Computerized Detection of Lesions in Diagnostic Images

Abstract

Computer-aided detection (CADe) has been an active research area in medical imaging. As imaging technologies advance, a large number of medical images are produced which physicians/radiologists must read. They may overlook lesions from such a large number of medical images. Consequently, CADe that provides suspicious lesions with radiologists/physicians is developed and becoming indispensable in their decision making to prevent them from overlooking lesions. Machine learning (ML) plays an essential role in CADe, because lesions and organs in medical images may be too complex to be represented accurately by a simple equation; modeling of such complex objects often requires a number of parameters that have to be determined by data. In this chapter, ML techniques used in CADe schemes for lung nodules in chest radiography and thoracic CT and those for the detection of polyps in CT colonography (CTC) are described, which include patch-/pixel-based ML and feature-based (segmented-object-based) ML.