Open Access
Add to BibSonomy
Abstract
There is a strong need for non-invasive detection of normal tissue from diseased one and a better understanding of the factors involved in the infection’s growth. Continuous monitoring of tissue samples at different time points is highly desirable. We demonstrate using the photoacoustic spectral response technique (PASR) for in situ analysis in a developing chicken embryo as a model (CAM) for anti-angiogenesis and vascular development. The photoacoustic technique is an emerging modality that is based on the acoustic detection of optical absorption of biological samples. The detected PA signals and their spectral response were used as a signature to determine the vasculature development pathology. Continuous monitoring of vascular growth and an anti-drug (Cisplatin) effect on vasculature has been done. PASR was investigated for the 10th day, 11th day, and 12th day control and inoculated egg samples. It shows that the dominant frequency of the PA spectral response for 10th day control and inoculated eggs lies between 0.45–0.52 MHz, whereas for 11th day and 12th day control eggs lie at 0.61 ± 0.152 MHz and 0.67 ± 0.001 MHz, respectively. The inoculated 11th and 12th day eggs lie at 0.35 ± 0.156 MHz and 0.16 ± 0.004 MHz, respectively. PASR could monitor the change in growth within a span of one day, which was not possible through the conventional imaging approach. This would open up a potential diagnostic technique for continuous monitoring of CAM assays.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Over the past century, biomedical sensing has evolved from doing radiography (X-Ray) to magnetic resonance imaging (MRI) [1–3]. However, high-resolution imaging and subsequent real-time visualization of tissues or organs are limited to a few microns of depth. Beyond this depth, deep tissue sensing can only be performed by fixing the tissue and follow up with histology [4]. This procedure involves sacrificing the animal or extracting the body’s tissue sample, thereby limiting the biologists to continuously monitor the pathology or assay’s progress. Biologists traditionally depend on the histology of excised tissues at different time points to study any dynamic process. With the advent of deep tissue sensing techniques, non-invasive in-vivo sensing has caught up with many clinical scientists [5–7]. Chick embryo chorioallantoic membrane (CAM) model is a prevalent animal model and readily available for tissue engineering [8,9]. Surface displacement measurement has been done for heterogenous CAM samples using 2D and 3D photoacoustic imaging [10]. Multiple development stages of chick embryos have been monitored using PA imaging [11]. Sonography has been applied to a CAM model to monitor tumor and vasculature growth [12]. None of the studies provides information about the mechano-biological properties of the sample. Many dynamic processes need a better understanding to adopt different strategies for therapy and treatment.
Recently, there have been considerable advances in non-ionizing, non-invasive biomedical tissue sensing that has demonstrated promise for early diagnosis and disease progression monitoring. One such technique is photoacoustic sensing [13,14]. The photoacoustic lens has been utilized for tissue characterization using spectral response [15]. Photoacoustic spectroscopy is another powerful tool to characterize different samples [16,17]. One such technique that has been developed in our lab is Photoacoustic Spectral Response (PASR) Technique. PASR is a non-invasive, non-contact pump-probe technique that can provide mechano-biological information of the tissue sample [18–21]. Briefly, PASR utilizes a nanosecond laser pulse to irradiate the tissues. Upon excitation, the sample experiences temperature excursion and subsequently relaxes in a non-radiative manner, thereby releasing the energy in the form of sound waves. An ultrasound sensor is used to detect these sound waves. In the PASR technique, the acoustic spectrum is obtained from the time domain signal, and the frequency spectra would provide signatures of the tissue sample. These spectra are utilized to distinguish between normal and pathological tissues.
A typical PA time-domain signal consists of three significant parameters, namely, amplitude (a), width (ζ), and relaxation time (X) (Fig. 1). These parameters depict different properties of the sample. Amplitude, relaxation time, and width of the PA signal are related to optical absorption, mechano-biological properties, and size of the sample, respectively. The relaxation time is related to the width of the signal as well. Figure 1(a) shows a simulated PA signal having 0.9 µs signal width that depicts the size of the sample, and corresponding frequency spectra is shown in Fig. 1(b), where the dominant frequency peak is at 6 MHz. This allows investigating how the signal's frequency content provides density/elasticity information of the sample to understand the mechano-biological properties.
In this paper, the PASR technique is proposed to be applied to study angiogenesis in cancer. With widespread interest in cancer and its treatment, a basic understanding of tumor dynamics and growth is of very high importance [22,23]. Angiogenesis plays a vital role in primary tumor development and also facilitates tumor invasion and metastasis. Angiogenic growth factors released by tumor cells are responsible for tumor neovascularization. Tumor microvascular networks possess unique pathological features distinguishing them from healthy blood vessels [24]. These include extremely high densities of tortuous leaky primitive micro-vessels. In general, angiogenesis is a hallmark of cancer, and for its treatment, we need anti-angiogenic compounds that inhibit the growth of these leaky angiogenic blood vessels and prevent angiogenic dependent diseases [25]. The angiostatic drugs like Cisplatin, Doxorubicin, etc., interact and interfere with the angiogenic cascade and with the factors released in neovascularization and result in reducing the density of blood vessels [26]. Differentiating reduction in blood vasculature (after anti-angiogenic drug treatment) from normal blood vessels followed by continuous monitoring of blood vessel decrease during tumor vasculature treatment would be the key questions from a sensing perspective.
In this context, continuous monitoring of vasculature in response to anti-angiogenic compounds is highly desirable to measure the efficacy of angiostatic compounds. The subsequent reduction in blood vasculature results in a reduction in tumor size. Here we investigated using the PASR technique to evaluate anti-angiogenesis in the chick embryo chorioallantoic membrane (CAM) model. In a developing chick embryo, membraneous allantois appears as early as 3.5 days of incubation of the fertilized egg. This vital membrane subsequently fuses with another membrane (mesodermal layer of chorion) to form a double layer of chorioallantoic membrane or CAM. The CAM is very rich in the vasculature and exhibits rapid growth of vessels, which mimic a tumor environment. This provides a quick, easy, and affordable model system to monitor large-scale screening of pharmacological compounds. Additionally, this CAM could provide a scaffold for the implantation of exogenous tumor cells (xenograft). In this study, we used a non-invasive PASR technique to monitor the drug response to mechano-biological properties of blood vessels in the mid chick embryo stage.
Mechano-biological properties of blood vasculature often bear the signature of its physiological conditions. These properties are rooted in its cellular composition and molecular orientation of the basic structural fibrils such as actin, and myosin, etc. [27]. Therapeutic interventions aiming to alter mechanical properties often results in favorable disease outcomes. For example, calcium channel blockers and angiotensin targeting molecules, etc., are clinical research studies aimed at targeting vascular alterations. Contextually, these changes in mechanical properties could be monitored non-invasively by photoacoustic sensing. Thus, the CAM model combined with PA-sensing is projected as a robust model system for drug screening.
Conventionally, the effect of the anti-angiogenic drug is verified through histological images or through multiple photographs taken through a conventional imaging technique (camera or a microscopic imaging system depending on the model). Literature shows that image processing needs to be applied in order to define the change in thickness of the blood vessels as the change in the thickness of blood vessels in a span of one day is not obvious (10th day to 11th day) [28]. Hence a real-time analysis of the change in the vasculature/density would not be possible using conventional stereo-microscopic images. PASR would be a potential tool that can provide this information with higher sensitivity. This paper proposes to continuously monitor the influence of the drug through the Photoacoustic Spectral Response technique (PASR).
The main objectives of this work are:
- ■ Exploring PASR as a potential tool to continuously monitor the effect of drug on a CAM assay.
- ■ Providing mechano-biological information of the CAM assay through acoustic spectra of the photoacoustic technique. This can give potential cues in better understanding of the assay compared to conventional microscopic imaging.
2. Materials and methods
2.1 Photoacoustic spectral response (PASR) system configuration
PASR technique consists of two parts: the excitation and the other is the acquisition. In the first part, the laser excitation was performed using a nanosecond Nd: YAG pulsed laser (10 ns, 0.65 mJ, 532 nm wavelength, 10Hz repetition rate and irradiation spot size 1 mm). The second is the acquisition of ultrasonic signals through ultrasonic transducers. The excitation laser light was guided onto the sample through optical components such as lenses, beam splitter, and diverging mirrors. A small percentage of the pulsed light has been diverted to an energy meter to measure the energy. This would determine the safety level of the technique. Laser energy used in the experiments was 0.65 mJ, and it is well below the safety limit according to American National Standards Institute recommendations [29].
In order to trigger the acquisition, a photodetector acquires a small portion of the Nd: YAG laser pulse. The pulse from the photodetector would be connected to an oscilloscope for triggering. The second part of the development is the acquisition. Ultrasound waves were acquired using an ultrasound transducer (3.5 MHz center frequency with 7 MHz bandwidth) that has been placed at the same height as the sample. In order to avoid any impedance mismatch, the sample container is filled with water. The time-domain signal from the ultrasound sensor (US) was captured in an oscilloscope after time-averaging for 128 times and then transferred to a computer (Fig. 2). To obtain the acoustic frequency spectra of the time domain PA signal, Fourier Transform was applied to the normalized time-domain PA signal. The frequency corresponding to the maximum magnitude is called the dominant frequency that has been utilized to understand the mechanobiological properties of a sample. In addition to the dominant frequency, spectral energy can also be used as a parameter to differentiate the sample. The spectral energy was calculated through the square of the magnitude of the (complex) Fourier transform of a PA signal.
Fig. 2. PASR experimental setup: BS = Beam splitter, PD = photodetector, and DSO = Digital storage oscilloscope
2.2 Sample preparation
2.2.1 Chick-embryo chorioallantoic membrane (CAM) model
Pathogen-free fertilized chicken eggs at various days of embryo development (EDD) were purchased from a local hatchery and disinfected using 75% spirit [30–32]. The eggs were then candled in the darkroom and incubated at 37°C and 65% relative humidity for embryogenesis [30]. On the next day, i.e., at EDD-10, eggs were candled to inspect the embryo viability and to trace the vascular area. The region was marked as the main operating window (2 × 2 cm), and a small 0.5 × 0.5 cm square opening is also marked on the eggshell at a 1cm distance from the main operating window. The operating window was cut carefully using a rotary tool to remove only the outer shell without penetrating the inner membrane. Then the CAM membrane was dropped by drilling one hole in the air sac and another hole in a square opening mark. After making a hole in the square opening, 30 µL of Phosphate buffer saline (PBS) is placed on the inner shell without removing it, and then one fine hole was made in that region [33].
2.2.2 Anti-cancer drug preparation and its administration
Anti-cancer drug (Cisplatin) and sterile normal saline (0.9% NaCl and pH 7.4) were purchased from a retail pharmacy. 15µM Cisplatin was used with 20 µl saline solution for administration on CAM membrane [34]. Topical administration of the drug on the CAM membrane was performed on EDD-10. Filter disks with the required amount of drug were loaded to the defined area of CAM. One batch of 10 eggs was labeled as control group, and another ten were kept as experimental group. Filter disks were gently grafted on the CAM membrane using blunt forceps. 20µl of final drug solution (in experimental batch) and the same volume of PBS (in control batch) was pipetted onto the filter disk slowly and allowed to soak in the disk. Following drug placement, eggs were sealed with parafilm and placed back into the incubator horizontally for further observation at 24h,48h, and 72h (Fig. 3).
2.2.3 Preparation steps for egg inoculation
- 1. Before starting the inoculation, the air sac was identified and marked using a pencil. Arrows indicate an air sac on Fig. 3(b).
- 2. The square opening mark on the eggshell was drilled, and the eggshell is removed using blunt forceps, and an arrow on Figs. 3(b) and (c) indicates the intact outer shell.
- 3. Using a drill, a small hole was made in the air sac region (cross marked area) to allow airflow into the egg (arrowhead). 30µL of PBS was placed onto the square opening mark on the eggshell.
- 4. The Place where PBS was previously dispensed, a Fine syringe needle was used to perforate the outer eggshell membrane.
- 5. Using a Rubber dropper, bulb pressure is applied to the perforated air sac region by attaching it to the drilled air sac region. The pressure released by the dropper generated an artificial air sac (white arrows) that should extend to the operating window.
- 6. The edges of the operating window were drilled. Eggshell was removed with blunt forceps.
- 7. Carefully remove the inner eggshell membrane with blunt forceps, not to introduce particles on the CAM (observed at 1 cm below the surface).
2.2.4 Cell culture and drug toxicity assay
Bab hamster kidney (BHK-21) cells (ATCC: City, Virginia, USA) were procured and were grown in a culture Dulbecco’s modified Eagle medium (DMEM) mixed with 5% heat-inactivated fetal bovine serum and Pen-Strep solution (100 units mL-1 penicillin and 100 units mL-1 streptomycin). The FBS, media, Pen-strep solution, and PBS solution were purchased from Invitrogen (Gibco) and Hi-Media, India, respectively. Cells were cultured with 5% CO2 in a humidified incubator (New Brunswick-Galaxy 48R) and at 37 °C. Cytotoxicity of cisplatin drug at various concentrations was evaluated using MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide) assay. About 6,000 cells were initially seeded onto a 96-well plate and were incubated at 37°C with 5% CO2 in the incubator. After proper cell attachment, the respective wells were treated with cisplatin drug at varying concentrations ranging from 3 to 66 µM (3.3 µM, 6.6 µM, 15 µM, 19 µM, 33 µM, 49 µM, 66 µM) were performed in a triplicate set for each concentration. The cells were treated with 10% SDS in 0.1N HCL to ensure killing was used as a positive control. After the treatment of 24h, the media was discarded, and wells were supplemented with 0.5 mg mL−1. MTT solution prepared in fresh medium(100µL) followed by incubation for 4 hr in the dark. After 4hr of treatment, the media containing MTT solution was replaced, followed by DMSO (200 µl), a solubilizing buffer, and dissolve the MTT dye formazan. Further, to dissolve the formazan crystal completely, the plate was kept on Rocker for 15 to 20 minutes, and then the absorbance was recorded at 590 nm in a microplate reader (Synergy H1BioTek multi-mode). Cell viability was counted by the equation of % of cells viability = Mean OD Treatment /Mean OD control *100.
Prior to experimentation, the eggs were observed through the operating window by removing the parafilm. Changes were observed in blood vessel diameter in experimental eggs. The thickness of blood vessels was found reduced, but there were no changes in control eggs as no significant blood vessel constriction was observed in them. To confirm observed changes in blood vessels, we detected the eggs using the photoacoustic technique, and we observed differences in peaks of the graph in experimental/inoculated and control eggs. Additionally, to rule out the adverse effects of the drug and advocate the drug tolerance dose, we performed cytotoxicity assays to find the safety limit. Mammalian cells (BHK-21) were treated with various concentrations (3 to 66 µM) of the drug. As an indicator of cytotoxicity, MTT assays were performed, and the absorbance of cells was taken at 590 nm, and high cell viability was measured. The results showed no significant adverse effect of the drug at 15 µM concentration, indicating the non-toxic nature of the Cisplatin drug used on the egg (Fig. 4). Hence, for subsequent experiments, a 15 µM concentration drug was used.
Fig. 4. Cell viability assessment by MTT assay – BHK-21 cells were seeded in 96- well plate and were treated with varying concentrations (µM) of Cisplatin (drug) for 24 hr. Positive control shows treatment of cells with 10% SDS in 0.1N HCL. Following treatment with MTT reagent and solubilizing buffer (DMSO), further absorbance was measured at 590 nm in a microplate reader. % Cell viability is represented by Bar chart, and SEM (Standard error of the mean) is indicated by Error bars. The Bar chart is depicting the mean ± SEM of 3 independent experiments(n=3).
2.3 Evaluation of blood vasculature change of CAM assay using microscopic image processing
Microscopic images of an egg sample on different days have been taken to monitor the blood vasculature continuously. After converting the microscopic image of the blood vessels to grayscale, several image processing methods have been applied to the image. Image contrast optimization, thresholding, etc., were applied in order to separate the blood vessels from the rest of the image. Image pixels belonging to the blood vessels were assigned the value of ‘1’ and ‘0’ to the rest. The detailed description is presented in the flowchart of Fig. 5.
Fig. 5. Flowchart describing the efficient image processing method applied to extract vessel area density and vessel length density
To characterize the obtained binary image, the following global parameters are determined:
- a. Vessel’s area density is defined as the number of pixels that belong to the vascular network on the image divided by the total area of the image,
- b. Vessel length density is defined as the number of pixels that belong to the skeleton of the vascular network on the image divided by the total area of the image.
3. Results and discussion
3.1 Photoacoustic spectral response (PASR) on the egg sample
In order to understand the sensitivity of the PASR technique on CAM experiments, a preliminary investigation of PASR was performed on the 5th day and 10th day control egg. The 10th day egg would have vasculature growth which results in an increase in density. The PASR results, as shown in Figs. 6(a) & 6(b), clearly show the frequency shift in these two egg samples. Figures 6(c) & 6(d) are the real-time images of the 5th and 10th day-control egg samples. It is clear that there is a considerable change in density as well as vasculature development of the chicken embryo.
Fig. 6. Frequency response of the time-domain PA signal of 5th and 10th day egg sample respectively: a &b, pictures of vasculature development on different days respectively: c &d
The dominant frequency of the 5th-day chicken embryo is 0.11MHz, whereas the 10th-day chicken embryo is 0.6 MHz, as observed from Figs. 6(a) & 6(b). The shifting of the frequency peak depicts an increase in the chicken embryo sample’s vascular development (A detailed explanation is provided in the discussion section). With this parameter’s help, we could characterize the fast-growing blood vessels mimicking tumor vasculature.
3.1.1 Effect of Cisplatin drug on inoculated eggs: monitoring using PASR
Platinum-based chemotherapy drugs are among the most powerful and widely used against cancer. Cisplatin, the most common platinum chemotherapy drug, was first approved in the US in 1978 [34]. 20 ul of cisplatin drug was placed on the filter disk in the 10th day experimental egg (inoculated egg) (Fig. 3(e)). The drug was inoculated on a 10th day egg. It is known that this drug is used for anti-angiogenesis. It would decrease the density of the blood vessels. Hence PASR experiments were carried out on the 10th day and 12th day eggs that are inoculated with the drug. Figure 7 shows the clear frequency shift between the two-time points. This proves that PASR can provide information well before visible distinction. This is important for continuous monitoring of samples. The absence and shifting of the dominant frequency peak of PA response justify the overgrown blood vessel reduction due to the cisplatin drug in the inoculated egg sample.
3.1.2 Effect of Cisplatin drug on time-domain PA signal
In this investigation, PA signals have been acquired for continuous monitoring of the 10th, 11th, and 12th day control and inoculated egg samples without applying any normalization (Fig. 8). PA signal has shown a gradual change in the time domain signal amplitude because of the optical absorption coefficient changes between control and cisplatin drug inoculated samples in Table 1. It gives information about the reduction of the blood vessel growth that can be monitored by a simple microscope but doesn’t give any information about the machno-biological properties of the sample.
Table 1. Average Peak to Peak PA signal amplitude of multiple control and inoculated egg samples with standard deviation (SD).
3.1.3 Normalization of time-domain PA signal to eliminate variation in absorption
The time-domain PA signal has been normalized before determining the frequency spectra. Normalization would remove the effect of the change in absorption/amplitude of the PA signal in the frequency spectra. A simulation Fig. 9 shows how normalization removes the variability due to optical absorption in the results (Fig. 9). Simulation targets of absorption of 1X, 1.5X, and 2 X were studied, and their time-domain response has been obtained. Further, the acoustic spectra were obtained that clearly show a difference. However, after normalization of the time domain signal, as shown in Fig. 9(b), the frequency spectra are the same for all the three targets.
3.1.4 Continuous drug monitoring using PASR
To understand the sample’s mechanobiological properties, we normalized the time domain signal amplitude and later performed post-processing to get PASR spectral response. This normalization has excluded the effect of optical absorption on the egg sample’s spectral response. Further, PA response has been investigated for multiple egg samples to assess the change in the mechano-biological properties such as density, elasticity, speed of sound, etc., of the inoculated sample and differentiate the inoculated from the control one.
After continuous monitoring of the effect of the cisplatin drug on the inoculated egg sample, PA response has been acquired for ten egg samples each on 10th, 11th, and 12th control and inoculated egg samples to assess the change in the mechanobiological properties. Figure 10. (a) shows that the dominant frequency of the PA spectral response for 10th day control and inoculated eggs lies between 0.45–0.52 MHz, whereas for 11th day control egg lies between 0.61 ± 0.152 MHz and inoculated egg lies between 0.35 ± 0.156 MHz (Fig. 10. (b)). On the 12th day control egg, the spectral frequency of the PA response lies between 0.67 ± 0.001 MHz, whereas, for an inoculated egg, the range is distinctly different, i.e., 0.16 ± 0.004 MHz (Fig. 10. (c)). Besides dominant frequency, extra spectral peak introduced in the 11th and 12th day results because of the heterodyne medium of the body of the egg sample. total spectral energy was also obtained. A change in these spectral parameters can also be used to differentiate the sample with identical absorption. Table 2 shows the spectral parameters of control and inoculated egg samples. From Table 2, it is clear that, in addition to the dominant frequency, the spectral energy can also be used as a parameter to differentiate the sample. The spectral energy was calculated through the square of the magnitude of the (complex) Fourier transform of a PA signal. There is a (4.65 ± 0.04) increase in the spectral energy between 10th day control and inoculated eggs. On 12th day difference increases by (8.77 ± 0.07) in the spectral energy of 12th day control and inoculated egg. Statistical results in Figs. 11(a), (b), and (c) give the comparative analysis and indicate the change in the properties of the egg samples. Section 3.4 shows the relation between spectral shift and the properties of the sample. Table 2 shows a significant change in spectral parameters to differentiate the sample, and this would pave the way to take this technique to in vivo diagnosis
Fig. 10. Acoustic spectral response of time-domain PA signal of a) 10th day, b)11th day, and c) 12th day egg sample
Table 2. Spectral parameters of control and inoculated egg sample with standard deviation.
3.2 Correlation of PASR results with microscopic images (image processing)
Microscopic images of the egg samples on 5th day, 10th day, and 11th day have been taken to monitor the change in the blood vasculature. Image processing is performed to differentiate the blood vessels in the image from the background non-vasculature area. Blood vasculature area density and vasculature length density were calculated from the binary images, and the results are shown in Table 3. It is clear that microscopic images can show difference only when substantial growth occurs (5th day to 10th day) (Fig. 12). However, one-day growth cannot be guaranteed even after image processing. On the contrary, PA could differentiate the egg samples with a day. This shows the accuracy and efficacy of the PASR technique for such bio-assays.
Fig. 12. Microscopic images of the (a) 5th (b) 10th and (c) 11th day control sample and (d), (e), & (f) are binary processed images. (g), (h) and (i) are skeleton-transformed images for blood vessel length measurement.
Table 3. Vasculature changes observed at microscopic images of the control egg sample with standard deviation.
Table 3 shows a significant density change between the 5th and 10th day egg samples, but there is no significant change between the 10th and 11th day egg samples. This shows that image analysis can provide significant change after a long duration but is not suitable for continuous monitoring.
3.3 PASR simulation studies: correlation of mechno-biological properties
In order to understand how sound speed and density affect the PA spectrum, a simulation study was performed in MATLAB using a k-Wave toolbox to investigate elasticity in biological tissues [35,36]. Three circular disc samples with identical size (1 mm diameter & 1000 kg/m3 medium density) and absorption were simulated with a change in sound speed from 1040 m/s, 1540 m/s to 1940 m/s Table 4. Another investigation has been done with three different sizes of the disc sample (2 mm, 4 mm, and 10 mm disc diameter and density 1000 kg/m3, medium sound speed 1040 m/s) to understand the spectral broadening and multiple reflections of the acoustic wave respect to the size of the sample. Figure 13(a) shows that as the size of the sample increases, there is a chance of narrowing the spectra, and multiple peaks may get introduced that is responsible for the pressure distribution. The egg samples were kept inside the 40 mm diameter plastic cup, this justifies the multiple peaks in the egg sample results in the Fig. 10 due to a large size of the cup and heterodyne medium of the blood vessel of the egg samples as, and Fig. 13(b) shows the dominant frequency shift in the spectrum from 1.1MHz to 2.0MHz.
Fig. 13. a). Spectral broadening with different sizes of sample b). Spectral response with different medium sound speed
Table 4. Properties of the propagation medium, initial pressure distribution, and given sensors
To understand the density effect on PASR, two identical size agarose samples were prepared with 0.025 g/ml and 0.075 g/ml of concentration of the agarose and kept black ink with constant concentration (4ml)15. Figure 14(a) shows the experimental PA results with the change in the density, and Fig. 14(b) shows a shift in the dominant frequency by 0.6–0.8MHz. This study clearly says that the spectral shift can be correlated with the tissue/sample density when the size of the same is identical. Therefore, the photoacoustic spectral response can be a great tool to investigate a tissue sample’s mechanobiological property.
Fig. 14. a). PA signal in the time domain with different density of mixture b). Spectral response of the time domain PA signal
It is prudent to discuss the advantage of the proposed PASR technique over the conventional PA technique. PA images obtained from conventional technique focuses on the amplitude of the time domain PA signal and the time of flight of the PA signal. Based on these two parameters, the tomographical image can be obtained through any of the time-reversal algorithms. The issue is that for samples that have analogous optical absorption, conventional PA imaging would fail. One such application explored using conventional PA imaging is to differentiate acute from chronic blood clots, which is a very useful for thrombosis applications. While PA imaging can successfully differentiate these two clots, they could not differentiate acute clots from blood. The reason is these two have almost identical absorption leaving little contrast in the PA images [37]. It is known that time-domain PA signal contains all the mechano-biological information. But this information is present in the rise and fall time of the PA signal, which is not used during PA tomographic imaging. Further, analyzing the Fourier spectra would be an easier way to find the differences, and there are many statistical analyses that can be performed. Hence, acquiring spectral information is proposed that can identify samples not only on optical absorption but also on the mechano-biological properties of the sample. This is a very important improvement in the PA technique to be used as a clinical diagnostic tool.
In terms of application, there are several applications that require quick screening before a detailed diagnostic procedure. For example, different types of thyroid cancers can be diagnosed through sonography techniques (ultrasound). However, Fine needle aspiration cytology (FNAC), which is very close to biopsy, would provide accurate results of the cancer diagnosis. FNAC is a single-point detection. Similar to FNAC, photoacoustic spectral response (PASR) is proposed to be a point detection technique. Unlike FNAC, the PASR technique can be acquired in a very quick time (few milliseconds) and hence can be performed at several locations of the sample. One such study has been performed by our group [37] where one particular variation of thyroid cancer can be distinguished through PASR, which was not possible using conventional diagnostic techniques.
4. Conclusion
We have demonstrated and analyzed the performance of the PASR system implemented on CAM Assay. The proposed PA experiment could detect a change in mechanobiological properties within a day span, which is not possible with any existing modalities even after the image processing. The photoacoustic investigation was performed for a fertilized CAM assay on the 10th, 11th, and 12th day with cisplatin drug embryonated egg samples. The proposed system could find applications in the fields where non-invasive or minimally invasive sensing is specifically required. The acquired PA signal’s spectral response shows a gradual frequency shift between control and inoculated eggs. The absence of the frequency peaks during anti-drug treatment of the inoculated experimental egg sample and correlation of PASR results with sample properties justifies the overgrown blood vessel reduction. Therefore, PASR can be explored more for continuous real-time monitoring of CAM assays and thus would be a potential diagnostic technique for several biomedical applications.
Funding
Science and Engineering Research Board (EMR/2017/004411).
Acknowledgments
The authors thank Mr. Sanchit Neema, Department of Biosciences and Biomedical Engineering, for technical discussions and laboratory support.
Disclosures
The authors declare that they have no conflict of interest.
Data availability
The datasets generated and analyzed in this paper are available from the corresponding author upon reasonable request.
Supplemental document
See Supplement 1 for supporting content.
References
1. A. P. Dhawan, “A review on biomedical image processing and future trends,” Comput. Methods Programs Biomed. 31(3-4), 141–183 (1990). [CrossRef]
2. J. Wallyn and N. Anton, “Biomedical Imaging: Principles, Technologies, Clinical Aspects, Contrast Agents, Limitations and Future Trends in Nanomedicines,” Pharm Res 36(6), 78 (2019). [CrossRef]
3. Z. Sun, K. H. Ng, and N. Ramli, “Biomedical imaging research: A fast-emerging area for interdisciplinary collaboration,” Biomed. Imaging Interv. J. 7, 1–4 (2011). [CrossRef]
4. N. Almadani, E. F. Thompson, B. Tessier-Cloutier, J. Pors, and L. Hoang, “An update of molecular pathology and shifting systems of classification in tumours of the female genital tract,” Diagnostic Histopathol. 26(6), 278–288 (2020). [CrossRef]
5. H. Wu, Z. Ji, and M. Li, “Non-invasive continuous blood-pressure monitoring models based on photoplethysmography and electrocardiography,” Sensors 19(24), 5543 (2019). [CrossRef]
6. K. S. Litvinova, I. E. Rafailov, A. V. Dunaev, S. G. Sokolovski, and E. U. Rafailov, “Non-invasive biomedical research and diagnostics enabled by innovative compact lasers,” Prog. Quantum Electron. 56, 1–14 (2017). [CrossRef]
7. M. Falk, C. Psotta, S. Cirovic, and S. Shleev, “Non-invasive electrochemical biosensors operating in human physiological fluids,” Sensors 20(21), 6352 (2020). [CrossRef]
8. I. Moreno-jiménez, G. Hulsart-billstrom, S. A. Lanham, and A. A. Janeczek, “The chorioallantoic membrane (CAM) assay for the study of human bone regeneration : a refinement animal model for tissue engineering,” Sci Rep 6(1), 32168 (2016). [CrossRef]
9. P. Nowak-sliwinska, T. Segura, M. L. Iruela-arispe, and L. Angeles, “HHS Public Access,” 17, 779–804 (2015).
10. S. A. Carp and V. Venugopalan, “Optoacoustic imaging based on the interferometric measurement of surface displacement,” J. Biomed. Opt. 12(6), 064001 (2007). [CrossRef]
11. M. Liu, B. Maurer, B. Hermann, B. Zabihian, M. G. Sandrian, A. Unterhuber, B. Baumann, E. Z. Zhang, P. C. Beard, W. J. Weninger, and W. Drexler, “Dual modality optical coherence and whole-body photoacoustic tomography imaging of chick embryos in multiple development stages,” Biomed. Opt. Express 5(9), 3150 (2014). [CrossRef]
12. J. Eckrich, P. Kugler, C. R. Buhr, B. P. Ernst, S. Mendler, J. Baumgart, J. Brieger, and N. Wiesmann, “Monitoring of tumor growth and vascularization with repetitive ultrasonography in the chicken chorioallantoic-membrane-assay,” Sci. Rep. 10(1), 18585 (2020). [CrossRef]
13. A. Wiacek and M. A. Lediju Bell, “Photoacoustic-guided surgery from head to toe [Invited],” Biomed. Opt. Express 12(4), 2079 (2021). [CrossRef]
14. J. Laufer, A. Jathoul, M. Pule, and P. Beard, “In vitro characterization of genetically expressed absorbing proteins using photoacoustic spectroscopy,” Biomed. Opt. Express 4(11), 2477 (2013). [CrossRef]
15. J. Heo, D. Biswas, K. K. Park, D. Son, H. J. Park, and H. W. Baac, “Laser-generated focused ultrasound transducer using a perforated photoacoustic lens for tissue characterization,” Biomed. Opt. Express 12(3), 1375 (2021). [CrossRef]
16. T. Schmid, “Photoacoustic spectroscopy for process analysis,” Anal. Bioanal. Chem. 384(5), 1071–1086 (2006). [CrossRef]
17. J. Li, W. Chen, and B. Yu, “Recent progress on infrared photoacoustic spectroscopy techniques,” Appl. Spectrosc. Rev. 46(6), 440–471 (2011). [CrossRef]
18. A. Gorey, S. Vasudevan, M. S. Ansari, P. Bhagat, S. Phatak, N. Sharma, and G. C. K. Chen, “Development of a compact laser-diode based frequency domain photoacoustic sensing system: Application of human breast cancer diagnosis,” Rev. Sci. Instrum. 90(11), 114101 (2019). [CrossRef]
19. D. Biswas, G. C. K. Chen, H. W. Baac, and S. Vasudevan, “Photoacoustic spectral sensing technique for diagnosis of biological tissue coagulation: In-vitro study,” Diagnostics 10(3), 133 (2020). [CrossRef]
20. D. Biswas, A. Kumari, G. C. K. Chen, S. Vasudevan, S. Gupta, S. Shukla, and U. K. Garg, “Quantitative Differentiation of Pneumonia from Normal Lungs: Diagnostic Assessment Using Photoacoustic Spectral Response,” Appl. Spectrosc. 71(11), 2532–2537 (2017). [CrossRef]
21. D. Biswas, S. Vasudevan, G. C. K. Chen, and N. Sharma, “Quantitative photoacoustic characterization of blood clot in blood: A mechanobiological assessment through spectral information,” Rev. Sci. Instrum. 88(2), 024301 (2017). [CrossRef]
22. M. Rajabi and S. A. Mousa, “The role of angiogenesis in cancer treatment,” Biomedicines 5, (2017).
23. N. Nishida, H. Yano, T. Nishida, T. Kamura, and M. Kojiro, “Angiogenesis in cancer,” Vasc. Health Risk Manag. 2(3), 213–219 (2006). [CrossRef]
24. A. R. Pries, A. J. M. Cornelissen, A. A. Sloot, M. Hinkeldey, M. R. Dreher, M. Höpfner, M. W. Dewhirst, and T. W. Secomb, “Structural adaptation and heterogeneity of normal and tumor microvascular networks,” PLoS Comput. Biol. 5(5), e1000394 (2009). [CrossRef]
25. D. Ribatti, B. Nico, A. Vacca, L. Roncali, P. H. Burri, and V. Djonov, “Chorioallantoic membrane capillary bed: A useful target for studying angiogenesis and anti-angiogenesis in vivo,” Anat. Rec. 264(4), 317–324 (2001). [CrossRef]
26. A. Vargas, M. Zeisser-Labouèbe, N. Lange, R. Gurny, and F. Delie, “The chick embryo and its chorioallantoic membrane (CAM) for the in vivo evaluation of drug delivery systems,” Adv. Drug Deliv. Rev. 59(11), 1162–1176 (2007). [CrossRef]
27. R. D. and K. C. Holmes, “Actin Structure and function,” 169–186 (2011).
28. S. Blacher, L. Devy, R. Hlushchuk, E. Larger, N. Lamandé, P. Burri, P. Corvol, V. Djonov, J. M. Foidart, and A. Noël, “Quantification of angiogenesis in the chicken chorioallantoic membrane (CAM),” Image Anal. Stereol. 24(3), 169–180 (2011). [CrossRef]
29. A. L. Augustoni, “Updated Laser Safety & Hazard Analysis for the ARES Laser System Based on the 2007 ANSI Z136 . 1 Standard,” (2007).
30. D. Ribatti, “The chick embryo chorioallantoic membrane as an in vivo assay to study antiangiogenesis,” Pharmaceuticals 3(3), 482–513 (2010). [CrossRef]
31. P. Kunz, A. Schenker, H. Sähr, B. Lehner, and J. Fellenberg, “Optimization of the chicken chorioallantoic membrane assay as reliable in vivo model for the analysis of osteosarcoma,” PLoS One 14(4), e0215312 (2019). [CrossRef]
32. A. Zijlstra, D. Mikolon, and D. G. Stupack, “Angiogenesis Assays in the Chick,” Angiogenes. Assays A Crit. Apprais. Curr. Tech. 294, 183–201 (2007).
33. M. Li, R. R. Pathak, E. Lopez-Rivera, S. L. Friedman, J. A. Aguirre-Ghiso, and A. G. Sikora, “The in ovo chick chorioallantoic membrane (CAM) assay as an efficient xenograft model of hepatocellular carcinoma,” JoVE 2015(104), 1–6 (2015). [CrossRef]
34. C. S. Kue, K. Y. Tan, M. L. Lam, and H. B. Lee, “Chick embryo chorioallantoic membrane (CAM): An alternative predictive model in acute toxicological studies for anti-cancer drugs,” Exp. Anim. 64(2), 129–138 (2015). [CrossRef]
35. B. E. Treeby and B. T. Cox, “k-Wave: MATLAB toolbox for the simulation and reconstruction of photoacoustic wave fields,” J. Biomed. Opt. 15(2), 021314 (2010). [CrossRef]
36. B. E. Treeby, J. Jaros, D. Rohrbach, and B. T. Cox, “Modelling elastic wave propagation using the k-Wave MATLAB Toolbox,” IEEE Int. Ultrason. Symp. IUS, 146–149 (2014).
37. A. Gorey, P. M. Jacob, D. T. Abraham, R. John, M. T. Manipadam, M. S. Ansari, G. C. K. Chen, and S. Vasudevan, “Differentiation of malignant from benign thyroid nodules using photoacoustic spectral response: A preclinical study,” Biomed. Phys. Eng. Express 5(3), 035017 (2019). [CrossRef]
