Computer-Aided Detection (CAD) of Breast-Cancer — Diagnostic Performance of a Further Improved System

Abstract

Instead of supreme spatial resolution and the depiction of subtile differences of contrast as a characteristic for today’s mammography the diagnostic performance mainly depends on the mammographer and his experience and training. So cross-checking the findings is usually practised. This time-consuming process led to the demand of computerized diagnostic aid to identify suspicious regions of interest, thus giving the opportunity to guide the radiologist to prior not appropriately noted lesions. The goal of our study was to compare the diagnostic performance of a first and second version of a computer-aided detection system on the same sample of mammograms depicting histologically verified breast-cancer; to clarify the possible progress in identifying suspicious findings on mammograms.