A Low-Cost Radar-based Domain Adaptive Breast Cancer Screening System
Samuel Claflin, Mohammad Arif Ul Alam · 2020
Over the past three decades, the advancements of breast cancer screening technologies such mammography, ultrasound, Magnetic resonance imaging (MRI) saved countless lives. The invention of mammography screening of breast cancer in the 1990s led a technological revolution which is, now-a-days, coupled with MRI and/or ultrasound to achieve diagnoses of much greater accuracy than previously attainable. However, these technologies (mammography, ultrasound, MRI) are not as widely available to patients as one might assume less fortunate countries (such as Bangladesh) often cannot afford the potentially enormous price tag that several ultrasound machines of sufficient quality for accurate diagnoses incurs. In this paper, we present a low-cost (<100 USD) millimeter Wave (mmWave) Radar sensor array (3– 10 GHz) imaging technology and a deep learning domain adaptation model-based breast cancer screening system. More specifically, (i) we develop a mmWave Radar sensor array (18 sensor antennas) based 2D imaging system; (ii) we develop a deep learning based domain adaptation model that can learn breast segmentation and cancer detection from expensive source data (mammography, ultrasound) and transfer the knowledge to less expensive target data (Radar images), (iii) we validated our system and methods by utilizing our existing mammography and ultrasound breast cancer screening data as well as 14 patients' Radar images collected from a third world country (Bangladesh).