Smartphone AI system may help detect glaucoma
Click Here to Manage Email Alerts
Key takeaways:
- An AI system detected definite glaucoma with higher sensitivity than vision center doctors in India.
- AI may help reduce the burden on glaucoma specialists.
An artificial intelligence system helped screen patients for glaucoma using a smartphone-based fundus camera, according to research presented at the American Glaucoma Society meeting.
According to Swati Upadhyaya, MD, the global supply of ophthalmologists is not sufficient to tackle the burden of glaucoma, and the problem is particularly evident in low-income areas. Some local vision centers in India have implemented an AI screening method to help reduce that burden.
“This is an offline AI and does not require the internet,” Upadhyaya said. “It gives a referral report within 10 seconds. It can be taken anywhere, done by any person, and the learning curve is very short.”
Upadhyaya and colleagues conducted a prospective study to compare the diagnostic accuracy of AI glaucoma detection vs. image grading by primary eye care center graders via telemedicine.
The study comprised 570 eyes of 299 participants. Upadhyaya said the AI algorithm had a sensitivity of 73.17% and specificity of 98.6% in detecting referrable glaucoma against the vision center doctors.
The sensitivity of AI to pick up definite glaucoma was 96.4% compared with 53.5% for vision center doctors when compared against a glaucoma specialist.
“Remidio Fundus on Phone with offline AI for glaucoma can be used for mass screening at remote places, schools, community halls and workplaces,” Upadhyaya said. “AI can help in reducing unnecessary referrals, and the burden of glaucoma specialists will reduce. They will be able to concentrate more on the real glaucoma cases who need expert advice and treatment.”