Predictive processing in the retina through evaluation of the omitted-stimulus response

Samantha I. Fradkin
While previous studies have demonstrated that individuals with schizophrenia demonstrate predictive coding abnormalities in high-level vision, it is unclear whether impairments exist in low-level predictive processing within the disorder. Evaluation of the omitted-stimulus response (OSR), i.e., activity following the omission of a light flash subsequent to a repetitive stimulus, has been examined previously to assess prediction within retinal activity. Given that little research has focused on
more » ... he OSR in humans, the present study investigated if predictive processing could be detected at the retinal level within a healthy human sample, and whether this activity was associated with high-level predictive processing. Flash electroretinography (fERG) was recorded while eighteen healthy control participants viewed a series of consecutive light flashes within a 1.96 Hz single-flash condition with a flash luminance of 85 Td · s, as well as a 28.3 Hz flicker condition with a flash luminance of 16 Td · s. Participants also completed the Ebbinghaus task, a context sensitivity task that assesses high-level predictive processing, and the Audio-Visual Abnormalities Questionnaire (AVAQ), which measures frequency of self-reported auditory and visual sensory distortions. For both conditions, within-group analyses were conducted to compare fERG amplitude and implicit time measurements following present-stimulus trials with those following omitted-stimulus trials. Additionally, mean omitted-stimulus waveforms reflecting averaged retinal responses across all subjects were examined for presence of an OSR. Results demonstrated the absence of an OSR within the 1.96 Hz condition and the presence of activity in response to omitted stimuli within the 28.3 Hz flicker condition that could represent an OSR. The amplitude of the OSR in the flicker condition was significantly correlated with the number of flicker repetitions prior to the first omission (rs=.57, p=.02), supporting the conclusion that this activity was at least partially predic [...]
doi:10.7282/t3-7r72-w857 fatcat:z6zovwfyhbbspjc4a3peuhynn4