Dermatology ·
DEXI dermoscopy score identifies 93.4% of cancerous lesions in Mayo Clinic study
Listen to this post
read by AI voices · all episodesTHE SHORT READ
In a quiet week for dermatology innovation, a retrospective Mayo Clinic study of 261 lesions found that the DEXI dermoscopy AI score identified 93.4% of cancerous lesions at a threshold of 3.0. Dr. Jenny Link, the study's lead author, said the tool requires further validation and refinement.
THE DETAILS
- AI Oct 23, 2026In the retrospective study, published in Mayo Clinic Proceedings, researchers analyzed 261 lesions with a DEXI score and a tissue-confirmed diagnosis from 178 patients who had total-body photography. DEXI, a deep learning tool trained on more than 66,000 annotated dermoscopy images, scores lesions from 0.0 to 10.0, and at a threshold of 3.0 it correctly identified 93.4% of cancerous and 75% of noncancerous lesions. The tool was less effective for severely atypical nevi and early evolving melanomas, and Dr. Jenny Link, a Mayo Clinic dermatologist, said it should not be used in isolation.
The weekly innovation roundup for dermatology, published every Sunday for the week before; every item links to its sources. Latest dermatology news · All specialties