Computational modeling to determine the physical characteristics of biological tissues for medical diagnosis

TL;DR

This paper presents a computational approach using computer simulations and parallel processing to determine physical characteristics of biological tissues, aiding personalized diagnosis of breast cancer through microwave radiothermometry.

Contribution

It introduces a method to estimate thermal properties of tissues from temperature data using least squares and parallel computing, enhancing diagnostic accuracy.

Findings

Good agreement with known solutions validates the method.

Parallel computing improves efficiency of inverse problem solving.

The approach supports personalized modeling of biological tissues.

Abstract

Timely diagnosis of breast cancer is an important task. This type of breast cancer is one of the most common diseases. The method of microwave radiothermometry is a promising direction for solving this problem. The method is based on measuring internal temperature of biological tissues in microwave frequency range. Computer simulations are used to improve the quality of diagnostics. Computer models make it possible to evaluate the effect of heat release in a malignant tumor on the thermal dynamics inside the mammary gland. It is necessary to build personalized models, taking into account the individual nature of the internal structure of the mammary gland in each patient. One of the problems is the determination of biophysical characteristics of biological components. Methods for determining these characteristics using computer simulations are proposed. The coefficient of thermal conductivity and specific heat of biological tissues are determined from known temperature distributions. Finding the physical parameters for a quasihomogeneous biological tissue is the first approximation for solving this problem. The least squares method is used as a solution method. The results obtained are in good agreement with previously known exact solutions, which indicates the applicability of this method for solving this class of problems. The efficiency of using parallel technologies in solving the inverse problem is investigated and the applicability of Open MP technology is demonstrated.