Abstract
In macular degeneration, the primary cause of vision loss among older adults, photoreceptor death in the retina results in diverse patterns of vision impairment. Complex visual tasks such as reading or navigation require both basic visual sensation (here, ‘low level vision’) as well as neural processes beyond basic sensation (here, ‘high-level vision’). Patients with similar retinal damage can differ widely in their performance on complex visual tasks, suggesting that compensation for this impairment varies between patients. Quantification of this compensation is a necessary step to understanding the neural mechanisms underlying compensatory visual strategies. Traditional tests like visual acuity and contrast sensitivity gauge only limited aspects of vision and may not capture the broader visual field, leading to discrepancies between test outcomes and real-world function. To bridge this gap, we developed a method using outcomes from the Macular Integrity Assessment (MAIA), a microperimetry method that evaluates sensitivity across the retina. Quantitatively comparing MAIA results across patients has been a challenge, especially given that lesions in central vision lead to worse impairment than peripheral vision. Here we introduce a method to quantitatively account for that difference, using the concept of the “Cortical Magnification Factor” (CMF). Different parts of visual cortex correspond to distinct regions of vision (retinotopic maps), and the CMF describes how much more cortex is devoted to each portion of the visual field. By weighting MAIA scores with CMF, we derived a measure called retinal functional health (RFH). RFH reliably reflects the clinical impression of the severity of a scotoma. RFH was significantly correlated to contrast sensitivity, as well as acuity. Further, models incorporating RFH were better predictors of high-level visual processing (aggregate performance on a range of complex visual tasks). These results validate our measure of RFH to compare scotoma severity across participants.