A recently introduced dermatoscopic method for diagnosis of early lentigo maligna (LM) is based on the absence of prevalent patterns of pigmented actinic keratosis (PAK) and solar lentigo/flat seborrheic keratosis (SL/SK). We term this the “inverse approach” OBJECTIVE: To determine whether training on the inverse approach increases the diagnostic accuracy of readers as compared to classic pattern analysis.
We used clinical and dermatoscopic images of histopathologically diagnosed LMs, PAKs and SLs/SKs. Participants of a dermatoscopy masterclass classified the lesions at baseline, after training on pattern analysis and the inverse approach. We compared their diagnostic performance among the 3 time points and to that of a trained convolutional neural network (CNN).
The mean sensitivity for LM without training was 51.5%, after training on pattern analysis increased to 56.7% and after learning the inverse approach to 83.6%. The mean proportion of correct answers at the 3 time points was 62.1%, 65.5% and 78.5%. The percentage of readers outperforming the CNN was 6.4%, 15.4% and 53.9%, respectively.
The experimental setting and the inclusion of histopathologically diagnosed lesions only.
The inverse approach, added to the classic pattern analysis, significantly improves the sensitivity of human readers for early LM diagnosis.

Copyright © 2020. Published by Elsevier Inc.

Author