MexicoHealthcare and life sciences2022Jalisco Health Services (OPD)
RETINA-IA: AI screening for diabetic retinopathy
Jalisco piloted AI retina screening in three clinics, with pre-diagnosis and referral in one to two days
Problem
Diabetic retinopathy can be treated if caught early, but screening in Jalisco's public clinics depended on scarce eye specialists.
Approach
Patients were called in, and their retinas were photographed in three clinics of the state health services. A deep-learning model graded each image. A team from the Jalisco College of Ophthalmologists checked the AI's pre-diagnoses, and patients who needed it were referred to specialists. The Inter-American Development Bank published a case study of the pilot, whose results were presented in August 2022.
Results
Takeaway
AI grading followed by ophthalmologist review allowed pre-diagnosis and referral within one to two days. The IDB reported good classification at each severity level but gave no accuracy figures or patient counts.
Categories
- Size
- Large enterprise
- Industry
- Healthcare and life sciences
- Business function
- Production and quality
- Innovation type
- Process innovationProduct and service innovation
- AI technology
- Computer vision
- Adoption
- Co-developed with a partner
- Evidence
- Independent study
Sources
Source summaries
The IDB reports on the RETINA-IA pilot in three Jalisco clinics. It describes the workflow of active patient recruitment, retina photos, deep-learning grading and ophthalmologist checks. It reports pre-diagnosis and referral within one to two days and says the model classified each retinopathy level well, without giving numbers.
Full details are in the original sources.
