Describe teleophthalmology and the emerging technologies in eye care.

Teleophthalmology

• The problem technology is being asked to solve is specific: in India the burden is rural while the expertise is urban, and no realistic programme of training will move enough ophthalmologists into villages. Teleophthalmology moves the image instead of the patient or the specialist
• Ophthalmology is unusually well suited to it, because so much of the diagnosis is visual and can be captured as an image — the anterior segment, the fundus, and increasingly optical coherence tomography
• The two modes — store-and-forward, in which images and data are captured locally and read later by a specialist elsewhere, which is cheap, needs no simultaneous connection and suits screening; and real-time video consultation, which permits interaction and suits triage, follow-up and specialist advice
• Established applications — diabetic retinopathy screening, which is the most successful and best-validated use; retinopathy OF prematurity screening with wide-field imaging, which is transformative because trained ROP screeners are extremely scarce and the window is measured in weeks; vision centre support, allowing an ophthalmic assistant to obtain a specialist opinion without referring the patient; glaucoma disc assessment; postoperative follow-up; and teaching and second opinions
• Requirements — adequate image quality, which is the usual limiting factor; trained local personnel to capture the images; reliable connectivity and power; protocols defining who reads and how quickly; and, critically, a referral and treatment pathway for those found
• Limitations that must be acknowledged — it cannot replace a slit-lamp examination, tonometry or gonioscopy; ungradable images are common with cataract or small pupils; medicolegal responsibility for a remote opinion must be defined; and data protection and consent apply as they would to any record

Artificial Intelligence and Other Developments

• Artificial intelligence has advanced furthest in ophthalmology of any specialty, for the same reason as teleophthalmology: large labelled image datasets with a clear outcome
• Validated and approved applications — diabetic retinopathy grading from fundus photographs, where several systems have regulatory approval and perform comparably to human graders; retinopathy of prematurity; glaucoma disc assessment; and age-related macular degeneration on optical coherence tomography
• Why it fits India particularly well — the bottleneck is trained graders, not cameras, and grading is exactly what these systems do; and the volume of screening required is beyond any feasible human workforce

One book of nineteen in the KAVACH series · mbbsadda.in