While doctors can identify genetic disorders with an accuracy between 70% and 75%, DeepGestalt—an AI project by FDNA—can do so with accuracy of 90% and over. By analyzing the face piece-by-piece, the system finds a corresponding disorder for that crop, moves on to the next and then pools the results to find the most proper diagnosis for the patient. For the 8% of the population that’s affected by rare genetic disorders, DeepGestalt is optimism for simpler identification (and treatment, when possible). Read more at Ars Technica.
AI Diagnoses Genetic Disorders Using Facial Photos
