Reliability and agreement testing of a new automated measurement method to determine extent of facial vitiligo using standardized UV images and a dedicated algorithm.

Quentin Marin Dit BertoudClémence BertoldKhaled EzzedineAmit G PandyaMarie CherelAlejandro Castillo MartinezMarie-Anne SeguyMarwa AbdallahJung Min BaeMarkus BöhmDavinder ParsadDavid RosmarinAlbert WolkerstorferPhilippe BahadoranManon BlaisePierre-Michel DugourdValérie PhilippoJean-Michel DelavalThierry Passeron

Published in: The British journal of dermatology (2023)

This study demonstrates that the Vitil-IA algorithm provides reliable assessment of facial involvement in vitiligo. It also underlines the limitations of the F-VASI score when performed by non-experts for homogeneous vitiligo depigmentation, and in all raters when depigmentation is heterogeneous.

Keyphrases

deep learning
machine learning
convolutional neural network
soft tissue
neural network