The atypical-mole (dysplastic nevus) syndrome is a genetically determined fault in melanocytes (pigment cells) of the skin. Afflicted individuals often have numerous, large, and 'atypical' (melanoma-appearing) moles scattered on their cutaneous surfaces. Importantly, these individuals are at high risk for developing malignant melanoma. Our studies show that the risk of developing malignant melanoma is higher than almost any other known risk factor for this cutaneous cancer. Furthermore, a gradation of risk exists depending on a personal and/or family history of malignant melanoma. Based on these observations, we published a new classification of the syndrome.

We also described the 'classic' atypical-mole syndrome in which patients have a triad of: 1) 100 or more moles, 2) at least one mole resembling melanoma, and 3) at least one mole 8 mm or larger in diameter. This subset of individuals has one of the highest risks for melanoma ever reported.

Most recently, our work has concentrated on the development of improved in vivo diagnostic methods in a differential diagnosis between atypical moles and malignant melanomas. Using digitized image analysis and computer vision, algorithms have been developed to differentiate these closely appearing melanocytic neoplasms. To accomplish this, we visualize lesions in vivo under 10X magnification with a dermoscope. Dermoscopic photographs are then digitized for computer analysis. In preliminary studies, 100% sensitivity but much lower specificity has been achieved. These studies, although preliminary, are exceedingly important in possibly providing a technique for the differentiation between atypical moles and early malignant melanomas.