Comparing the number-needed-to-biopsy ratio for melanoma diagnosis by artificial intelligence as a medical device, teledermatology and face-to-face models of care

To the best of our knowledge, this is the first UK study to compare the number-needed-to-biopsy (NNB) ratio for melanoma diagnosis across three different models of care, incorporating artificial intelligence as a medical device (AIaMD) alongside teledermatology. We present encouraging results from our centre, with NNB ratios that are comparable among all three models and lower than the average NNB ratio reported in a worldwide meta-analysis. We highlight the use of AIaMD in the correct identification of benign lesions and encourage further work to improve and validate performance, thereby maintaining patient safety. Finally, we consider a hybrid model as a potential strategy to streamline current pathways and address the increasing demands on urgent skin cancer services.