Harnessing Artificial Intelligence to Improve Glaucoma Clinical Trials

The aim of this project is to use artificial intelligence tools to increase the efficiency of clinical trials for glaucoma.

Clinical trials for glaucoma can be difficult because of burdens related to ongoing eye testing, limitations of tools used to follow disease progression, and participant selection. Jithin Yohannan, MD, MPH, and his colleagues plan to use artificial intelligence (AI) approaches to make trials more efficient.

Participants in glaucoma trials must be followed using repeated vision testing, which can be burdensome over a long period and yield variable results. Dr. Yohannan and his team plan to use AI tools to screen for patients who are a good fit for these repeated tests and at high risk for disease progression. The tools are expected to be able to make accurate predictions at the first patient exam.

The researchers also want to expand outcomes in these trials beyond results on these visual tests. They plan to develop further AI tools that can use data from eye exam imaging to predict the trajectory of vision change over time, extending trial outcomes beyond vision testing.

The AI models we develop will predict which patients will most likely get worse over time AND be good at taking tests to identify this worsening. Additionally, our AI models will use 3D scans of the optic nerve head to predict changes in visual function over time.

Dr. Yohannan and his team predict that their findings will make clinical trials of new glaucoma therapies faster and less costly, translating into faster assessment and approval of candidate treatments for glaucoma.