New AI Tool Helps Dermatologists Assess Skin Lesions for Cancer Risk
- Researchers at the Mayo Clinic have found an artificial intelligence tool called Dermoscopy Explainable Intelligence, or DEXI, that provides dermatologists with numerical scores to help assess whether skin...
- In the retrospective study evaluated by the Mayo Clinic News Network, researchers analyzed 261 skin lesions from 178 patients who underwent total-body photography, sequential digital dermoscopy imaging, and...
- Beyond generating a numerical risk score, DEXI utilizes explainable AI to highlight specific image features—such as asymmetry, border irregularities, and color variations—that contribute to the final assessment.
Researchers at the Mayo Clinic have found an artificial intelligence tool called Dermoscopy Explainable Intelligence, or DEXI, that provides dermatologists with numerical scores to help assess whether skin lesions are cancerous, according to a study published in Mayo Clinic Proceedings. The deep-learning tool is integrated into advanced imaging systems used to capture and track skin lesions over time. It assigns lesions a score ranging from 0.0 to 10.0 based on specific image characteristics linked to cancer risk, according to a study published in Mayo Clinic Proceedings.
DEXI Scores and Cancer Risk Thresholds
In the retrospective study evaluated by the Mayo Clinic News Network, researchers analyzed 261 skin lesions from 178 patients who underwent total-body photography, sequential digital dermoscopy imaging, and subsequent tissue biopsies. Noncancerous lesions recorded an average DEXI score of 2.5. Melanomas averaged a score of 6.0, while nonmelanoma skin cancers averaged 7.8, showing that cancerous lesions receive significantly higher scores than benign ones. Using a threshold of 3.0, where any score of 3.0 or higher is considered positive for potential skin cancer, DEXI correctly identified 93.4 percent of cancerous lesions and 75 percent of noncancerous lesions.
Clinical Application and Scoring Mechanics
Beyond generating a numerical risk score, DEXI utilizes explainable AI to highlight specific image features—such as asymmetry, border irregularities, and color variations—that contribute to the final assessment. These details give clinicians concrete data points to weigh against their own physical evaluations. Prior to the study, I would review a patient’s DEXI score, but I wouldn’t know exactly what to do with the number,
says Jenny Link, M.D., a dermatologist at Mayo Clinic and lead author of the study, as reported by the Mayo Clinic News Network. Now I can confidently use the score as another piece of data to inform my clinical judgment on whether I recommend a biopsy or monitoring.
Current Study Limitations and Validation Needs
Despite its high sensitivity for identifying malignancies, the algorithm carries notable limitations that prevent it from functioning as a standalone diagnostic instrument. The system does not account for critical patient-specific variables, including age, current medications, and personal history of melanoma. Researchers noted that the tool proved less effective when evaluating severely atypical moles and early evolving melanomas. Dr. Link emphasized that the technology requires further validation and refinement before broader clinical adoption, stating that the score must always be interpreted alongside direct clinical findings and individual patient factors rather than in isolation.