Reading

About 3.7 million Americans are visually disabled. Of these, about 200,000 are blind, and the rest have low vision. Our broad goal is to use the principles of vision science to understand the reading difficulties of people with low vision. A major goal of the current research plan is to focus on the adaptability of the low-vision reader as a method for rehabilitation. 

Our recent research has shown that the concept of the visual span provides a key to understanding reading deficits. The visual span is the number of letters on a line of text that are identified reliably without moving the eyes during reading. Decreasing acuity in peripheral vision results in a small visual span, even for normal vision. This small visual span is further diminished as a consequence of eye disease. The visual span shrinks in size in peripheral vision, a key factor in explaining the severe reading deficits of people with central-field loss. We have developed empirical methods for measuring spatial and temporal characteristics of the visual span, and a computational model (Mr. Chips) for studying its role in eye-movement behavior in reading. 

In a recent study, we have shown an expansion of the visual span in peripheral vision through training, with a corresponding increase in reading speed. In Series 1 of our current research plan, we are exploring the potential of perceptual learning (training) for enlarging the visual span and for improving reading speed in low vision. This research will include study of the roles of spatial attention and positional coding in determining the size of the visual span. We will also measure the growth of the visual span in children as they learn to read. 

Recent evidence for adult brain plasticity in sensory systems makes it likely that reorganization of visual cortex plays a role in the adaptation of people with low vision. In Series 2, we are conducting experiments with functional magnetic resonance imaging (fMRI) to examine cortical reorganization and contrast coding in low vision, and to examine the possibility of tactile processing in the visual cortex of people with low vision. 

In a third series, we are conducting secondary data analyses of two large sets of low-vision reading data (MNREAD test), from well-characterized groups with cataract and macular scotomas. The goals are to 1) assess the relationships between measures of reading performance and traditional measures of visual impairment, controlling for non-visual factors such as cognitive status; 2) Develop a new reading ability scale; and 3) compare the patterns found in #1 and #2 for the cataract and macular-scotoma groups.