SCH: Multimodal Retina Image Alignment and Applications

NIH RePORTER · NIH · R01 · $273,134 · view on reporter.nih.gov ↗

Abstract

PROJECT SUMMARY (See instructions): Overview: Progress in the characterization and treatment of retinal disease involves epidemiological and natural history studies which include genetic and environmental risk factor evaluation as well as clinical trials. As treatments advance, it is important to be able to scientifically analyze and interpret a large amount of information that can be procured from different areas and even points on the retina and evaluate retinal structure and function over time and in response to therapies. Currently there is a proliferation of technology to provide data and this data comes from many instruments and from different companies. It is increasingly difficult for any one person or reading center to evaluate this information. The goal of this proposal is to develop deep-learning based multi modal retinal image processing methods to help the ophthalmologist to quickly detect and diagnose diseases. Intellectual Merit: As treatments advance, it is important to be able to scientifically analyze and interpret a large amount of information that can be procured from different areas and even points on the retina and evaluate retinal structure and function over time and in response to therapies. Currently there is a proliferation of imaging technology producing images from many instruments and from different companies. It is increasingly difficult for any one person or reading center to reliably review the multiple types of imaging available in a patient with a retinal disease nor overlay these on each other to properly analyze retinal structure and function and determine correlations and predictive value of these tests to clinical outcomes and covariates like sex, age, race and concurrent medications. The ability to co-localize all of this data and use artificial intelligence (Al) to help with these analytics will advance the field of understanding and treating retinal disease. The objective of this proposal is to develop deep-learning based multimodal retinal image registration methods to help the ophthalmologist to quickly detect and diagnose retinal diseases.

Key facts

NIH application ID
10916222
Project number
5R01EY033847-04
Recipient
UNIVERSITY OF CALIFORNIA, SAN DIEGO
Principal Investigator
Cheolhong An
Activity code
R01
Funding institute
NIH
Fiscal year
2024
Award amount
$273,134
Award type
5
Project period
2021-09-30 → 2025-08-31