Modeling of Textile Antenna Sensors for Breast Cancer Detection Using Artificial Intelligence Techniques

Abstract

This paper proposes modelling and simulations for textile antenna-based sensors acting as tumor detection system for breast using artificial intelligence. The proposed textile sensors are lightweight, comfortable to wear with low fabrication cost at microwave frequencies. The sensors are modelled with a 3-layer breast phantom with skin, gland, and fatty tissue, along with a tumor with high water content. Simulations are conducted using three dimensions electromagnetic wave CST Studio simulator with 4 elements of antenna-based sensors. The utilized sensor operates on breast phantom in the band from 3.5 to 5 GHz with an average gain of 6 dBi. Tumors are simulated with diameters from 5 mm to 15 mm at the centre of breast and 45 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">o</sup> offset of the centre. For tumor detection, different machine learning techniques are applied to classify between malignant and benign cells using Scattering parameters. These techniques provide accurate estimation for tumor detection at different tumor sizes. 98% classification accuracy is realized for centre and shifted tumor using developed machine learning techniques.