High-Resolution Breast Cancer Detection Using AOA-Optimized mm-Wave Antenna and GRU Classifier

Abstract

Breast cancer detection at an early stage remains an open challenge due to limitations in the resolution, cost, and portability of conventional imaging systems. This paper presents a high-resolution framework for breast cancer detection by integrating an Angle-of-Arrival-optimized millimeter-wave antenna with an intelligent deep-learning classifier. The antenna design is optimized using an evolutionary MAOA optimization algorithm to enhance directional gain, penetration capability, and spatial resolution, allowing for better localization of malignant tissues within heterogeneous breast phantoms. Backscattered mm-wave signals are preprocessed and input to a GRU-based neural classifier that learns temporal-spectral features associated with the presence of tumors. Experimental simulation and prototype measurements of the proposed system demonstrate superior detection accuracy with reduced false-positive rates and improved resolution compared with other millimeter-wave and microwave imaging approaches. The integration of optimized antenna design with a sequence-aware GRU model presents a promising pathway toward the realization of noninvasive, compact, and highly reliable breast-cancer screening technologies.