7. Project Summary Abstract Glaucoma is one of the leading causes of preventable blindness, and currently available treatments are not sufficient to halt progression in many patients. While much has been learned about the biology of glaucoma, development of new forms of treatment has been stymied by three barriers: high between-subject variability in ganglion cell number in normal eyes, high within-subject variability for perimetry in patients with glaucoma, and the slow rate of progression of the disease. The proposed research integrates neural modeling and clinical research to develop improved methods for diagnosing glaucoma and for assessing progression towards blindness. The results are intended to improve measures for both clinical trials and ongoing patient care, while at the same time improving basic science understanding of the pathophysiology of glaucoma and providing guidance for biological studies of the disease process. Innovative uses of clinical devices will guide testing with custom systems, and statistical analyses will utilize the synergy between structural and functional measures of glaucomatous damage. High-resolution retinal imaging of retinal nerve fiber layer (RNFL) will be performed on patients with glaucoma using a custom advanced adaptive optics scanning laser ophthalmoscope (AOSLO) as well as custom use of spectral domain optical coherence tomography (SD-OCT). High-resolution perimetry will be performed in corresponding regions of the visual field, using custom stimuli that are resistant to optical artifacts that affect conventional perimetry. Specific Aim 1 will determine the role of ganglion cell dysfunction in between-subject differences in the amount of visual loss corresponding to clinically observed reflectance defects, and assess longitudinal changes. Specific Aim 2 will utilize the synergy between structural and functional measures to develop perimetric algorithms that interact dynamically with information about RNFL structure. Specific Aim 3 will develop methods that dramatically improve the ability to detect wedge defects that are poorly sampled by conventional perimetry.