How do we perceive the world?

Perception is not a passive process of registering the world.
Rather, the brain can make use of prior knowledge to actively predict possible future states.
This can help to make sense of the incoming sensory signals and make quick and accurate decisions.
The Predictive Brain Lab sets out to understand how our brain uses prior knowledge to aid perception and decision-making processes.

Anticipation of future events

We showed that early visual cortex performs pattern completion by recreating a stimulus sequence after only a subset of the visual sequence is provided. Interestingly, this ‘pre-played’ sequence was temporally faster compared to the actual sequence and linked to behavioral performance during perception.

Layered fMRI reveals feedforward and feedback processing

Ultra-high resolution fMRI allows for functional data to be acquired from separate cortical layers. Using knowledge of the layer-specific terminations of feedforward and feedback connections in visual cortex, we use this to separate top-down from bottom-up visual responses. For example, increased fMRI responses to an illusory shape stimulus from top-down feedback are only found in the deep layers of V1.

Expectations enhance perception

Here, we combined human magnetoencephalography (MEG) with multivariate decoding techniques to resolve the timing of neural signals induced by expectations. We found that, when auditory cues induced the expectation of a grating of a particular orientation, this evoked a neural template of the expected grating already before the grating was presented. These results suggest a mechanism by which the brain predicts future sensory stimuli in order to enhance perception.

Latest news

Paper published on familiarity facilitating perceptual performance by increasing visual processing speed

Jan 16, 2020 | Publications

Most things in life we have seen before. Familiarity with a stimulus leads to an attenuated neural response to the stimulus. Alongside this attenuation, recent studies have also observed a truncation of stimulus-evoked activity for familiar visual input. One proposed...

Paper published on neural mechanisms for implicit context learning

Jan 2, 2020 | Publications

Humans are remarkably good at knowing where to expect relevant stuff, given a particular context. For example, we intuitively know that we should expect a computer mouse next to the keyboard, and below the monitor. It has been claimed that we can learn these patterns...

Paper published in Journal of Vision

Nov 27, 2019 | Publications

Our everyday decisions about what we see are often based on ambiguous and unstable sensory input. For instance, we make about three eye movements per second, greatly shifting the visual information that meets our eyes. Yet, despite such instabilities in sensory input,...

Tweets by predictivebrain

How to get to the lab

Donders Institute

Address

Floris de Lange, PhD
Donders Institute (Centre for Cognitive Neuroimaging), Room 01.117
Kapittelweg 29
P.O. Box 9101
NL – 6500 HB Nijmegen
the Netherlands

Phone: +31 24 36 10658
Fax: +31 24 36 10989
e-mail: floris.delange@donders.ru.nl