COMPUTER-AIDED DIAGNOSIS IN SCREENING MRI OF WOMEN WITH EXTREMELY DENSE BREASTS

Abstract

Women with extremely dense breasts (Breast Imaging Reporting and Data System [BI-RADS] class D), that is, breasts containing a large amount of fibroglandular tissue, have a 2 to 6 times higher risk of developing breast cancer than women with very fatty breasts. Moreover, these cancers are harder to detect on mammography due to the low contrast between fibroglandular tissue and tumor tissue and overlapping tissue. Additional MRI screening for these women has proven to detect additional cancers and reduce the number of interval cancers at the cost of false positive referrals and increased workload. In this research computer aided methods were developed to score screening MRI examination of women with extremely dense breasts. To reduce the workload computer-aided triaging was developed to dismiss normal breast examinations (i.e., examinations that show no lesions) from radiological review without missing malignant disease. The method showed potential to dismiss 39.7% of normal bilateral breast examinations. Computer-Aided diagnosis (CAD) was developed to distinguish between benign and malignant lesions in the first round of the dense trial. The presented method was able to correctly classify 41.5% of benign BI-RADS 3 and BI-RADS 4 lesions as benign without missing malignant lesions. Both methods were validated on data acquired in the second screening round and showed robustness with similar results in another dataset.