Design and Optimization of Non-Invasive Microwave Sensors for Cancer Detection

Abstract

This paper presents the design, optimization, and validation of a non-invasive planar microwave resonator for early detection of malignant tissue based on dielectric contrast. Operating in the 3–4 GHz band, the sensor was optimized through fullwave simulations to maximize the frequency–permittivity sensitivity while maintaining compact geometry and safe exposure levels. The optimized design exhibits a resonant downshift of 40– 60 MHz for malignant inclusions with ε<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><i>r</i></sub> ≈ 50 and diameters of 8–10 mm located at depths up to 12–15 mm. The measured detection uncertainty (<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">u_f</i> ≈ 8–10 MHz) is significantly smaller than the tumor-induced perturbation, enabling reliable differentiation between healthy and malignant phantoms. Experimental validation demonstrates close agreement with simulations, with resonance deviations typically below 5%. The specific absorption rate (SAR) remained well under IEEE C95.1 limits for an input power of 1 mW, confirming safe operation. These results establish the proposed sensor as a compact and sensitive platform for microwave-based tissue abnormality detection and provide a quantitative framework for future clinical extensions targeting smaller or deeper lesions.