Frontiers in ophthalmology · 2025
Synergistic AI-resident approach achieves superior diagnostic accuracy in tertiary ophthalmic care for glaucoma and retinal disease.
СтатьяНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Camacho-García-Formentí Dalia, Baylón-Vázquez Gabriela, Arriozola-Rodríguez Karen, Avalos-Ramirez Enrique, Hartleben-Matkin Curt, Valdez-Flores Hugo, Hodelin-Fuentes Damaris, Noriega Alejandro
Аннотация
INTRODUCTION: Artificial intelligence (AI) shows promise in ophthalmology, but its potential in tertiary care settings in Latin America remains understudied. We present a Mexican AI-powered screening tool and evaluate it against first-year ophthalmology residents in a tertiary care setting in Mexico City.
METHODS: We analyzed data from 435 adult patients undergoing their first ophthalmic evaluation using an AI-based platform and first-year ophthalmology residents. The platform employs an Inception V3-based multi-output classification model with 512 × 512 input resolution to capture small lesions when detecting retinal disease. To evaluate glaucoma suspects, the system uses U-Net models that segment the optic disc and cup to calculate cup-to-disc ratio (CDR) from their vertical heights. The AI and resident evaluations were compared with expert annotations for retinal disease, CDR measurements, and glaucoma suspect classification. In addition, we evaluated a synergistic approach combining AI and resident assessments.
RESULTS: For glaucoma suspect classification, AI outperformed residents in accuracy (88.6% vs. 82.9%, p = 0.016), sensitivity (63.0% vs. 50.0%, p = 0.116), and specificity (94.5% vs. 90.5%, p = 0.062). The synergistic approach achieved a higher sensitivity (80.4%) than ophthalmic residents alone or AI alone (p < 0.001). AI's CDR estimates showed lower mean absolute error (0.056 vs. 0.105, p < 0.001) and higher correlation with expert measurements (r = 0.728 vs. r = 0.538). In the retinal disease assessment, AI demonstrated higher sensitivity (90.1% vs. 63.0% for medium/high risk, p < 0.001) and specificity (95.8% vs. 90.4%, p < 0.001). Furthermore, differences between AI and residents were statistically significant across all metrics. The synergistic approach achieved the highest sensitivity for retinal disease (92.6% for medium/high risk, 100% for high risk).
DISCUSSION: AI outperformed first-year residents in key ophthalmic assessments. The synergistic use of AI and resident assessments showed potential for optimizing diagnostic accuracy, highlighting the value of AI as a supportive tool in ophthalmic practice, especially for early career clinicians.
METHODS: We analyzed data from 435 adult patients undergoing their first ophthalmic evaluation using an AI-based platform and first-year ophthalmology residents. The platform employs an Inception V3-based multi-output classification model with 512 × 512 input resolution to capture small lesions when detecting retinal disease. To evaluate glaucoma suspects, the system uses U-Net models that segment the optic disc and cup to calculate cup-to-disc ratio (CDR) from their vertical heights. The AI and resident evaluations were compared with expert annotations for retinal disease, CDR measurements, and glaucoma suspect classification. In addition, we evaluated a synergistic approach combining AI and resident assessments.
RESULTS: For glaucoma suspect classification, AI outperformed residents in accuracy (88.6% vs. 82.9%, p = 0.016), sensitivity (63.0% vs. 50.0%, p = 0.116), and specificity (94.5% vs. 90.5%, p = 0.062). The synergistic approach achieved a higher sensitivity (80.4%) than ophthalmic residents alone or AI alone (p < 0.001). AI's CDR estimates showed lower mean absolute error (0.056 vs. 0.105, p < 0.001) and higher correlation with expert measurements (r = 0.728 vs. r = 0.538). In the retinal disease assessment, AI demonstrated higher sensitivity (90.1% vs. 63.0% for medium/high risk, p < 0.001) and specificity (95.8% vs. 90.4%, p < 0.001). Furthermore, differences between AI and residents were statistically significant across all metrics. The synergistic approach achieved the highest sensitivity for retinal disease (92.6% for medium/high risk, 100% for high risk).
DISCUSSION: AI outperformed first-year residents in key ophthalmic assessments. The synergistic use of AI and resident assessments showed potential for optimizing diagnostic accuracy, highlighting the value of AI as a supportive tool in ophthalmic practice, especially for early career clinicians.