What Type of Man Against Machine?

Dear Editor I request clarification of the type of man (or indeed woman) pitted against the machine [1.Haenssle H.A. Fink C. Schneiderbauer R. et al.Man against machine: diagnostic performance of a deep learning convolutional neural network for dermoscopic melanoma recognition in comparison to 58 dermatologists.Ann Oncol. 2018; 29: 1836-1842Abstract Full Text Full Text PDF PubMed Scopus (663) Google Scholar]. The authors note that 22.4% of the dermatologists showed a higher diagnostic performance than machine. However, the paper does not analyse this group who triumphed over Google. They warrant closer study. The authors initially describe the trial participants as ‘readers’ drawn from the International Dermoscopy Society (IDS) and then as dermatologists. The authors have stratified results by a timeframe of experience in dermatoscopy. However, an additional matrix would be helpful in assessing the results. The IDS, supported by FotoFinder Systems GmbH, has an open membership. The level of declared specialisation within its membership varies. It has close to 500 members describing themselves as specialists in General Practice rather than Dermatology. The membership is drawn from countries with significant differences in the incidence of melanoma. Both of these factors can be argued to impact on the results of this study. It seems quite possible that the readers who outperformed the machine were predominantly from high incidence melanoma countries with specialist dermatologist qualifications. Such a finding would be important to allow the medical community to assess the place in the diagnostic pathway of this system. I would propose that the authors present the data analysed taking into account these factors. None declared.