APPROVE · Autistic Perception and the Predictive Role of Visual Experience

Autistic Perception and the Predictive Role of Visual Experience

How we make sense of the huge number of things we see around us? We move our eyes at least 3 times per second and the brain is constantly bombarded with very different information. How do we combine this information to get a stable and reliable representation of the visual world we are experiencing? “The autistic world” can be quite different from ours, and right now little is known on how socio-communicative difficulties and behavioral stereotypes in adulthood arise from childhood. One of the main challenges of the brain is to distinguish between external movements and movements that come from our own eye-movements. How to achieve a stable representation of the world despite the high-speed motion on the retina due to own eye movements? The sensorimotor system must act against this self-produced stimulus to prevent the perception of disturbing motion for every eye-movement. That means that the brain must have a precise idea of where the eyes are before and after an eye movement. An efference copy is a copy of the motor command that is sent to the sensory areas which makes a prediction about on where the eye will be after an eye movement, for example a saccade. This signal is then confronted to the actual feedback provided from the motor areas. If the predicted signal and the actual signal never match, perception should encounter problems in maintain stability. A constant lack of match between prediction and sensory feedback we hypothesized could lead over time to the build time of weaker prior experiences. Within this framework, recent theories propose that individuals with ASD may demonstrate a reduced reliance on prior information when making perceptual judgments. The Bayesian perspective suggests that the predictive system deficits in autism stem from a diminished Bayesian prior. From this perspective, the brain unconsciously infers information about the world by applying Bayes rule: it assesses the probability of a prediction given sensory data, considering the likelihood of sensory data given that prediction and the Bayesian prior. The Bayesian prior is essentially a probabilistic distribution representing the expectation of the environment being in a particular state before any observations are made. Formal objectives of this Marie Skłodowska Curie Action (MSCA) have been to (a) to test the hypothesis that efference copy processing is weakened in ASD; (b) to test the idea that weak efference copy processing explains the inability to integrate contextual information in ASD; and (c) to test the efference copy hypothesis in a realistic environment.

Data: CORDIS, © European Union