An artificial intelligence (AI) system developed at the University of Yamanashi, Japan, can now help medical professionals assess the severity of Psoriasis. The study was published online in the form of a short report in the Journal of the European Academy of Dermatology and Venereology.
Takashi Okamoto and their colleagues from the University of Yamanashi have created a simplified Psoriasis Area and Severity Index (PASI) system (Single-Shot PASI) by associating AI models capable of assessing psoriasis severity.
During the model development, the team used 705 psoriasis images of the front and back of patients. 10 images were used to validate the deep learning system.
Thirteen board-certified dermatologists and nine medical students scored the test sets without any assistance from AI. Subsequently, they referred to the scores of AI and then reevaluated their own previous scores.
With the assistance of artificial intelligence, standard deviation among evaluators and mean absolute differences from AI scores were significantly reduced.
The Single-Shot PASI system decreases the pressure of scoring psoriasis severity for dermatologists. Even when the AI scores from the Single-Shot PASI system are not used directly, referencing them later reduces deviations in the evaluation between dermatologists.
Moreover, the AI application and Single-Shot PASI system can be an instant tool to check psoriasis severity for patients with Psoriasis objectively. The Single-Shot PASI system is hoped to be used by many dermatologists and patients to address Psoriasis.