Computational Approaches to Skin Cancer Detection using Torsion Wave Elastography and Inverse Problem

Abstract

Accurate estimation of skin tumor tissue parameters is critical for diagnosing and treating dermatology. However, signi cant challenges in this endeavor are posed by the multilayered structure of the skin. The aim of this study is to address these challenges through the use of two fractional rheological models: Spring-pot (SP) and Kelvin-Voigt fractional derivative (KVFD) models. These models are integrated with Torsional wave elastography (TWE) and Probabilistic inverse problem (PIP) techniques. The goal is to improve the estimation of skin tissue parameters in both laboratory and clinical settings. The research will be conducted in two phases. In the rst step, twelve bilayer tissue mimicking phantoms are tested. These phantoms simulate the properties of skin tissue using a specially developed TWE probe. The potential of the SP and KVFD models to estimate skin tissue parameters will be explored through inverse problem solving. Experimental data will be validated using ultrafast imaging from the Verasonics Research System. Pearson correlation, Dynamic time warping (DTW), and time-frequency plotting are used to evaluate the correlation between these models and experimental results. In the second phase, the methodology is applied to a group of healthy volunteers. The aim of this phase is to understand the variability of skin tissue properties among di erent individuals. It will also re ne the application of the models in a realistic clinical context. This study is a contribution to the eld of dermatologic diagnostics by investigating the feasibility and e cacy of the SP and KVFD models in conjunction with TWE. The aim of the research is to improve the diagnostic process for skin tumors, with potential implications for future developments in dermatological research and clinical applications. The potential of the methodology in the eld of dermatology is highlighted by its application in a range of scenarios, from controlled laboratory settings to real cases.