Focal quality metrics for the objective evaluation of confocal microwave images

Abstract

Confocal microwave imaging for breast cancer detection relies on accurate knowledge of the average dielectric properties of the patient-specific breast. When accurately estimated, coherent addition will occur at the tumor site, producing a clear and sharp image thereof. Conversely, if the average dielectric properties are poorly estimated, a blurred, unfocused image will be reconstructed, potentially obscuring cancerous lesions. Several methods have been proposed to estimate the patient-specific average dielectric properties, for example, time-of-flight estimation. However, such methods are specific to the individual imaging hardware, can be susceptible to multipath propagation and assume the chosen paths are representative of the whole volume. In this paper, a novel method to estimate the patient-specific average dielectric properties is presented, based on focal quality metrics (FQMs); used historically to measure the clarity and focus of microscopic or digital photographic images. These FQMs are applied to confocal microwave breast images to assess their focus, and hence estimate the patient-specific average dielectric properties. In this way, FQMs can be used to generate the optimum microwave image of the breast. The performance and robustness of these FQMs for microwave breast imaging applications is examined in this paper and preliminary results are presented and discussed.