Schaap MJ, Cardozo NJ, Patel A, et al. - Findings show the potential of Convolutional Neural Networks (CNNs) to automatically and objectively perform image-based Psoriasis Area and Severity Index (PASI) scoring at an anatomical region level. The performance of CNNs, for erythema, desquamation and induration scoring, was similar to that of physicians, while for area scoring CNNs performed better than physicians on image-based PASI scoring.

• Performance of image-based automated PASI scoring in anatomical regions by CNNs was assessed, and performance of CNNs was compared to image-based scoring by physicians.

• For erythema, desquamation, induration and area, the intraclass correlation coefficients (ICCs) between the CNN and real-life scores of the trunk region were 0.616, 0.580, 0.580 and 0.793 respectively, and similar findings were evident for the arms and legs region.

• There were slightly higher ICCs between the CNN and real-life scores for erythema (0.616 vs 0.558), induration (0.580 vs 0.573) and area scoring (0.793 vs 0.694) vs image-based scoring by physicians.

• The performance of physicians was slightly better than the CNN on desquamation scoring (0.580 vs 0.589).