Pattern Analysis in Dermoscopic Images

In this chapter an extensive review of algorithmic methods that automatically detect patterns in dermoscopic images of pigmented lesions is presented. Pattern Analysis seeks to identify specific patterns, which may be global and local. It is the method most commonly used for providing diagnostic accuracy for cutaneous melanoma. In this chapter, a description of global and local patterns identified by pattern analysis is presented as well as a brief explanation of algorithmic methods that carry out the detection and classification of these patterns. Although the 7-Point Checklist method corresponds to a different diagnostic technique than pattern analysis, it can be considered as a simplification of it as it classifies seven features relatedwith local patterns. For this reason, themain techniques to automatically assess the 7-Point Checklist are briefly explained in this chapter.