This proposal is directed at the development, application, and evaluation of a novel computer assisted diagnosis (CAD) method for the detection and classification of clustered calcifications from digitized mammograms. The long term aim of the research is the design of fully automated CAD methods that will have the role of """"""""a second reader"""""""" in the interpretation of digital mammograms with a significant impact on the standardization of reading, sensitivity, specificity, and telemammography.
The specific aims of the proposed study included: (a) the development of a CAD method that starts with the automatic extraction of calcifications from digitized mammograms using novel wavelet and quadrature mirror filter methodology and proceeds with computer-extracted features of the segmented calcifications for the automatic benign/malignant differentiation of the extracted classifications using neural networks with faster and more efficient training algorithms; (b) the development of a database with normal and abnormal (benign and malignant) cases with calcifications representative of clinical practice; (c) the initial evaluation of each module of the proposed CAD methodology using simulation studies and carefully selected sets of mammograms with ground truth determined by biopsy and experts; (d) the investigation of the effect of image resolution on the CAD outcome and on the design of each module of the algorithm aiming at the implementation of a flexible and possibly image independent technique; (e) the development of a user-friendly, open-architecture computer interface that will allow performance reading of digitized mammograms and CAD effectively from computer monitors aiming at future teleradiology applications; and (f) the conduct of a clinical study of the proposed CAD technique using receiver operating characteristic (ROC) analysis.
| Kallergi, Maria; Heine, John J; Berman, Claudia G et al. (2004) Improved interpretation of digitized mammography with wavelet processing: a localization response operating characteristic study. AJR Am J Roentgenol 182:697-703 |
| Kallergi, Maria (2004) Computer-aided diagnosis of mammographic microcalcification clusters. Med Phys 31:314-26 |
| Kallergi, M; Carney, G M; Gaviria, J (1999) Evaluating the performance of detection algorithms in digital mammography. Med Phys 26:267-75 |
| Kallergi, M; Gavrielides, M A; He, L et al. (1998) Simulation model of mammographic calcifications based on the American College of Radiology Breast Imaging Reporting and Data System, or BIRADS. Acad Radiol 5:670-9 |
