ABSTRACT
Different score-based dermoscopy algorithms (e.g., ABCD, Menzies, seven-point checklist, CASH) have been developed to assist clinicians in distinguishing DN from melanoma (Stolz et al., 1994; Menzies, 2009; Argenziano et al., 1998). Many DN that are clinically suspicious for melanoma can be correctly identifi ed as benign using such dermoscopy algorithms. Yet, there still remains a signifi cant portion of DN that score in the “suspicious” range implying the algorithm cannot distinguish these DN from melanoma with assurance. This is due to the considerable overlap in the primary morphology between DN and melanoma not only clinically but also dermoscopically. Pattern analysis, the most frequent method used by experienced dermoscopists, has been shown to better discriminate between many DN and melanomas as compared with other scoring algorithms (Argenziano et al., 2003). Pattern analysis involves determining the overall (global) dermoscopic pattern (Table 7b.1, Fig. 7b.3) based on the presence or absence of network, globules, homogenous (structureless) areas, and identifying the presence or absence of specifi c local features (Table 7b.2). In evaluating patterns, the observer must consider not only the overall dermoscopic pattern, but also the number and distribution of colors and structures, whether the dermoscopic structures are “regular” or “irregular” (see chap. 6a for further details), and whether the colors and structures are distributed symmetrically or asymmetrically and in an organized or disorganized manner.
