Bio Phantoms Mimicking the Dielectric and Mechanical Properties of Human Skin Tissue at Low-Frequency Ranges

Tissue phantoms are widely used as substitute materials for real tissue validation of various newly emerging biomedical technologies such as ultrasound (US), computed tomography (CT), and magnetic resonance imaging (MRI). However, there is no specific recipe for fabricating skin-mimicking phantoms which can mimic both the mechanical and dielectric properties of human skin at lower frequency ranges.


Introduction
Tissue-mimicking (TM) phantoms are vivid models of real human tissue that exhibit realistic properties of tissues in certain areas (Porter et al., n.d.). As real human tissue samples are difficult to obtain and store (Bot et al., 2009) (Singh et al., 2016), tissue phantoms are making a significant contribution to the characterization of the new imaging technologies and medical training. Human skin is the heaviest and vastest organ of the human body, which performs critical functions to human health, especially in regulation and protection. The development of the skin phantoms will facilitate the development of biomedical applications and contribute to skin clinical research, particularly for cosmetic, dermatology, and detection of cutaneous pathology  (Garrett et al., 2014) (Moll and Dennis, n.d.) (Sugihara et al., 1991) (Meaney et al., 2012).
Although the field of tissue-mimicking phantoms manufacturing is becoming more attractive and many researchers have achieved active explorations in it, there is no specific recipe for fabricating skin-mimicking phantoms that can properly match mechanical and electrical properties of human skin. The dielectric properties of most existing skin-mimicking phantoms were measured at high frequencies (normally over 500MHz) to satisfy the requirement of microwave imaging technology (Meaney et al., 2012) (Popovic et al., 2005). To date, less research has been done on phantoms mimicking human skin at low frequencies because of the error introduced due to the electrode-polarization effect.
However, this relatively blank research area has attracted more attention recently with the gradual clinical application of low-frequency technologies such as electrical impedance tomography (EIT) (Riu and Anton, 2010) (Ahn et al., 2010). The adjustable frequency of currently commercial devices of EIT for clinical use is below 150 kHz (Orschulik and Menden, 2017).

Method
As an ex experimen in this rese

Oil-in-Gelatin TMM
The investigated materials are mainly a mixture of gelatin solution and a solution of mixed oil (50 vol% kerosene and 50 vol% safflower oil). The volume content of the mixed oil (from 0 to 40%) is the variation factor to tune the properties of the prepared samples (details in table 3).
For comparison purposes, samples without adding oil were also prepared. The five main steps of preparing the oil-in-gelatin TMMs are described below.
Step 2: Add the solvent produced in Step I using a quantitative pipette into a new small beaker. Add the desired amount of deionized (DI) water and mix the solution at room temperature until the solution becomes white.
Step 3: Add dry mass gelatin into the solution and keep stirring. Heat and stir the mixture on a hotplate at 100℃ of temperature until the mixture becomes transparent (see figure 6). Step 4: Prepare the desired amount of mixed oil (50 vol% kerosene and 50 vol% safflower oil) in another beaker and heat at 50 ℃. Pour the gelatin mixture prepared in step 3 and liquid surfactant into the beaker.

Lignin/ Graphene Nanopowder TMM
Desired quantities of Lignin (0-6wt%) and graphene nanopowder (0 to 0.15%) were added to the gelatin mixture prepared above in Step III (see figure 8). The rest of the recipe was the same as used to prepare oil-in-gelatin mixtures. The concentrations of each ingredient can be seen in tables 4 and 5.     Table 6. C

Gelatin
Desired qu solution w The oil inc oil percent   Table 8. C

Shapin
The dime Appropria data.
The  Vol. 14, No. 7;2020 A circular cutter consisting of two concentric blades with diameters 44.6mm and 52.6mm was designed and manufactured. Homogeneous film samples of the TMMs were generated by pouring the molten mixture (above 45℃) into a petri dish with a height of 5mm to solidify. After the mixture was cured, the cutting tool was used to cut a circular film of desired dimensions ( Figure 13). Figure 13. 3D design of a cutting tool and a sample cut in the desired shape

Mechanical Testing of the TMMs
A compression test was performed on the samples to evaluate their Young's Modulus (YM) using a TA.XT.Plus Texture Analyzer from Stable Micro Systems (Surrey, UK). The settings listed in table 9 were used to obtain a stress-strain curve, from which YM was calculated.

Five types results obt
The mecha the ingredi  or Sum of Squ n equation (7) to evaluat    Figure 19. Cole-Cole plot for the 20 vol% oil sample

Mechanical Properties
Lignin Figure 20. The stress-strain plot of oil-in-gelatin sample on adding 4 wt% lignin      Desired mechanical properties were achieved on using gelatin, DI mixed oil, and lignin for preparing TMMs, as discussed in section IV, therefore only electrical properties of the remaining TMMs were tested.       Resistance, R (Ohms)

Discussion
On testing the mechanical properties of oil in gelatin TMMs, it was observed that the YM increases on increasing the concentration of oil in gelatin. The range of YM of the oil-in-gelatin phantoms varied from 8 kPa to 50 kPa. This range is well within the YM of human skin, measured by indentation and suction techniques (5kPa to 260 kPa), as stated in section I.
The samples became more fragile and stickier when large volumes of oil were added in gelatin, therefore adding more than 30vol% oil in gelatin is not advisable.
The changes in permittivity and conductivity with a frequency sweep can be seen in figures 17 and 18 respectively. At lower frequencies, no significant change in the permittivity was observed on increasing the oil concentration. This outcome is different than the study performed by Lazebnik [26] in 2005 at higher frequencies (500MHz-20GHz). The contradiction in results can either be due to the addition of solvents-p-toluic acid (powder) and n-propanol in oil in gelatin TMMs or due to the polarization effect at low frequencies. However, the polarization effect was reduced using SSE curve fitting in Matlab, as discussed in section II, and the best Cole-Cole fit was achieved (see figure 19). The conductivities of the TMMs increased on increasing the oil concentration, which is in alignment with the trend in literature.
The addition of lignin and graphene nanopowder further reduced the YM of the TMMs (tables 12 and 13), although the YM of TMMs (25kPa-33kPa) was closer to human skin than oil-in-gelatin TMMs. Adding more than 6% lignin to the mixture wasn't feasible due to high viscosity and poor agitation of the mixture. The increase in YM on the addition of graphene nanopowder can be due to the suspension of insoluble graphene powder in the oil-gelatin based emulsions.
An increase in the permittivity and conductivities of the TMMs was observed on increasing the concentration of lignin and graphene nanopowders in oil-in-gelatin mixtures. This trend is in agreement with the findings in the literature (Lan et al., n.d.).
The measured data for both lignin and graphene nanopowders agreed with the Cole-Cole plot (figures 24 and 27), indicating their ability to mimic biological tissues.
As seen from figure 28, the permittivity was almost unchangeable on increasing the concentration of gelatin in DI, although the conductivity of the gelatin-DI mixture was closer to wet and dry skin at 25 wt% concentration of gelatin in DI.
However, on adding 5wt% mixed oil in gelatin-DI (see figure 30), the permittivity and conductivity dropped by 17067 and 0.42 S/m, respectively, which is closer to the values for dry and wet skin. Similarly, the addition of lignin reduced the permittivity and conductivity of the samples at low frequencies.

Conclusion
This study examined the mechanical and electrical performance of five kinds of TMMs at a frequency range of 20 Hz to 300 kHz. Amongst them, the oil-in-gelatin based TMMs were mainly standing on the contributions in (Lazebnik et al., 2005) (Lan et al., n.d.). The shreds of evidence from this study confirm that in general these tested TMMs can reach the mechanical requirement for fabricating skin phantoms.
The mechanical properties of TMMs were tested using compression tests based on the assumption that they are homogeneous in nature. More accurate research in the future should consider the heterogeneous and anisotropic properties of human skin.
This study fulfills the requirement of preparing high-quality skin phantoms using readily available materials at a lower frequency range. The electrode polarization effect, which occurs mainly at low frequencies was also removed using model fitting in Matlab, and a resemblance to the Cole-Cole plot was found in all TMMs.
All proposed skin phantom materials reflect some regularity between the dielectric properties and ingredients' concentrations. Although some exceptions cannot be explained, for example, the oil-in-gelatin based TMMs didn't show a substantial increase in permittivity at higher concentrations of oil in contrast with the findings in the literature.
The concentrations of oil, the lignin or graphene nanopowder can be further increased, but this will affect the Young's Modulus of the TMMs.
In conclusion, we propose the use of gelatin-DI TMMs doped with mixed oil or lignin for low-frequency applications. A concentration of 5-15wt% of mixed oil (50 vol% kerosene and 50 vol% sunflower oil) or 2-6wt% of lignin in Gelatin and DI mixtures are the most suitable choices for preparing TMMs for the skin. Future work can be done in testing the shelf life and biocompatibility of these materials.