An Automatic AI-Based Algorithm That Grades the Scalp Surface Exfoliating Process From Video Imaging: Application to Dandruff Severity and Its Validation on Subjects of Different Ages and Ethnicities

The study evaluates a new AI-based imaging device that grades scalp exfoliation, using dandruff as a model. The device, equipped with a camera and various lighting conditions, was trained on 3,600 images from 234 subjects of different ages and ethnicities. In the validation phase, 460 images from 192 subjects were analyzed, showing a high correlation (0.952; p < 0.001) between expert assessments and device gradings, with a Mean Average Error of 0.16 grading units. The device's accuracy varied slightly with ethnic background and phototypes. The study concludes that this device offers a valid, rapid, and easy method for assessing scalp exfoliation and could aid dermatologists in diagnosing scalp disorders, with potential applications in other cosmetic areas.