Abstract
Glaucoma diagnosis, management and research depend on complex judgments of the optic disc (or optic nerve head), visual field and intraocular pressure. The current standard of optic disc evaluation requires qualitative observer judgments of stereoscopic photographs of the optic disc, a less than optimal method. Despite much research, no methods have yet conclusively improved over this conventional Approach: Contemporary optic disc analyzers typically use instrument-specific image capture methods and derive quantitative estimates for various anatomical features of the optic disc. Our goal is to improve the methods of optic disc diagnosis by applying advanced image analysis methods from computer engineering to the essential diagnostic problem in glaucoma - detecting change or stability in optic disc images over time. Expertise at the University of Pennsylvania in clinical glaucoma, translational research (R. Stone, PI; E. Miller, J. Piltz-Seymour and others) and biostatistics (M. Maguire, G.-S. Ying) is merged with engineering expertise in computer image analysis at Sarnoff Corporation (B. Hanna, H. Sawhney, and others) in a Bioengineering Research Partnership with four Specific Aims: 1) Develop and validate robust registration algorithms for automatic alignment of optic disc images; 2) Develop an automated multiple-view analysis approach to extract relative, local change parameters from optic disc stereo images; 3) Develop interactive tools to assist in observer grading of optic disc images and in clinical interpretation of the automated change detection and stereoscopic algorithms; and 4) Conduct initial validation studies of the optic disc change detection tools. Our plan to address stereo primarily differs from other approaches to optic nerve analysis, but it offers many advantages for validation, clinical care and research not possible with the instrument-specific formats of contemporary fundus analyzers. Requiring only a personal computer and software to analyze optic disc images, our approach is clinically intuitive, can accommodate improvements in software and camera technology, is compatible with many image formats, permits use of archived fundus photos and is cost-effective. The refined approach to stereo recovery will permit robust detection of optic disc stability or change, and it offers great promise for advancing optic nerve diagnosis in glaucoma.
Agency: National Institute of Health (NIH)
Institute: National Eye Institute (NEI)
Type: Research Project (R01)
Project #: 1R01EY017299-01A1
Project Start: 2007-09-30
Project End: 2012-08-31
- Designed and developed novel algorithms for 3D alignment and change detection from fundus images captured at intervals of several years for early Glaucoma diagnosis.
- Designed and developed novel algorithms for 3D optic disc reconstruction from uncalibrated, highly degenerate stereo pair of fundus imagery.
- Designed and developed novel algorithms for vessel detection and localization in small field-of-view fundus imagery.
- Guided development of algorithms for interactive vessel extraction in premature fundus imagery for early Retinopathy of Prematurity (RoP) diagnosis.