The goal of the proposed research is to develop computer-aided diagnosis (CAD) schemes to assist radiologists in diagnosing breast cancer from mammograms. The hypothesis of this project is that CAD will improve radiologists' ability to interpret mammograms, so that both the number of missed cancers and the number of women unnecessarily sent to biopsy can be reduced. This proposal is specifically for the development of CAD schemes for the detection and the classification (benign versus malignant) of microcalcifications from digital mammograms. This project will continue to refine the current CAD schemes, towards our goal of clinically viable CAD schemes.
The specific aims of this proposal are: (1) To improve the performance of our detection scheme by detecting subtle clusters based on their brightest microcalcifications; by decreasing the number of false positives by eliminating those caused by calcified vessels or obvious benign calcifications; and by improving the robustness of the scheme; (2) To study the effect of scoring methodology on the measured performance of a detection scheme and then to devise the """"""""best"""""""" scoring method for clustered microcalcifications; (3) To improve the performance of the classification scheme: by enlarging our database of benign and malignant cases; by identifying additional computer-extracted features; and by utilizing information from multiple sources: magnification views, two views of the same breast, and the same views of the same breast taken at different times; (4) To combine the detection and classification schemes, thereby fully automating the classification scheme. This will entail examining the effects on the classification scheme of incomplete detection of all the microcalcifications within a cluster, of false positive signals being included in a true cluster, and of false positive clusters. Using this information, the applicants proposed to develop an interface between the detection and classification schemes; and (5) To perform a preliminary clinical evaluation of the classification scheme.
Showing the most recent 10 out of 19 publications