Using Artificial Intelligence to Compute Severity of Alopecia Tool Scores

Researchers have developed an artificial intelligence (AI) system to objectively quantify alopecia areata (AA) and assign Severity of Alopecia Tool (SALT) scores, which are traditionally cumbersome and inconsistent. The AI system analyzes four images of a person's head using a deep learning model to produce a scalp segmentation and a heat map indicating hair loss percentage. Trained on 823 images, the system was evaluated on 188 images from 47 subjects, achieving an intraclass correlation coefficient (ICC) of 0.97 with human scores, indicating excellent correlation. The AI's performance was consistent across different skin tones and hair lengths, though discrepancies arose in severe cases or with new, white, short hair. The study supports AI's potential in SALT scoring, with future plans for a 3D model to enhance accuracy.