Poster F94, Tuesday, March 27, 8:00-10:00 am, Exhibit Hall C
Deficit of Prediction Ability as A Potential Cause of Phantom Noise in Autism Spectrum Disorder
Jyh-Jong Hsieh1, Yukie Nagai2, Minoru Asada1; 1Osaka University, 2National Institute of Information and Communications Technology
Previous studies report that some people with autism spectrum disorder (ASD) suffer from phantom noise in the visual and/or auditory perception, especially when the environmental stimuli suddenly change. However, the underlying mechanism is still unclear. We propose a hypothesis that the deficit of prediction ability might be a cause of phantom noise because the stochastic resonance in the brain produces stronger noise to compensate the information loss caused by the impaired prediction. We developed a computational model to verify our hypothesis in the acoustic perception. This model adopted a feedback control to determine the optimal noise level that maximize the information content originally exists in the sensory signal. We suggest that the predictor in this model employs the information from preceding signals, other sensory modalities, and/or prior knowledge to recognize incoming signals. We manipulated time window of referred preceding signal to examine its effect on hyper neural responses. Experiments first demonstrated a core benefit of stochastic resonance on the detection of subthreshold stimuli when the time window was long enough. By contrast, when the time window was too short, overly intense noise was induced, which might be perceived as phantom noise. This result supports our hypothesis that the phantom noise could occur with impaired prediction from the preceding signal.
Topic Area: PERCEPTION & ACTION: Multisensory