ABSTRACT
Computer-aided diagnosis (CAD) refers to an approach to diagnostic practice in which the radiologist in the interpretation of medical images is assisted by a computer, which analyzes the image data with a mathematical algorithm implemented as computer software. Currently, only computer-aided detection (CADe) has been introduced in clinical practice. e goal of CADe is to help radiologists detect cancers or other abnormalities that they might miss without the assistance of CADe. In CADe, the computer identies a number of possible lesions or abnormalities in the images and presents the results to the radiologist, who then must decide whether the computer detection results are true lesions or true abnormalities or false positives. is chapter discusses several methodological issues for the assessment of the clinical eect of CAD, particularly of CADe.
