Visual Crowding

Summary Crowding is a perceptual phenomenon in which a clearly discernible stimulus becomes unrecognizable because of nearby `distractor' stimuli. Crowding affects core aspects of visual processing, including feature integration, scene perception and reading. Because crowding is stronger in non-foveal vision, it strongly affects individuals with central vision loss (e.g. macular degeneration), with substantial consequences for their quality of life. Crowding has been explored extensively in human psychophysical studies. Functional imaging and electrophysiological studies in humans have provided some neural correlates of crowding in early and midlevel visual cortex, but the neural underpinnings of crowding remain largely unexplored and poorly understood. This project will determine how crowded displays affect the representation of sensory information by neuronal populations in low and midlevel cortex (encoding), how crowding affects the manner in which this sensory information is used to make perceptual judgments (decoding), and whether the effects of crowding can be mitigated by brief periods of sensory experience. In Specific Aim 1, we will record from neuronal populations in primary visual cortex (V1) and V4 of macaque monkeys. We will test the hypothesis that crowding corrupts sensory representations in these areas. We will determine how crowded displays affect neuronal responsivity, tuning, and variability, and interneuronal noise correlations. Making use of recent advances in understanding population codes, we will assess how chances in these response features combine to affect encoding of visual information with crowding. In Specific Aim 2, we will train animals to perform a fine orientation discrimination task, for target stimuli presented in isolation and with distractors. We will use a stimulus paradigm that allows us to compute “psychophysical kernels”—a method for assessing perceptual strategy by which different elements of a visual display are combined to make decisions. We will pair these behavioral measures with recordings of neuronal population in V1 and V4, to infer the read-out strategy used by the animal to relate sensory responses to perceptual decisions. These experiments will test the hypothesis that crowding arises in part from suboptimal read out of sensory information. In Specific Aim 3, we will test whether brief periods of visual adaptation can mitigate crowding. We hypothesize that adaptation can be used to improve the salience of novel stimuli, with heightened salience resulting in improved perceptual performance. We will test our hypothesis by measuring the population information for target stimuli in isolation or with distractors, under control conditions and different adaptation states. We will also test whether adaptation alters the read-out strategy used by animals in our orientation discrimination task. These experiments will reveal which aspects of crowding are most plastic, and test for a novel functional benefit of sensory adaptation. Together our aims will provide a comprehensive investigation of the neural basis of crowding, providing knowledge needed to develop therapeutic and behavioral strategies to alleviate the quality of life issues caused by crowding.