Imaging for Early Microwave Breast Cancer Quantitative Diagnosis

This letter proposes a near-field confocal microwave imaging (CMI) approach to improve the accuracy of breast cancer identification. An iterative process integrates the results of the time-delay and point spread function methods as mutual prior information, which enables precise phase and amplitude compensation in the spherical wave model. Frequency-domain zero-padding and inverse discrete Fourier transform methods are applied to optimize imaging accuracy while maintaining efficiency. The proposed method is validated in a breast cancer detection scenario, where the structural similarity index metric is employed to evaluate reconstructed energy distribution. Results indicate a 52.59% improvement in distribution accuracy compared to traditional CMI.