Department of Dermatology and Cutaneous Surgery (Dr. Usatine) and Department of Family and Community Medicine (Drs. Usatine and Shama), University of Texas Health, San Antonio; Memorial Sloan Kettering Skin Cancer Center, Hauppauge, NY (Dr. Marghoob); Department of Dermatology and Cutaneous Surgery, University of Miami, Fla (Dr. Jaimes) usatine@uthscsa.edu
Dr. Marghoob discloses that he provides free advice to Heine, Canfield, and 3GEN regarding their products. He has also received honoraria to speak for 3GEN and to test equipment for Heine, 3GEN, and Canfield.
The other authors reported no potential conflict of interest relevant to this article.
The first of 2 major algorithms that can be used to guide dermoscopic analysis is a modified pattern analysis put forth by Kittler.10 This descriptive system based on geometric elements, patterns, colors, and clues guides the observer to a specific diagnosis without categorizing lesions as being either melanocytic or nonmelanocytic. Because this is not the preferred method of the authors, we will move on to Method 2.
The second method, a 2-step algorithm, is a qualitative system that guides the observer through differentiating melanocytic from nonmelanocytic lesions in order to differentiate nevi from melanoma (FIGURE 2). At the same time, it serves as an aid to correctly diagnose non-melanocytic lesions. The 2-step algorithm forms the foundation for the dermoscopic evaluation of skin lesions in this article.
Not all expert dermoscopists employ structured analytical systems or methods to reach a diagnosis. Because of their vast experience, many rely purely on pattern recognition. But algorithms can facilitate non-experts in dermoscopy in the differentiation of nevi from melanoma or, simply, in differentiating the benign from the malignant.
Although each algorithm has its unique criteria, all of them require training and practice and familiarity with the terms used to describe morphologic structures. The International Dermoscopy Society recently published a consensus paper designating some terms as preferred over others.11