This article has been reviewed according to Science X's editorial process and policies. Editors have highlighted the following attributes while ensuring the content's credibility:

Autonomous artificial intelligence (AI) in urgent skin cancer pathways can increase clinical capacity for dermatologists, according to a study presented at the annual meeting of the European Academy of Dermatology and Venereology, held from Sept. 30 to Oct. 3 in Vienna.

Lucy Thomas, M.B.Ch.B., from Imperial College London, and colleagues implemented class III artificial intelligence as a medical device (AIaMD) across two hospital sites. Sequential validation was applied (dermoscopic confirmation and image quality assessment), followed by hierarchical classification prioritizing higher-risk diagnoses.

Cases classified as benign were discharged autonomously, without clinician review, while high-risk cases were triaged for assessment by a teledermatologist (TD). To assess diagnostic performance, outcomes were compared with face-to-face (F2F), TD-only and AI-assisted TD pathways.

Overall, 8,391 patients were managed via the autonomous pathway; the consent rate for autonomous decision-making was 86%. The researchers found that autonomous AIaMD discharge rates were 37% and 32% at sites 1 and 2, respectively, with actual discharge rates of 31% and 25%, respectively, after exclusions; TD discharged an additional 24% and 25%, respectively.

Routine follow-up was reduced with autonomous AIaMD versus standard TD (12% versus 27%), while the biopsy rate was reduced compared with conventional F2F care (26% to 29% versus 43%). False negatives were rare; no adverse outcomes were identified within available follow-up. Compared with the F2F pathways, estimated clinician time savings were 2,851 hours, representing a 62% gain in clinical capacity.

Source link