions is different from the vascularity supporting benign and normal tissue and interstitial fluid pressure is elevated around malignant tumors. Such vessels in the breast also behaves differently in response to pressure-regulated stimuli. Moreover, considering that the most sensitive response wavelength center of the selected optical sensor is about 840 nm, we can add longer wavelength excitation LEDs to the source plate in subsequent work to more accurately recover the content of, for example, lipids and collagen, etc. The content of these chromophores is also related to the benign and malignant of lesion.
For the treatment of breast cancer, such as NAC treatment, the experimental results in this paper demonstrate the ability of the proposed compact-DOTB system to provide contrast indicators of [HBT] and [StO2] before and after the intervention to determine whether the treatment is effective. The proposed system meets the requirements of easy maintenance and compactness in clinical applications, while its cost is competitive with the existing breast-DOT devices.
6. Conclusions
In this paper, we introduce a compact fiber-free parallel-plane multi-wavelength DOT imager with 3D imaging capability for breast tumor diagnosis and NAC process monitoring. The light source matrix is composed of surface-mounted LEDs integrated with three wavelengths, and the detection matrix consists of the miniaturized optical sensors each encapsulating a high-sensitive PD and a low-noise current amplifier, capable of providing 3136 SD pairs of measurement data at each excitation wavelength. The fiber-free design ensures the consistency between source/detection channels to a large extent, while effectively reducing the complexity of system construction and maintenance.
Funding
National Key Research and Development Program of China (2016YFC0103800); National Natural Science Foundation of China (61471279, 61901342); Natural Science Basic Research Program of Shaanxi Province (2021JZ-29, 2021SF-131).
Disclosures
The authors declare that there are no conflicts of interest related to this article.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. M. Akraml, M. Iqbal, M. Daniyal, and A. U. Khan, “Awareness and current knowledge of breast cancer,” Biol. Res. 50(1), 33 (2017). [CrossRef]
2. D. Grosenick, H. Rinneberg, R. Cubeddu, and P. Taroni, “Review of optical breast imaging and spectroscopy,” J. Biomed. Opt. 21(9), 091311 (2016). [CrossRef]
3. H. G. Welch, P. C. Prorok, A. J. O’Malley, and B. S. Kramer, “Breast-Cancer Tumor Size, Overdiagnosis, and Mammography Screening Effectiveness,” N. Engl. J. Med. 375(15), 1438–1447 (2016). [CrossRef]
4. B. J. Tromberg, Z. Zhang, A. Leproux, T. D. O’Sullivan, A. E. Cerussi, P. M. Carpenter, R. S. Mehta, D. Roblyer, W. Yang, K. D. Paulsen, B. W. Pogue, S. D. Jiang, P. A. Kaufman, A. G. Yodh, S. H. Chung, M. Schnall, B. S. Snyder, N. Hylton, D. A. Boas, S. A. Carp, S. J. Isakoff, and D. Mankoff, “Predicting Responses to Neoadjuvant Chemotherapy in Breast Cancer: ACRIN 6691 Trial of Diffuse Optical Spectroscopic Imaging,” Cancer Res. 76(20), 5933–5944 (2016). [CrossRef]
5. Q. Zhu, S. Tannenbaum, S. H. Kurtzman, P. DeFusco, A. R. Jr, H. Vavadi, F. Zhou, C. Xu, A. Merkulov, P. Hegde, M. Kane, L. Wang, and K. Sabbath, “Identifying an early treatment window for predicting breast cancer response to neoadjuvant chemotherapy using immunohistopathology and hemoglobin parameters,” Breast Cancer Res. 20(1), 56 (2018). [CrossRef]
6. M. L. Altoe, K. Kalinsky, A. Marone, H. K. Kim, H. Guo, H. Hibshoosh, M. Tejada, K. D. Crew, M. K. Accordino, M. S. Trivedi, D. L. Hershman, and A. H. Hielscher, “Effects of neoadjuvant chemotherapy on the contralateral non-tumor-bearing breast assessed by diffuse optical tomography,” Breast Cancer Res. 23(1), 16 (2021). [CrossRef]
7. M. L. Altoe, A. Marone, H. K. Kim, K. Kalinsky, D. L. Hershman, A. H. Hielscher, and R. S. Ha, “Diffuse optical tomography of the breast: a potential modifiable biomarker of breast cancer risk with neoadjuvant chemotherapy,” Biomed. Opt. Express 10(8), 4305–4315 (2019). [CrossRef]
8. J. E. Gunther, E. A. Lim, H. K. Kim, M. Flexman, M. Altoe, J. A. Campbell, H. Hibshoosh, K. D. Crew, K. Kalinsky, D. L. Hershman, and A. H. Hielscher, “Dynamic Diffuse Optical Tomography for Monitoring Neoadjuvant Chemotherapy in Patients with Breast Cancer,” Radiology 287(3), 778–786 (2018). [CrossRef]
9. J. M. Cochran, D. R. Busch, A. Leproux, Z. Zhang, T. D. O’Sullivan, A. E. Ceryssi, P. M. Carpenter, R. S. Mehta, D. Roblyer, W. Yang, K. D. Paulsen, B. Pogue, S. D. Jiang, P. A. Kaufman, S. H. Chung, M. Schnall, B. S. Snyder, N. Hylton, S. A. Carp, S. J. Isakoff, D. Mankoff, B. J. Tromberg, and A. G. Yodh, “Tissue oxygen saturation predicts response to breast cancer neoadjuvant chemotherapy within 10 days of treatment,” J. Biomed. Opt. 24(2), 021202 (2019). [CrossRef]
10. B. W. Pogue, M. Testorf, T. Mcbride, U. Osterberg, and K. Paulsen, “Instrumentation and design of a frequency-domain diffuse optical tomography imager for breast cancer detection,” Opt. Express 1(13), 391–403 (1997). [CrossRef]
11. B. W. Pogue, S. P. Poplack, T. O. Mcbride, W. A. Wells, K. S. Osterman, U. L. Osterberg, and K. D. Paulsen, “Quantitative hemoglobin tomography with diffuse near-infrared spectroscopy: pilot results in the breast,” Radiology 218(1), 261–266 (2001). [CrossRef]
12. H. Jiang, Y. Xu, N. Iftimia, J. Eggert, K. Klove, L. Baron, and L. Fajardo, “Three-dimensional optical tomographic imaging of breast in a human subject,” IEEE Trans. Med. Imaging 20(12), 1334–1340 (2001). [CrossRef]
13. B. W. Pogue, S. D. Jiang, H. Dehghani, C. Kogel, S. Soho, S. Srinivasan, X. M. Song, T. D. Tosteson, S. P. Poplack, and K. D. Paulsen, “Characterization of hemoglobin, water, and NIR scattering in breast tissue: analysis of intersubject variability and menstrual cycle changes,” J. Biomed. Opt. 9(3), 541–552 (2004). [CrossRef]
14. T. Yates, J. C. Hebden, A. Gibson, N. Everdell, S. R. Arridge, and M. Douek, “Optical tomography of the breast using a multi-channel time-resolved imager,” Phys. Med. Biol. 50(11), 2503–2517 (2005). [CrossRef]
15. C. Q. Li, H. Z. Zhao, B. Anderson, and H. B. Jiang, “Multispectral breast imaging using a ten-wavelength, 64×64 source/detector channels silicon photodiode-based diffuse optical tomography system,” Med. Phys. 33(3), 627–636 (2006). [CrossRef]
16. M. L. Flexman, M. A. Khalil, R. Al Abdi, H. K. Kim, C. J. Fong, E. Desperito, D. L. Hershman, R. L. Barbour, and A. H. Hielscher, “Digital optical tomography system for dynamic breast imaging,” J. Biomed. Opt. 16(7), 076014 (2011). [CrossRef]
17. M. L. Flexman, H. K. Kim, J. E. Gunther, E. A. Lim, M. C. Alvarez, E. Desperito, K. Kalinsky, D. L. Hershman, and A. H. Hielscher, “Optical biomarkers for breast cancer derived from dynamic diffuse optical tomography,” J. Biomed. Opt. 18(9), 096012 (2013). [CrossRef]
18. F. El-Ghussein, M. A. Mastanduno, S. D. Jiang, B. W. Pogue, and K. D. Paulsen, “Hybrid photomultiplier tube and photodiode parallel detection array for wideband optical spectroscopy of the breast guided by magnetic resonance imaging,” J. Biomed. Opt. 19(1), 011010 (2014). [CrossRef]
19. Y. Zhao, B. W. Pogue, S. J. Haider, J. Gui, R. M. diFlorio-Alexander, K. D. Paulsen, and S. D. Jiang, “Portable, parallel 9-wavelength near-infrared spectral tomography (NIRST) system for efficient characterization of breast cancer within the clinical oncology infusion suite,” Biomed. Opt. Express 7(6), 2186–2201 (2016). [CrossRef]
20. Y. Zhao, W. R. Burger, M. W. Zhou, E. B. Bernhardt, P. A. Kaufman, R. R. Patel, C. V. Angeles, B. W. Pogue, K. D. Paulsen, and S. D. Jiang, “Collagen quantification in breast tissue using a 12-wavelength near infrared spectral tomography (NIRST) system,” Biomed. Opt. Express 8(9), 4217–4229 (2017). [CrossRef]
21. C. H. Schmitz, D. P. Klemer, R. Hardin, M. S. Katz, Y. L. Pei, H. L. Graber, M. B. Levin, R. D. Levina, N. A. Franco, W. B. Solomon, and R. L. Barbour, “Design and implementation of dynamic near-infrared optical tomographic imaging instrumentation for simultaneous dual-breast measurements,” Appl. Opt. 44(11), 2140–2153 (2005). [CrossRef]
22. L. C. Enfield, A. P. Gibson, N. L. Everdell, D. T. Delpy, M. Schweiger, S. R. Arridge, C. Richardson, M. Keshtgar, M. Douek, and J. C. Hebden, “Three-dimensional time-resolved optical mammography of the uncompressed breast,” Appl. Opt. 46(17), 3628–3638 (2007). [CrossRef]
23. W. Zhang, L. H. Wu, J. Li, X. Yi, X. Wang, Y. M. Lu, W. T. Chen, Z. X. Zhou, L. M. Zhang, H. J. Zhao, and F. Gao, “Combined hemoglobin and fluorescence diffuse optical tomography for breast tumor diagnosis: a pilot study on time-domain methodology,” Biomed. Opt. Express 4(2), 331–348 (2013). [CrossRef]
24. A. Pifferi, P. Taroni, A. Torricelli, F. Messina, R. Cubeddu, and G. Danesini, “Four-wavelength time-resolved optical mammography in the 680-980 nm range,” Opt. Lett. 28(13), 1138–1140 (2003). [CrossRef]
25. D. Grosenick, K. T. Moesta, H. Wabnitz, J. Mucke, C. Stroszczynski, R. Macdonald, P. M. Schlag, and H. Rinneberg, “Time-domain optical mammography: initial clinical results on detection and characterization of breast tumors,” Appl. Opt. 42(16), 3170–3186 (2003). [CrossRef]
26. J. P. Culver, R. Choe, M. J. Holboke, L. Zubkov, T. Durduran, A. Slemp, V. Ntziachristos, B. Chance, and A. G. Yodh, “Three-dimensional diffuse optical tomography in the parallel plane transmission geometry: Evaluation of a hybrid frequency domain/continuous wave clinical system for breast imaging,” Med. Phys. 30(2), 235–247 (2003). [CrossRef]
27. R. Choe, S. D. Konecky, A. Corlu, K. Lee, T. Durduran, D. R. Busch, S. Pathak, B. J. Czerniecki, J. Tchou, D. L. Fraker, A. DeMichele, B. Chance, S. R. Arridge, M. Schweiger, J. P. Culver, M. D. Schnall, M. E. Putt, M. A. Rosen, and A. G. Yodh, “Differentiation of benign and malignant breast tumors by in-vivo three-dimensional parallel-plate diffuse optical tomography,” J. Biomed. Opt. 14(2), 024020 (2009). [CrossRef]
28. A. Hagen, D. Grosenick, R. Macdonald, H. Rinneberg, S. Burock, P. Warnick, A. Poellinger, and P. M. Schlag, “Late-fluorescence mammography assesses tumor capillary permeability and differentiates malignant from benign lesions,” Opt. Express 17(19), 17016–17033 (2009). [CrossRef]
29. Q. Q. Fang, S. A. Carp, J. Selb, G. Boverman, Q. Zhang, D. B. Kopans, R. H. Moore, E. L. Miller, D. H. Brooks, and D. A. Boas, “Combined Optical Imaging and Mammography of the Healthy Breast: Optical Contrast Derived From Breast Structure and Compression,” IEEE Trans. Med. Imaging 28(1), 30–42 (2009). [CrossRef]
30. D. Grosenick, A. Hagen, O. Steinkellner, A. Poellinger, S. Burock, P. M. Schlag, H. Rinneberg, and R. Macdonald, “A multichannel time-domain scanning fluorescence mammograph: Performance assessment and first in vivo results,” Rev. Sci. Instrum. 82(2), 024302 (2011). [CrossRef]
31. S. A. Carp, A. Y. Sajjadi, C. M. Wanyo, Q. Q. Fang, M. C. Specht, L. Schapira, B. Moy, A. Bardia, D. A. Boas, and S. J. Isakoff, “Hemodynamic signature of breast cancer under fractional mammographic compression using a dynamic diffuse optical tomography system,” Biomed. Opt. Express 4(12), 2911–2924 (2013). [CrossRef]
32. H. Y. Ban, M. Schweiger, V. C. Kavuri, J. M. Cochran, L. Xie, D. R. Busch, J. Katrasnik, S. Pathak, S. H. Chung, K. Lee, R. Choe, B. J. Czerniecki, S. R. Arridge, and A. G. Yodh, “Heterodyne frequency-domain multispectral diffuse optical tomography of breast cancer in the parallel-plane transmission geometry,” Med. Phys. 43(7), 4383–4395 (2016). [CrossRef]
33. B. B. Zimmermann, B. Deng, B. Singh, M. Martino, J. Selb, Q. Q. Fang, A. Y. Sajjadi, J. Cormier, R. H. Moore, D. B. Kopans, D. A. Boas, M. A. Saksena, and S. A. Carp, “Multimodal breast cancer imaging using coregistered dynamic diffuse optical tomography and digital breast tomosynthesis,” J. Biomed. Opt. 22(4), 046008 (2017). [CrossRef]
34. J. M. Cochran, D. R. Busch, L. Lin, D. L. Minkoff, M. Schweiger, S. Arridge, and A. G. Yodh, “Hybrid time-domain and continuous-wave diffuse optical tomography instrument with concurrent, clinical magnetic resonance imaging for breast cancer imaging,” J. Biomed. Opt. 24(05), 1 (2019). [CrossRef]
35. E. Y. Chae, H. H. Kim, S. Sabir, Y. Kim, H. Kim, S. Yoon, J. C. Ye, S. Cho, D. Heo, K. H. Kim, Y. M. Bae, and Y. M. Choi, “Development of digital breast tomosynthesis and diffuse optical tomography fusion imaging for breast cancer detection,” Sci. Rep. 10(1), 13127 (2020). [CrossRef]
36. R. M. Al Abdi, B. Deng, H. H. Hijazi, M. Wu, and S. A. Carp, “Mechanical and hemodynamic responses of breast tissue under mammographic-like compression during functional dynamic optical imaging,” Biomed. Opt. Express 11(10), 5425–5441 (2020). [CrossRef]
37. A. Cerussi, N. Shah, D. Hsiang, A. Durkin, J. Butler, and B. J. Tromberg, “In vivo absorption, scattering, and physiologic properties of 58 malignant breast tumors determined by broadband diffuse optical spectroscopy,” J. Biomed. Opt. 11(4), 044005 (2006). [CrossRef]
38. G. Q. Yuan, U. Alqasemi, A. Chen, Y. Yang, and Q. Zhu, “Light-emitting diode-based multiwavelength diffuse optical tomography system guided by ultrasound,” J. Biomed. Opt. 19(12), 126003 (2014). [CrossRef]
39. J. M. Cochran, A. Leproux, D. R. Busch, T. D. O’Sullivan, W. Yang, R. S. Mehta, A. M. Police, B. J. Tromberg, and A. G. Yodh, “Breast cancer differential diagnosis using diffuse optical spectroscopic imaging and regression with z-score normalized data,” J. Biomed. Opt. 26(02), 026004 (2021). [CrossRef]
40. H. Vavadi, A. Mostafa, F. Zhou, K. M. S. Uddin, M. Althobaiti, C. Xu, R. Bansal, F. Ademuyiwa, S. Poplack, and Q. Zhu, “Compact ultrasound-guided diffuse optical tomography system for breast cancer imaging,” J. Biomed. Opt. 24(2), 021203 (2019). [CrossRef]
41. S. S. Spink, F. Teng, V. Pera, H. M. Peterson, T. Cormier, A. S. Budge, D. Chargin, S. Brookfield, A. T. Eggebrecht, N. Ko, and D. Roblyer, “High optode-density wearable diffuse optical probe for monitoring paced breathing hemodynamics in breast tissue,” J. Biomed. Opt. 26(06), 062708 (2021). [CrossRef]
42. F. Teng, T. Cormier, A. Sauer-Budge, R. Chaudhury, V. Pera, R. Istfan, D. Chargin, S. Brookfield, N. Y. Ko, and D. M. Roblyer, “Wearable near-infrared optical probe for continuous monitoring during breast cancer neoadjuvant chemotherapy infusions,” J. Biomed. Opt. 22(1), 014001 (2017). [CrossRef]
43. A. T. Eggebrecht, S. L. Ferradal, A. Robichaux-Viehoever, M. S. Hassanpour, H. Dehghani, A. Z. Snyder, T. Hershey, and J. P. Culver, “Mapping distributed brain function and networks with diffuse optical tomography,” Nat. Photonics 8(6), 448–454 (2014). [CrossRef]
44. B. S. Nichols, A. Llopis, G. M. Palmer, III S. S. McCachren, O. Senlik, D. Miller, M. A. Brooke, N. M. Jokerst, J. Geradts, R. Greenup, and N. Ramanujam, “Miniature spectral imaging device for wide-field quantitative functional imaging of the morphological landscape of breast tumor margins,” J. Biomed. Opt. 22(2), 026007 (2017). [CrossRef]
45. B Dolgoshein, V. Balagura, P. Buzhan, M. Danilovb, L. Filatovd, E. Garuttif, M. Groll, A. Ilyin, V. Kantserov, V. Kaplin, A. Karakash, F. Kayumov, S. Klemin, V. Korbel, H. Meyer, R. Mizukb, V. Morgunovb, E. Novikovb, P. Pakhlovb, E. Popova, V. Rusinovb, F. Sefkow, E. Tarkovskyb, and I. Tikhomirovb, “Status report on silicon photomultiplier development and its applications,” Nucl. Instrum. Methods Phys. Res. A 563(2), 368–376 (2006). [CrossRef]
46. C. Ma, R. Chen, J. Zhao, M. Xu, H. Wu, Y. Sheng, Y. Wang, and S. Zhu, “Fiber-free parallel-plane continuous wave breast diffuse optical tomography system,” Proc. SPIE 11229, 112290L (2020). [CrossRef]
47. A. Lyons, F. Tonolini, A. Boccolini, A. Repetti, and D. Faccio, “Computational time-of-flight diffuse optical tomography,” Nat. Photonics 13(8), 575–579 (2019). [CrossRef]
48. C. Geuzaine and J. F. Remacle, “Gmsh: a three-dimensional finite element mesh generator with built-in pre- and post-processing facilities,” Int. J. Numer. Methods Eng. 79(11), 1309–1331 (2009). [CrossRef]
49. X. Intes, “Time-Domain Optical Mammography SoftScan: Initial Results,” Acad. Radiol. 12(8), 934–947 (2005). [CrossRef]
50. H. Dehghani, M. E. Eames, P. K. Yalavarthy, S. C. Davis, S. Srinivasan, C. M. Carpenter, B. W. Pogue, and K. D. Paulsen, “Near infrared optical tomography using NIRFAST: Algorithm for numerical model and image reconstruction,” Commun. Numer. Methods Eng. 25(6), 711–732 (2009). [CrossRef]
51. F. Gao, J. Li, L. Zhang, P. Poulet, H. Zhao, and Y. Yamada, “Simultaneous fluorescence yield and lifetime tomography from time-resolved transmittances of small-animal-sized phantom,” Appl. Opt. 49(16), 3163–3172 (2010). [CrossRef]
52. H. J. van Staveren, C. J. Moes, J. van Marie, S. A. Prahl, and M. J. van Gemert, “Light scattering in Intralipid-10% in the wavelength range of 400-1100 nm,” Appl. Opt. 30(31), 4507–4514 (1991). [CrossRef]
53. G. Marquez, L. V. Wang, S. P. Lin, J. A. Schwartz, and S. L. Thomsen, “Anisotropy in the absorption and scattering spectra of chicken breast tissue,” Appl. Opt. 37(4), 798–804 (1998). [CrossRef]
54. J. W. Hoi, H. K. Kim, C. J. Fong, L. Zweck, and A. H. Hielscher, “Non-contact dynamic diffuse optical tomography imaging system for evaluating lower extremity vasculature,” Biomed. Opt. Express 9(11), 5597–5614 (2018). [CrossRef]
55. D. R. Busch, R. Choe, T. Durduran, and A. G. Yodh, “Towards noninvasive characterization of breast cancer and cancer metabolism with diffuse optics,” PET Clin. 8(3), 345 (2013). [CrossRef]
56. Q. Zhu and S. Poplack, “A review of optical breast imaging: Multi-modality systems for breast cancer diagnosis,” Eur. J. Radiol. 129, 109067 (2020). [CrossRef]
57. Y. Wang, J. Li, T. Lu, L. Zhang, Z. Zhou, H. Zhao, and F. Gao, “Combined diffuse optical tomography and photoacoustic tomography for enhanced functional imaging of small animals: a methodological study on phantoms,” Appl. Opt. 56(2), 303–311 (2017). [CrossRef]
