Artificial intelligence. To some people, this term may evoke images of the android Data from Star Trek, or Skynet from the Terminator franchise. While we have a long way to go before machines are sentient, researchers are developing useful applications of artificial intelligence (AI) that can help people right now. For sufferers of chronic blinding diseases like glaucoma, AI is being adapted to both diagnosis and disease management. In fact, before long, a solution could be as close as your cell phone.
Making AI Work for Glaucoma Patients
Intelligent computers require two things to work: a processing algorithm, which is a step-by-step procedure to solve a problem or reach a decision, and a whole lot of the right data to run through that algorithm. Put in a medical context, with enough useful data to compare with a specific patient’s data, the processing algorithm can reliably arrive at an accurate diagnosis. Because the process a physician goes through to come up with an accurate diagnosis is complex and very hard to properly model, AI can be used to contain costs, screen, determine risk, and help doctors make decisions.
An example of using AI for eye disease can be found in China, where researchers launched a program to study the application of AI to congenital cataracts. Congenital cataracts (CC) is a rare disease that causes irreversible vision loss, and breakthroughs related to CC have contributed substantially to medical science. The researchers developed a three-fold AI system called CC-Cruiser, which included:
- Identification networks for screening for CC in populations
- Evaluation networks for risk stratification among CC patients
- Strategist networks to assist in treatment decisions by ophthalmologists
CC-Cruiser was both accurate and efficient in providing comprehensive evaluations of disease severity (lens opacity) with respect to three different indices (opacity area, density and location) for risk stratification, as well as a reference for treatment decisions for CC patients, while the strategist networks provide the final treatment decision (surgery or follow-up) on the basis of results from both the identification networks and the evaluation networks.
Going Forward
Proof that the concept behind AI is sound has finally led some players in the industry to look at developing systems that can be brought to the marketplace. One such company, VISULYTIX, presented at GRF’s Glaucoma 360 event to introduce its new technology: an inexpensive smartphone clip-on optic nerve scanner that they say will help diagnose glaucoma. This is exciting news, especially when thinking about how to diagnose and monitor glaucoma in developing countries that may not have access to doctors or expensive medical equipment.
Of course, this early in the game, it is best to be cautious. One day, AI will be able to do some amazing things for glaucoma sufferers, especially in the areas of diagnosis and disease management, but it will take more work. For now, though, there is no replacement for the sound professional judgment of an experienced doctor who knows you and your eyes.
Reference: Nature Biomedical Engineering 1, Article number: 0024 (2017) doi:10.1038/s41551-016-0024
Posted on March 10, 2017; Reviewed on April 4, 2022
