As part of the National Center for Research and Development program LIDER IX
the T-SKIN project
“THz multifunctional diffractive system to enhance imaging of skin cancer”
is being carried out at the Warsaw University of Technology, Faculty of Physics.
Duration: 1.01.2019-31.12.2021
Funding amount: 1 198 375 PLN
Project leader: Agnieszka Siemion, PhD Eng.
Report 2019:
Development of a skin tissue phantom model for tests with the use of terahertz radiation
The aim of the study was to develop a human tissue model for the terahertz wave. Firstly, the literature on this subject was analyzed. The analyzed articles referred to the measurements of human skin tissue parameters, both healthy and suffering from basal-cell carcinoma. On the basis of the results obtained for these tissues in spectroscopic studies, the values of refractive index and absorption coefficient for the terahertz wave were determined. Literature analysis made it possible to formulate theoretical assumptions regarding the values of the aforementioned coefficients for the tissue model tested with a terahertz wave at a frequency of 0.5 THz.
The next step was to determine which material is the most suitable for creating a tissue model. In that case, again, on the basis of the literature, it was decided to use TX-151 substance in the process of development of the tissue model.
During the studies, 3 main series of measurements were carried out and 45 samples consisting of a mixture of TX-151, water and other additives were made. The developed tissue models were examined with the terahertz wave using the TPS Spectra 3000 spectroscope located at the Institute of Optoelectronics at the Military University of Technology in Warsaw. We managed to distinguish several compositions which can successfully imitate human tissues in terms of the absorption coefficient and refraction index.
During the first attempts to create tissue phantoms, it turned out that handling a mixture of water and TX-151 requires appropriate preparation. The sample should be thin, stiff, but not so thick that it leads to breaking. Many different proportions of ingredients were tested before developing the mixture allowing to obtain solidifying and non-cracking samples. The large amount of powder in relation to water caused the sample to stiffen too quickly and it also resulted in a total lack of control when it comes to its thickness. However, the large amount of water in relation to the powder caused the sample to stiffen very slowly or not at all. When it comes to certain compositions of samples, there were air bubbles or lumps of undiluted powder and cracks appeared after solidifying. Table 1 shows the observations of the behavior of samples with a given ratio of water, powder and additive after the first attempts to mix them to obtain a homogeneous substance.
Table 1 Comments regarding the behavior of the mixture of water and TX-151 during the first attempts to obtain a homogenous substance.
| The ratio of the mixture water : powder : additive | Appendix | Notes |
| 7 : 1 : – | The sample does not solidify | |
| 2 : 1 : – | The sample solidifies, cracks appeared | |
| 4 : 1 : 1 | Oil | The sample does not solidify |
| 5 : 1 : 1 | Oil | The sample does not solidify |
| 3 : 1 : 2 | Chalk | The sample solidifies, lumps appeared |
| 3 : 2 : 1 | Salt | The sample solidifies, cracks appeared |
The results for the first successful measurement series were presented on charts for the refractive indexes and absorption coefficients of samples with different thickness but the same composition. The relations between the refractive index or absorption coefficient and the frequency of terahertz radiation for samples number 1 are shown in Fig. 1 and Fig. 2.
Each series of the measurements allowed to determine which compositions should be considered when trying to create a model of human tissue for a terahertz wave at a frequency of 0.5 THz. During the research, it was possible to obtain samples that met the theoretical assumptions of the tissues several times.
The attempts to create a healthy tissue model while taking the absorption coefficient into consideration allowed to obtain the following results. Table 2 shows the samples with absorption coefficients at a frequency of 0.5 THz which are the closest to the theoretical assumptions.
Table 2 The samples with the best fit to the theoretical assumptions for a healthy tissue in terms of the absorption coefficient.
| Number of sample | Thickness of sample [mm] | α [mm-1] |
| 1 | 0,4 | 10,00 |
| 2 | 0,3 | 9,70 |
| 5 | 0,3 | 9,80 |
| 6 | 0,4 | 10,00 |
| 7 | 0,4 | 10,70 |
| 8 | 0,4 | 9,80 |
| 25 | 0,5 | 9,70 |
| 45 | 0,3 | 10,70 |
Taking the information from the table above into account, it turns out that in order to obtain a tissue model with a refractive index equal to that of healthy human skin, one needs to create a mixture of water and TX-151 powder in the proportion of 1: 1 and a thickness of 0.4 mm or in the proportion of 4:1 with added salt, the water to salt ratio is 10:1 and the sample thickness is 0.4 mm. These two compositions showed a perfect fit out of all 45 samples with the theoretical assumptions regarding the skin’s refractive index at the frequency of 0.5 THz.
In addition to trying to determine an ideal model of healthy human skin, the samples created during the research were also measured when it comes to their similarity to the theoretical assumptions about the parameters of the most common skin cancer BCC. Table 3 contains the samples with the refractive index values closest to the values of the refractive index for BCC.
Table 3 The samples with the best fit to the theoretical assumptions for a diseased tissue in terms of the refractive index.
| Number of sample | Thickness of sample [mm] | n |
| 6 | 0,1 | 2,55 |
| 10 | 0,3 | 2,31 |
| 14 | 0,5 | 2,22 |
| 20 | 0,5 | 2,23 |
| 35 | 0,3 | 2,23 |
| 36 | 0,3 | 2,36 |
| 36 | 0,5 | 2,32 |
| 37 | 0,5 | 2,18 |
The results obtained from the samples containing flour (20 and 35) turned out to be the closest to the value of the refractive index for BCC at the frequency of 0.5 THz. Interestingly, both samples had an almost identical composition, they had a different amount of additive and a different thickness. Therefore, a conclusion can be drawn that the model of the affected tissue successfully representing the value of the BCC refractive index is a mixture of water, TX-151 powder and flour.
Due to the conducted research, it was also possible to distinguish several models corresponding to the assumptions concerning the BCC absorption coefficient. These results are presented in Table 4.
Table 4 The samples with the best fit to the theoretical assumptions for a diseased tissue in terms of the absorption coefficient
| Number of sample | Thickness of sample [mm] | α [mm-1] |
| 3 | 0,3 | 12,60 |
| 4 | 0,3 | 11,40 |
| 6 | 0,1 | 13,70 |
| 24 | 0,3 | 12,22 |
| 44 | 0,5 | 12,10 |
The sample which was the mixture of water and TX-151 in the proportion 3: 1 without any addition turned out to be the best fitted model of the diseased tissue in terms of the absorption coefficient for the terahertz wave at a frequency of 0.5 THz.
The most interesting model, which showed the similarity of the tested parameters and the theoretical assumptions, is the sample number 6 (which had the thickness of 0.1 mm) with the addition of salt.