Unimodal Approach for Detection and Classification of Breast Cancer Using Mask Scoring R-CNN

Breast cancer, a highly aggressive disease affecting women, presents abnormalities like masses, calcification, and lumps, causing more deaths than any other cancer. Early prediction and accurate diagnosis are essential to better patient outcomes. Breast cancer detection techniques include clinical examinations and self-examinations, with mammography being the main method. Complementary technologies like MRI and Biopsy aid in further assessment. Technological progress has led to 3D mammography, molecular imaging, and genetic testing. There are drawbacks to breast cancer detection techniques, such as overdiagnosis, overtreatment, and false-positive results. While biopsies entail risk, and MRI can lead to anxiety and overtreatment. There is a risk of overdiagnosis, and certain techniques might not identify some forms of breast cancer. In order to decrease false positives, increase accuracy, and lower risks, more research is required. The current paper suggests a Unimodal Approach for Detection and Classification of Breast Cancer Using Mask Scoring R-CNN. Mammogram imaging in breast cancer detection is improving through technological advancements, including high-frequency transducers, advanced software, and techniques. Automated and Computer-Aided Diagnosis systems, specialized training, and combining ultrasound with other imaging methods create a comprehensive diagnostic approach. Continuous research and development promise further enhancements. Comparative analysis against existing methods involves various metrics such as accuracy, true positive rate, true negative rate, precision, F-score, balanced classification rate, Youden's index, Jaccard Index, employing a mammogram database. Results from experiments emphasize the superiority of the proposed approach in effectively detecting, localizing, and classifying instances of breast cancer across multiple occurrences. As a diagnostic tool aligned with clinical objectives, it promises accurate anomaly diagnoses associated with breast cancer.