JAMA Dermatology (formerly Archives of Dermatology) reported a study (Arch Dermatol. 2012;148:1083-1084) on the diagnosis of pigmented lesions by dermatologists compared with the MelaFind device (Mela Sciences), a handheld light unit that produces a 3-dimensional computer image of a pigmented lesion with subsequent recommendation of “high disorganization” versus “low disorganization.” Of 47 pigmented lesions (23 melanoma; 24 nonmelanoma), MelaFind recommended biopsy in 44 cases and no biopsy in 3 cases, with 1 lesion that was truly melanoma (96% sensitivity; 8% specificity), compared with a range of 48% to 100% (mean, 80%) sensitivity and 4% to 71% (mean, 43%) specificity among the study dermatologists. The authors quote in their conclusion paragraph that MelaFind is a “very sensitive tool to guide dermatologists.”
What’s the issue?
A recent patient forcefully suggested to me that our department should “keep up with the times” and invest in MelaFind. MelaFind’s Web site states that it is “not a screening device.” The physician fact sheet says that it is “the world’s first and only multi-spectral, non-invasive, painless, and 100% objective and automated computer vision technology that evaluates clinically atypical pigmented skin lesions and classifies them unambiguously and clearly based upon their level of 3-dimensional morphological disorganization.” Given these claims and its US Food and Drug Administration approval, how can we not own one? The word unambiguous has never been used to describe the clinical decision making and diagnosis of pigmented lesions, but we all hope for it. Quite well-known leaders in our field have spoken highly of this technology, though I await better-powered studies and more poignant data, or perhaps just a better grasp of how it works. Can someone aid us in understanding its true niche in clinical practice?