Flicker photophobia as an experience of inefficient coding

Project Summary A canonical feature of migraine is visual discomfort (i.e., “photophobia”), with particular sensitivity to flicker (time-varying modulations of light). We lack a mechanistic understanding of this symptom generally, and specifically require a framework that unites the phenomenon of discomfort with the properties of the visual

environment, perception, and neural response. Such a synthesis may be offered by recent work in information- optimal representation. Computationally “efficient” representations represent the statistical structure of the environment and maximize sensory information storage. Recent research in experimental psychology has

shown that these “efficient coding” models account for aspects of human sensory judgments and the properties of neural activity. Importantly, these models explain how changes in the statistics of the visual environment to lead to changes in perception. Our project is motivated by the idea that photophobia is an experience of

“inefficient” information processing. Over three Aims we will apply the efficient coding framework to understand the properties of flicker exposure, perception, and neural representation in typical observers and in people with migraine and photophobia. Using personal light-logging devices, we will test the idea that people with migraine

and photophobia experience a systematically different visual world. Using psychophysical and discomfort measures we will test for the effects of stimulus properties upon flicker perception, and for differences between people with migraine and headache free controls. Finally, we will examine the neural representation of visual

flicker using functional MRI to test for the signature of efficient coding in distributed neural responses.