(C-81) Accounting for variable skin thickness in quantifying tissue optical properties in small animal models

Introduction:: Diffuse optical spectroscopy is a non-invasive and non-ionizing method that uses visible and near-infrared light to provide quantitative estimates of tumor biology based on light interaction with tissue. Prior to clinical translation, optical spectroscopy studies typically use small animal models, such as tumor xenografts in mice, to validate the technology and biomarkers. Both analytical and empirical models of light-tissue interaction assume constant absorption by the overlying mouse skin to quantify tumor optical properties. However, as cancerous cells proliferate, they displace other cells causing the surrounding tissues to stretch. As the tissues stretch, their thickness around the tumor will vary. This implies that the thickness of the skin will be different at different tumor volumes, and assuming a constant thickness could potentially lead to errors in the quantification of tumor optical properties. While previous studies, such as Sabino et al., have measured the absorbance spectrum of mouse skin, the findings are only limitedly useful because the study only looks at a set skin thickness, so it is unclear how light transmission would behave if the tissue were thinner. Due to this ambiguity, this study sought to quantify the changes in absorbance and transmission for varying amounts of skin stretch.

Materials and Methods:: Skin excised from a C57BL/6 mouse was mounted into a biaxial stretching device (Figure 1). A liquid crystal tunable filter (LCTF, CRi) was mounted onto a white light LED (Thorlabs), and the light was collimated using a collimating adapter (Photometrics) (Figure 1B). Prior to data collection, the skin samples were pretensed for 20 cycles to mitigate the risk of them tearing during stretching. Afterward, samples were brought back to 0 mm displacement. The skin was stretched at seven different displacements at 0.5 mm increments from 0 to 3 mm, and intensity images were captured with a CMOS camera (Basler). Integration times and LED driver output power were adjusted to avoid oversaturating the sensor. Neutral density filters were also placed and/or removed from the illumination path to further mitigate oversaturating the sensor. All images were normalized, and averages at each wavelength were calculated using a Python script (Visual Studio Code) to determine transmission dependence on the stretch.

Results, Conclusions, and Discussions:: Our preliminary results demonstrate large differences in wavelength-dependent absorbance at different skin stretches. We plan to present data from optical spectroscopy studies with and without incorporating tissue thickness into light-transport models to determine the influence of skin thickness on the quantification of tumor optical properties. The results of this study will be useful to improve upon current light-transport models and enable better analysis of pre-clinical and clinical data.

Acknowledgements (Optional): :

References (Optional): :