Hypo-priors in autistic visual perception

A recent account of autism proposes that autistic perception is characterized
by a reduced integration of sensory processing with prior world knowledge:
so-called *hypo-priors*. In the context of visual processing, for example, this
means that input to the sensory cortices of autistic individuals is less
influenced by prior knowledge and experience than it is within healthy
individuals. Previous research has shown that perceptual expectations strongly
modulate activity in the early sensory cortices of healthy individuals. In
visual perception, these modulations reduce overall V1 activity and yet
'sharpen' the representation of expected stimuli, leading to a more efficient
representation of expected stimuli compared to unexpected stimuli. However, in
the case of hypo-priors, the modulation of sensory input by expectation should
not occur.

We will evoke perceptual expectations in and record brain activity using
functional magnetic resonance imaging (fMRI) and magnetoencephalography (MEG).
Using pattern analysis techniques on the fMRI, we can investigate the
representational content of brain activity and thus estimate what content
visual cortex is representing during the task. Then, by investigating the
neural osciallations collected by MEG, we can analyze how these representations
and the effect of expectations on them are implemented by the brain's
physiology.

The knowledge obtained here will be informative about brain functioning in
autism spectrum disorder. Moreover, it could be used to develop and inform new
treatment approaches in the future.

The purpose of this study is to test whether perceptual expectations modulate
V1 activity consistent with *hypo-priors* in autistic individuals.

Cross-sectional, observational study with two to three sessions. Participants
first complete questionnaires to assess their IQ and autistic traits and are
familiarized with the MRI-scanner. During the second session, participants
complete a visual task in the MRI-scanner. During a third session on a second
day, the visual tasks are repeated in the MEG scanner.

This study is entirely safe. The first session involves completing standardized
behavioural questionnaires. The second session involves performing a visual
task in the MRI-scanner for approximately one hour. Using standard safety
procedures, the risk posed by MRI to the participant can be considered
negligible. MEG, which is used in the third session, does not pose any risk to
subjects. Moreover, the Donders Institute has vast experience with neuroimaging
studies within patient groups.

Although there is no immediate therapeutic or clinical benefit of this study,
it will result in a more detailed understanding of autistic brain functioning
and the knowledge obtained can be used in future research on diagnostic markers
and treatment approachesfor ASD.


This study is group related and can therefore not be done in an adult
population. Although ASD can be reliably diagnosed in adulthood, it is foremost
a developmental disorder that is characterized by an onset early in life.
Sensory abnormalities are most striking in childhood and adolescence and less
pronounced in adulthood. This research is group-related to children and
adolescents in so far that we aim to obtain knowledge that is relevant for the
development of new treatment approaches for young patients, in whom possible
treatment would be most effective. Results from studies involving adult
participants would not generalize well to the population we target. It is
therefore crucial to study abnormal sensory behaviors and their neural
substrates at the stage where these are most typical.