Neuroimaging of visual perception

Visual perception is initiated by retinal stimulation and guided by the brain*s
existing knowledge about the visual world * known as prior assumptions. Our
visual cortex contains several complete maps of our visual world (more than
16), each with a distinct representation of our visual environment. At early
stages representations will correspond to the physical retinal stimulations.
But at later stages our prior assumptions will intervene to give rise to
representations that match the perceptual qualities of the visual scene. This
leads to our main research question: How do low-level sensory stimulation and
high-level prior assumptions interact to give rise to visual perception?

We will use functional Magnetic Resonance Imaging (fMRI) to address this
question. These fMRI signals are closely related to neuronal signals.
Conventional fMRI data-analyses primarily tests hypotheses about the
localization of certain functions (*activation*). Computational data-analyses
techniques, on the other hand, allow additional neuronal properties to be
extracted from the same fMRI data, including visual field maps, receptive
fields and the reconstruction of the representation of each map. This leads to
our secondary research aims; to extend and validate novel computational
data-analysis techniques and to apply these techniques in subjects with visual
impairments to reveal any potential diagnostic uses.

The main objective is to describe the interactions between physical sensory
stimulations and high-level prior assumptions that give rise to visual
perception. In two secondary objectives, we will advance novel model-based
neuroimaging data-analysis methods and apply these novel data-analysis
techniques in visually impaired subjects.

Using functional MRI, subjects are scanned on while viewing various visual
stimuli, ranging from synthetic stimuli (e.g. checkerboards) to photos and
movies. The main focus will be on novel ways to analyse the data in order to
derive new neuronal properties from the fMRI signal. We will compare how these
properties vary across different visual field maps, with image and perceptual
qualities and across subject populations.

Scanning will take approximately 45-60 min in total per session for each
subject. Functional MRI is a non-invasive technique, so there is no need for
special preparation for the subject. There are no known risks associated with
functional MRI acquisition. The data are used for research purposes only.
However, severe abnormalities may be noticed, in which case a specialist
(radiologist) may be asked for advice, upon decision of the research team. If
the specialist confirms that medical treatment is indicated, then the subject
will be notified. Besides financial remuneration, there are no benefits from
participation in this study for the subjects.

The measurements of eye movements will take about 1 hour. There are no risks
associated with eye movement recordings.