Greater use of AI may lead to advances in glaucoma research
A partnership between artificial intelligence systems and clinicians may lead to mutual advancements in glaucoma research and clinical practice, according to a review published in the British Journal of Ophthalmology.
“Given that glaucomatous changes to the optic nerve head tissues are irreversible, timely and reliable structural and functional evaluation of the eye could help in the early diagnosis of glaucoma,” Sripad Krishna Devalla, PhD, department of biomedical engineering at the national University of Singapore, and colleagues wrote. “In recent years, AI-based systems have started to revolutionize the health care industry. ... A number of AI studies have been proposed for the diagnosis and management of glaucoma based on the interpretation of functional and/or structural information of the eye.”
In a review summarizing the role of AI in glaucoma, researchers reported that modern systems have successfully increased the speed and diagnostic power of ocular imaging modalities through data collation, abnormality detection and referral relevance. Further, AI has the potential to reduce the workload for clinicians while minimizing diagnostic errors and cost.
While the many benefits of AI allude to a need for broader use, many challenges in its clinical translation still exist due to the subjective nature of technology and the quality of training data. Researchers further note the AI requirements for clinical acceptance are indefinite, which poses an issue for regulatory approval. Further clinical studies comparing AI to traditional diagnostic methods are needed to better understand its capabilities.
“Clinicians in the future may expect a plethora of AI tools to assist them in the day-to-day diagnosis and management of glaucoma,” Devalla and colleagues concluded. “While the persistence of new clinical and technical challenges is undeniable, one cannot dismiss the ways in which AI could positively impact ophthalmic research and clinical practice in glaucoma.”
Perspective
Back to Top
The application of AI algorithms in health care is becoming increasingly more prevalent. Machine learning, in its various forms, can be applied to the management of many different conditions. Can glaucoma be counted among these conditions? Quite possibly, yes. In general terms, AI can be used to help detect the presence of a condition or to help determine the severity of a known condition. It can also often assess a disease state over time for stability or progression. All of these things can be potentially helpful in glaucoma.
As it is applied to glaucoma, AI typically evaluates optic nerve photographs, OCT measurements or visual field tests. It can also evaluate two or more of those tests in combination, thus improving reliability. In all circumstances, the performance of AI algorithms is highly dependent upon the quality of the data that they are supplied with. The idiom “garbage in, garbage out” most definitely applies.
There are programs that can be used to help differentiate glaucomatous from nonglaucomatous eyes and programs that can help to detect progression of glaucoma over time. There are many potential applications for such technology.
For example, in areas where health care resources are scarce and difficult to access, screening platforms can be developed to evaluate large groups of people for the presence of glaucoma. Resources can then be applied to those who are most in need. In more developed areas, AI could potentially be utilized to more definitively determine if progression has or has not occurred. In any event, AI would not be used to replace practitioners, but would instead be used to allow practitioners to be more efficient and more effective.
Collapse
Read more about