Here, we demonstrate how agarose-based tissue-mimicking optical phantoms are made and how their optical properties are determined using a conventional optical system with an integrating sphere.
This protocol describes how to make agarose-based tissue-mimicking phantoms and demonstrates how to determine their optical properties using a conventional optical system with an integrating sphere. Measuring systems for the acquisition of the diffuse reflectance and total transmittance spectra are constructed with a broadband white light source, a light guide, an achromatic lens, an integrating sphere, a sample holder, an optical fiber probe, and a multi-channel spectrometer. An acrylic mold consisting of two rectangular acrylic pieces and a U-shaped acrylic piece is constructed to create an epidermal phantom and a dermal phantom with whole blood. The application of a sodium dithionite (Na2S2O4) solution to the dermal phantom enables the researcher to deoxygenate hemoglobin in red blood cells distributed in the dermal phantom. The inverse Monte Carlo simulation with the diffuse reflectance and total transmittance spectra measured by a spectrometer with an integrating sphere is performed to determine the absorption coefficient spectrum µa(λ) and the reduced scattering coefficient spectrum µs'(λ) of each layer phantom. A two-layered phantom mimicking the diffuse reflectance of human skin tissue is also demonstrated by piling up the epidermal phantom on the dermal phantom.
Optical phantoms are objects mimicking the optical properties of biological tissues and have been widely used in the biomedical optics field. They are designed so that the optical properties, such as light scattering and absorption coefficients, match with those of living human and animal tissues. Optical phantoms are generally used for the following purposes: simulating the light transport in biological tissues, calibrating a newly developed optical system design, evaluating the quality and performance of existing systems, comparing the performance between systems, and validating the ability of the optical methods to quantify the optical properties1,2,3,4,5. Therefore, easy-to-get substances, a simple fabrication process, a high reproducibility, and an optical stability are required for making optical phantoms.
Various types of optical phantoms with different base materials such as aqueous suspension6, gelatin gel7, agarose gel8,9,10, polyacrylamide gel11, resin12,13,14,15,16, and room-temperature-vulcanizing silicone17 have been reported in previous literature. It has been reported that gelatin- and alginate-based gels are useful for optical phantoms with heterogeneous structures18. Alginate phantoms have a suitable mechanical and thermal stability for evaluating photothermal effects such as laser ablation studies and laser-based hyperthermia studies18. Agarose gels have the ability to fabricate heterogeneous structures, and their mechanical and physical properties are stable for a long time18. High-purity agarose gels have a very low turbidity and a weak optical absorption. Therefore, optical properties of agarose-based phantoms could easily be designed with the appropriate light scattering and absorbing agents. Recently, styrene-ethylene-butylene-styrene (SEBS) block copolymers19 and polyvinyl chloride (PVC) gels20 have been reported as interesting phantom materials for optical and photoacoustic techniques.
Polymer microspheres7,12,21,22, titanium oxide powder1, and lipid emulsions23,24,25,26 such as milk and lipid emulsion are used as light scattering agents, whereas black ink27,28 and molecular dyes29,30 are used as light absorbers. Diffuse reflectance spectra of most living organs are dominated by the absorption of oxygenated and deoxygenated hemoglobin in red blood cells. Therefore, hemoglobin solutions31,32 and whole blood8,9,10,33,36 are often used as light absorbers in the phantoms for a diffuse reflectance spectroscopy and multispectral imaging.
The method described in this article is used to create an optical phantom mimicking the light transport in biological tissues and to characterize its optical properties. As an example, a two-layered optical phantom mimicking optical properties of human skin tissue is demonstrated. The advantages of this method over alternative techniques are the ability to represent diffuse reflectance spectra of living biological tissues in the visible to near-infrared wavelength region, as well as the simplicity to make it, using easily available materials and conventional optical instruments. Therefore, the optical phantoms made by this method will be useful for the development of optical methods based on diffuse reflectance spectroscopy and multispectral imaging.
1. Construction of a Conventional Diffuse Reflectance and Total Transmittance Spectroscopic System
Note: Construct the measuring systems for the diffuse reflectance and total transmittance spectra using a broadband white light source, a light guide, an achromatic lens, an integrating sphere, a sample holder, an optical fiber, and a multi-channel spectrometer. The role of the light trap is to remove the specular reflection component from the reflectance spectrum. The sample holder of the integrating sphere consists of a mounting plate and a dovetail and spring-loaded clamp assembly that holds the sample against the port. The dovetail and spring-loaded clamp assembly are removed from the sample holder and a hand-made cubic pedestal of polystyrene foam is attached to the mounting plate instead. The layouts of the optical components, shown in Figure 1a and 1b, can be referred to for the construction procedure for the diffuse reflectance measurements and the total transmittance measurements, respectively.
2. Preparation of an Acrylic Mold
Note: An acrylic mold that consists of two rectangular acrylic pieces and a U-shaped acrylic piece is constructed to create a monolayer gel phantom. Figure 2 can be referred to for this construction procedure.
3. Preparation of Base Material
4. Preparation of Skin-mimicking Optical Phantoms
Note: A coffee solution is used to mimic the absorption spectrum of melanin. The coffee solution contains a brown pigment called melanoidin. The absorption spectrum of melanoidin has been reported to be similar to that of melanin10.
5. Acquisition of the Diffuse Reflectance Spectra
6. Acquisition of the Total Transmittance Spectrum
7. Estimating the Absorption and Light-scattering Properties
Note: A set of the diffuse reflectance spectrum and the total transmittance spectrum is saved to the hard drive of a personal computer and analyzed offline. An inverse Monte Carlo simulation8,38,39,40 is then performed to estimate the absorption coefficient spectrum µa(λ) and the reduced scattering coefficient spectrum µs’(λ). In this inverse Monte Carlo simulation, the estimated scattering coefficient µs, under the assumption that the anisotropy factor g is 0, is regarded as the reduced scattering coefficient µs’. Both the reflectance and the transmittance data are used for a single simulation run. The detailed algorithm used in this protocol has been reported in previous literature8,39. We estimated the absorption coefficient spectrum µa(λ) and the reduced scattering coefficient spectrum µs’(λ) of an epidermal layer from a set of the diffuse reflectance spectrum and the total transmittance spectrum obtained from the epidermal layer. In the same way, we estimated µa(λ) and µs’(λ) of a dermal layer from a set of the diffuse reflectance spectrum and the total transmittance spectrum obtained from the dermal layer.
Figure 3 shows the representative estimated spectra of the reduced scattering coefficient and the absorption coefficient for the epidermal phantom and dermal phantom. The results shown in Figure 3 are the averages of ten measurements of both reflectance and transmittance spectra. The reduced scattering coefficient µs' has a broad scattering spectrum, exhibiting a higher magnitude at shorter wavelengths. The spectral features correspond to the typical scattering spectra of soft tissues. The absorption coefficient µa of the epidermal phantom decays exponentially as the wavelength increases, which is similar to the absorption spectrum of melanin. The absorption coefficient spectrum of the epidermal phantom layer and that of melanin41 were fitted by an exponential function as:
The value of B for the epidermal layer was calculated to be 0.011, whereas that for melanin was estimated to be 0.009. The wavelength dependence of the absorption coefficients µa for the dermal phantom containing oxygenated blood and deoxygenated blood is dominated by the spectral characteristics of oxygenated hemoglobin and deoxygenated hemoglobin, respectively.
Figure 4 shows representative digital color photographs of the two-layered skin phantoms. Figure 4a shows a cross-sectional image of the two-layered skin phantom. Figure 4b and 4c show top views of the 3-by-3 phantom matrix containing oxygenated blood and deoxygenated blood, respectively. The rows from top to bottom have coffee solution concentrations Cc of 5%, 10%, and 20%. The columns from left to right have blood concentrations Cb of 0.2%, 0.4%, and 0.6%. The color of the phantom becomes darker as the value of Cc in the epidermal layer increases, whereas it turns pink as the value of Cb increases. The phantom with oxygenated blood has a more reddish color than that with deoxygenated blood. Those variations represent the change in skin color due to physiological conditions such as tanning and hypoxemia, respectively.
Figure 5 shows an example of representative measured diffuse reflectance spectra obtained from the two-layered skin tissue phantoms having different conditions for (Figure 5a) the concentration of coffee solution Cc, (Figure 5b) the concentration of whole blood Cb, and (Figure 5c) the oxygenated state of blood. In Figure 5a, the diffuse reflectance in a shorter wavelength region is significantly decreased in comparison with that in a longer wavelength region as the value of Cc becomes larger. This is due to the strong light absorption by the coffee solution in the shorter wavelength region (see Figure 3b). Figure 5b shows the remarkable change in diffuse reflectance in the middle wavelength region with the value of Cb, which represents the strong light absorption by hemoglobin in the wavelength range from 500 to 600 nm. The difference in spectral feature between oxygenated hemoglobin and deoxygenated hemoglobin and isosbestic points of hemoglobin are clearly observed in the diffuse reflectance spectra shown in Figure 5c.
Figure 1: Schematic diagram of the experimental apparatus. These panels show the set-up for measuring (a) diffuse reflectance spectra and (b) total transmittance spectra. Please click here to view a larger version of this figure.
Figure 2: Steps in the preparation of agarose-based optical phantoms. These panels show (a) the making of an epidermal layer phantom and (b) the making of a dermal layer phantom. Please click here to view a larger version of this figure.
Figure 3: The representative estimated optical properties of phantoms. (a) This panel shows the average reduced scattering coefficient spectrum µs'(λ) of the epidermal and dermal layers. (b) This panel shows the absorption coefficient spectra µa(λ) of the epidermal layer and dermal layers. Please click here to view a larger version of this figure.
Figure 4: The representative digital color photographs of the two-layered skin phantoms. (a) This panel shows a cross-sectional view of the two-layered skin phantom. (b) This panel shows the top view of the 3-by-3 phantom matrix containing oxygenated blood. (c) This panel shows the top view of the 3-by-3 phantom matrix containing deoxygenated blood. The rows from top to bottom have coffee solution concentrations Cc of 5%, 10%, and 20%. The columns from left to right have blood concentrations Cb of 0.2%, 0.4%, and 0.6%. Please click here to view a larger version of this figure.
Figure 5: The representative measured diffuse reflectance spectra obtained from the two-layered skin tissue phantoms. These panels show the diffuse reflectance spectra of the phantoms with different conditions of (a) the concentration of coffee solution Cc, (b) the concentration of whole oxygenated blood Cb, and (c) the oxygenated state of blood. Please click here to view a larger version of this figure.
The most critical step in this protocol is the temperature control of the base material. The temperature to maintain the base material ranged from 58 to 60 °C. If the temperature is more than 70 °C, a denaturation of both the lipid emulsion and the whole blood will occur. As a consequence, the optical properties of the phantom will deteriorate. If the temperature is less than 40 °C, the base material will be ununiformly gelled and, thus, the light scattering and absorption agents will be heterogeneously distributed in the phantom. Although the base material is kept at 60 °C, suctioning it with a syringe lowers the temperature. The temperature of the base material lowers to 50 °C when it is added to the blood solution.
The optical phantoms described in this article suffer from short useable lifetimes which are usually limited to no more than one day. The useable lifetimes might be extended by encapsulating the phantom with the base material in the sealed container or by using a preservative. The 1-mm-thick epidermal layer phantom is an order of magnitude greater than the human epidermal thickness. In this protocol with the acrylic mold, however, it was difficult to create a layer thickness less than 0.5 mm. To reduce the expected effects of this thickness on the measured diffuse reflectance spectra of the phantoms, the scattering and absorption coefficients of the epidermis phantom were regulated so that the diffuse reflectance spectrum showed the similar spectrum to that of human skin. A spin-coating method42 looks promising for making a layer thinner than 0.5 mm. The values of µa (λ) and µs' (λ) for human skin are reported in the literature43.
The uniform distribution of melanin or bilirubin in an agar phantom layer might be difficult using the protocol described here because those chromophores are not completely soluble in water. The use of melanoidin extracted from roasted coffee beans and tartrazine can be used as comparable or substitute materials for melanin and bilirubin, respectively. The inverse Monte Carlo simulation used for estimating the optical properties from the measured diffuse reflectance and the total transmittance is relatively time-consuming due to its iterative fashion. Another light transport calculation model such as the adding-doubling method44 can be used to shorten the calculation time. The reduced scattering coefficient µs' is a lumped optical property incorporating the scattering coefficient µs and the anisotropy factor g. To estimate µs and g separately, the collimated transmittance of a phantom must be measured in addition to the total transmittance and the diffuse reflectance38,40. In the present study, we did not measure the refractive index for each layer. We set the refractive index of water as published in literature45 in the input data file for the inverse Monte Carlo simulation instead since the agarose gel consists mainly of water. We assumed that there is no difference in the refractive indices between the two layers. We also used the nominal value for the refractive index of glass (e.g., n = 1.524 at λ = 546.1 nm) for the Monte Carlo simulations.
It is advantageous that this protocol, with one integrating sphere instead of two integrating spheres, is cost-effective. On the other hand, using a single integrating sphere is time-consuming since the arrangement of the integrating sphere must be changed according to whether the measurement is for a total transmittance or for a diffuse reflectance. It is advantageous that the protocol described in this article can extend to create monolayer or multilayer optical phantoms with various shapes, sizes, and inclusions by changing the design of the molds. The surfaces of the phantom layers were wetted immediately after they were taken out of their mold. Therefore, the epidermal layer and dermal layer were adhered together by stacking the second layer closely onto the first layer. It might be possible to solidify the second layer directly on the first one, rather than fabricating them separately and attaching them afterward. In that case, however, it may be difficult to accurately make a thin epidermal layer with a uniform layer thickness. We sandwiched the phantom between the glasses to prevent a drying of the phantom. We considered the optical properties and thickness of glass in the inverse Monte Carlo simulation. Therefore, there is no effect on the estimated optical properties of the phantoms. The significance of the present technique with respect to existing methods is its ability to represent the diffuse reflectance spectra of living tissues in the visible to near-infrared wavelength region. The optical phantoms made by this protocol will be available for validation of newly developed optical methods based on diffuse reflectance spectroscopy and spectrocolorimetry.
The authors have nothing to disclose.
Part of this work was supported by a Grant-in-Aid for Scientific Research (C) from the Japanese Society for the Promotion of Science (25350520, 22500401, 15K06105) and the US-ARMY ITC-PAC Research and Development Project (FA5209-15-P-0175, FA5209-16-P-0132).
150-W halogen-lamp light source | Hayashi Watch Works Co., Ltd, Tokyo, Japan | LA-150SAE | |
Light guide | Hayashi Watch Works Co., Ltd, Tokyo, Japan | LGC1-5L1000 | |
Integrating Sphere | Labsphere Incorporated, North Sutton, NH, USA | RT-060-SF | |
Port adapter | Labsphere Incorporated, North Sutton, NH, USA | PA-050-SMA-SF | |
Light trap | Labsphere Incorporated, North Sutton, NH, USA | LTRP-100-C | |
Spectralon white standard with 99% diffuse reflectance | Labsphere Incorporated, North Sutton, NH, USA | SRS-99-020 | |
Optical fiber | Ocean Optics Inc., Dunedin, Florida, USA | P400-2-VIS-NIR | |
Miniature Fiber Optic Spectrometer | Ocean Optics Inc., Dunedin, Florida, USA | USB2000 | |
Achromatic lens | Chuo Precision Industrial Co.,Ltd, Tokyo, Japan | ACL-50-75M | |
Intralipid | Fresenius Kabi AB, Uppsala, Sweden | Intralipid 10% | |
Coffee (Blendy Mocha Blend Regular Coffee) |
Ajinomoto AGF, Inc. Tokyo, Japan | Unavailable | |
Whole blood | Nippon Bio-Test Laboratories Inc. Saitama, Japan | 0103-2 | |
Agarose | Nippon Genetics Co., Ltd, Tokyo, Japan | NE-AG02 | |
Cooking heater | TOSHIBA CORPORATION Tokyo, Japan | HP-103K |