The Value of Annual Glaucoma Screening for High-Risk Adults Ages 60 to 80

Artificial intelligence at the national eye institute

PURPOSE OF REVIEW

This review highlights the artificial intelligence, machine learning, and deep learning initiatives supported by the National Institutes of Health (NIH) and the National Eye Institute (NEI) and calls attention to activities and goals defined in the NEI Strategic Plan as well as opportunities for future activities and breakthroughs in ophthalmology.

RECENT FINDINGS

Ophthalmology is at the forefront of artificial intelligence-based innovations in biomedical research that may lead to improvement in early detection and surveillance of ocular disease, prediction of progression, and improved quality of life. Technological advances have ushered in an era where unprecedented amounts of information can be linked that enable scientific discovery. However, there remains an unmet need to collect, harmonize, and share data in a machine actionable manner. Similarly, there is a need to ensure that efforts promote health and research equity by expanding diversity in the data and workforce.

SUMMARY

The NIH/NEI has supported the development artificial intelligence-based innovations to advance biomedical research. The NIH/NEI has defined activities to achieve these goals in the NIH Strategic Plan for Data Science and the NEI Strategic Plan and have spearheaded initiatives to facilitate research in these areas.

Structure-function models for estimating retinal ganglion cell count using steady-state pattern electroretinography and optical coherence tomography in glaucoma suspects and preperimetric glaucoma: an electrophysiological pilot study

PURPOSE

To derive and validate structure-function models for estimating retinal ganglion cell (RGC) count using optical coherence tomography (OCT) and steady-state pattern electroretinography (ssPERG) parameters in glaucoma suspects (GS) and preperimetric glaucoma (PPG).

METHODS

In this prospective cross-sectional study, 25 subjects (50 eyes) were recruited at the Manhattan Eye, Ear, and Throat Hospital. Subjects underwent comprehensive eye examinations, OCT, standard automated perimetry (SAP), and ssPERG testing. Eyes were divided into three groups based on the Global Glaucoma Staging System: healthy (N = 30), GS (N = 10), and PPG (N = 10) eyes. The combined structure-function index (CSFI), which estimates retinal ganglion cell count (eRGC_CSFI) from SAP and OCT parameters, was calculated in each study subject. Two prediction formulas were derived using a generalized linear mixed model (GLMM) to predict eRGC_CSFI from ssPERG parameters, age, and average retinal nerve fiber layer thickness (ARNFLT) in 30 eyes selected at random (training group). GLMM predicted values were cross-validated with the remaining 20 eyes (validation group).

RESULTS

The ARNFLT, ssPERG parameters magnitude (Mag) and magnitudeD (MagD), and eRGC_CSFI were significantly different among study groups (ANOVA p ≤ 0.001). Pearson correlations demonstrated significant associations among ARNFLT, ssPERG parameters, and eRGC_CSFI (r² ≥ 0.31, p < 0.001). Two GLMMs predicted eRGC_CSFI from Mag (eRGC_Mag) and MagD (eRGC_MagD), respectively, with significant equations (F(3,18), F(3,19) ≥  58.37, R² = 0.90, p < 0.001). eRGC_Mag and eRGC_MagD in the validation group (R² = 0.89) correlated with eRGC_CSFI similarly to the training group. Multivariate pairwise comparisons revealed that eRGC_Mag and eRGC_MagD distinguished between healthy, GS, and PPG eyes (p ≤ 0.035), whereas independent Mag, MagD, and ARNFLT measures did not distinguish between GS and PPG eyes.

CONCLUSION

This pilot study offers the first combined structure-function models for estimating RGC count using ssPERG parameters. RGC counts estimated with these models were generalizable, strongly associated with CSFI estimates, and performed better than individual ssPERG and OCT measures in distinguishing healthy, GS, and PPG eyes.