Deep Learning to Discriminate Arteritic From Nonarteritic Ischemic Optic Neuropathy on Color Images.
AI model distinguishes arteritic from nonarteritic anterior ischemic optic neuropathy on eye photos.
Deep Learning to Discriminate Arteritic From Nonarteritic Ischemic Optic Neuropathy on Color Images.
IMPORTANCE: Prompt and accurate diagnosis of arteritic anterior ischemic optic neuropathy (arteritic anterior ischemic optic neuropathy) from giant cell arteritis and other systemic vasculitis can contribute to preventing irreversible vision loss from these conditions.
To develop, train, and test a deep learning system (deep learning system) to discriminate arteritic anterior ischemic optic neuropathy from nonarteritic anterior ischemic optic neuropathy on color fundus images during the acute phase.
DESIGN, SETTING, AND PARTICIPANTS: This was an international study including color fundus images of 961 eyes of 802 patients with confirmed arteritic anterior ischemic optic neuropathy and nonarteritic anterior ischemic optic neuropathy.
AI correctly told apart the two causes of sudden optic nerve vision loss in 92.6% of cases.
In the training and validation sets, 374 (54.9%) of patients were female, 301 (44.2%) were male, and 6 (0.9%) were of unknown sex; the median (range) age was 66 (23-96) years.
The accuracy of the 2 experts for classification of the same dataset was 74.3% (95% CI, 66.7-81.9) and 81.6% (95% CI, 74.8-88.4), respectively.
A deep learning system showing disease-specific averaged class-activation maps had greater than 90% accuracy at discriminating between acute arteritic anterior ischemic optic neuropathy from nonarteritic anterior ischemic optic neuropathy on color fundus images, at the eye level, without any clinical or biomarker information.
A deep learning system that identifies arteritic anterior ischemic optic neuropathy could improve clinical decision-making, potentially reducing the risk of misdiagnosis and improving patient outcomes.