The comment that the aim of any melanoma diagnostic system is to remove all lesions that may be melanoma while minimizing excision of benign lesions is sound. There are a number of descriptive statistics that provide relevant data toward this end: sensitivity, specificity, diagnostic accuracy (true-positive melanoma/[true-positive melanoma + false-positive melanoma + false-negative melanoma]); positive predictive value (true positive/[true positive + false positive]); and negative predictive value (true negative/[true negative + false negative]).1 Dr Bystryn's claim2 is that such instrumentation should focus on ensuring 100% sensitivity in diagnosing melanoma at the expense of a lower specificity, negative predictive value, and perhaps diagnostic accuracy. This approach is also sound, provided the specificity is not low enough to result in an unrealistic health cost and morbidity due to needless excisions.
Menzies SW. Epiluminescence Microscopy Diagnostic Criteria With Follow-up Computer-Based Monitoring of "Less Suspicious" Lesions May Increase Sensitivity for the Diagnosis of Melanoma While Maintaining Adequate Specificity. Arch Dermatol. 2001;137(3):378–379. doi:
Artificial Intelligence Resource Center
Customize your JAMA Network experience by selecting one or more topics from the list below.