The Value Of CAD For Breast Microcalcification

ObjectiveMicrocalcification plays a very important role in detection of breast cancer, especially early-stage breast cancer. The purpose of the study is to evaluate the diagnose effect of CAD system and discuss the relationship between microcalcification imaging characteristics and pathology and CAD system by using full-field digital mammography and computer-assisted detection.MethodsThis study included 111 cases of microcalcifications(including 5 patients with bilateral lesions), treated in our department from 2009 to 2010. Record item consisted age, pathology, breast parenchymal density, microcalcification shape, microcalcification distribution, number per unit area, location and accompanied by mass et al. The relationship between microcalcification imaging characteristics and pathology and the detection of CAD was analyzed.ResultsThere were 50 malignant lesions(45.0%)and 61 benign lesions(55.0%). Four variables had significant value for pathologic analysis, including microcalcification shape(P<0.001), distribution(P<0.001), number of per unit area(P<0.001), and combination with a mass(P<0.001). Furthermore, microcalcification shape(P=0.017), number of per unit area(P=0.023) had a certein extent in the detection of CAD. However, the current CAD iis unuseful to distinguish the malignant from microcalcification lesions. The sensitivity and specificty of CAD was 94.0%, 14.8% respectively for microcalcification detection.Conclusions1. Microcalcification imaging features, including shape, distribution, number of per unit and combination with mass, can help improve diagnostic accuracy. 2. The current CAD system is unuseful to distinguish the malignant one from microcalcification lesions. Its purpose is to improve the detection rate of microcalcification.3. Combination of imaging features of microcalcification and CAD system can improve the detection rate of breast cancer.