Written by: Dr Norman Nieder-Heitmann, Ophthalmologist from Paarl Eye Centre
MBChB (Stell), Dip Ophth (SA),
FC Ophth (SA), MMED Ophth
(UCT)
Diabetes mellitus (DM) refers to a condition where the body is unable to produce or respond to the hormone insulin. This in turn, leads to impaired metabolism of carbohydrates and ultimately raised blood sugar. Raised blood glucose affects virtually every organ in the human body, the eyes being no exception. This global epidemic has significant financial implications, accounting for approximately 10% of most countries’ health budgets.
Diabetic retinopathy (DR) is the most common cause of poor vision in diabetics and without timely diagnosis and treatment, it may lead to irreversible blindness. This microvascular disease has a prevalence of 30 – 40% amongst diabetic patients of which 10% is vision threatening.
With approximately 6000 ophthalmologists and more than 400 000 000 diabetics world-wide, timeous diagnosis, surveillance and treatment of diabetic retinopathy becomes a virtually impossible task. This is even more evident in middle to low-income countries where resources are limited.
Artificial intelligence (AI) has become a valuable screening tool in overcoming the above problem. AI is defined as the ability of computer systems to perform complex, independent tasks that require human-like intelligence, such as visual processing, speech recognition, or decision making.
In the case of DR, large image datasets of retinal photographs are used to detect and flag the presence of disease. It acts as a triage system to determine which patients with urgent disease should be seen preferentially by a specialist, while those with nonurgent needs may wait for a routine appointment. In the absence of symptoms and early signs of retinopathy, the patients can be scheduled for annual screening.
Not only does AI increase patient access to DR screening, but it also significantly reduces the cost and burden of unnecessary doctor consultations. With the onset of ultra-widefield retinal cameras that’s able to take clear pictures through an undiluted pupil, the screening process only takes a few minutes. Once the images are captured and interpreted by an AI program, they are remotely reviewed by an ophthalmologist.
DR is a treatable disease if detected early. With the help of technology, screening programs will hopefully be ramped up in the near future.
Paarl Eye Centre:
Tel 021 871 1368
20 Bo Lady Grey Street
Paarl, 7646
www.paarleyecentre.co.za
