Artificial Intelligence Holds Promise for Earlier Glaucoma Detection

Using deep-learning techniques that encompassed hundreds of raw optical coherence tomography (OCT) images of healthy and glaucomatous eyes, IBM Research, in collaboration with New York University, conducted a study to determine if visual function can be accurately measured directly from structures in the eye imaged using noninvasive techniques. The researchers, who presented their findings at the recent Association for Research in Vision and Ophthalmology meeting, discovered that data obtained with retina imaging can help to assess the presence of glaucoma. The study was able to accurately estimate the visual field index (VFI) from a single 3-dimensional raw OCT image of the optic nerve with Pearson correlation of 0.88. Capturing that metric with artificial intelligence could lay the groundwork for future technologies that can potentially use this analysis to quickly estimate a patient’s visual function. This could give practitioners access to precise information — without the need for multiple and time-intensive tests — when gathering data for a glaucoma diagnosis.

Conventional OCT structural measurements, such as retinal nerve fiber layer thickness and ganglion cell inner plexiform layer thickness, could not achieve this degree of accuracy, despite both layers being known target locations of glaucoma. This deep-learning study suggests the structural measurements captured by OCT contain information highly correlated with functional measurements.

Another important challenge of glaucoma cited by the researchers is its rate of progression, which requires the careful analysis of data from multiple visits. They addressed this issue using machine learning, which has shown that visual function results at future visits could be accurately forecast. The ability to do this could one day help practitioners to better predict the progression and onset of the disease and adjust treatments accordingly.