Quantification and Classification of Aqueous and Vitreous Inflammation in Uveitis Using Deep Learning Analysis of Ultrahigh-Resolution OCT

NIH RePORTER · NIH · R21 · $247,800 · view on reporter.nih.gov ↗

Abstract

PROJECT SUMMARY Uveitis is a common cause of blindness in the United States. The specific cause of uveitis is often unknown, making it difficult to implement targeted testing and treatment. By harnessing the power of artificial intelligence (AI) and ultrahigh-resolution optical coherence tomography (OCT), we will develop novel tools to help diagnose and classify uveitis by characterizing the distribution and type of inflammatory response within the eye. OCT will be used to image multiple areas in the aqueous and vitreous media of the eye and count the density of inflammatory cells. AI analysis will further classify cell types to aid in differential diagnosis. Protein levels in the media will be estimated based on the intensity of background scattering. Together, the cell count and protein level can be used to monitor the patient’s disease severity and response to treatment. The precision of OCT is ideal for clinical trials, where precise outcome measures can accelerate drug development and save costs. To further these goals, specific aims of the project are: (1) Develop ultrahigh-resolution OCT for imaging the anterior and vitreous media of the eye. (2) Quantify and classify inflammation in the aqueous and vitreous humor of uveitis patients. We will develop OCT-based biomarkers to aid in the differential diagnosis of uveitis and assessment of disease severity.

Key facts

NIH application ID
10952558
Project number
1R21EY036563-01
Recipient
OREGON HEALTH & SCIENCE UNIVERSITY
Principal Investigator
Yan Li
Activity code
R21
Funding institute
NIH
Fiscal year
2024
Award amount
$247,800
Award type
1
Project period
2024-09-01 → 2026-08-31