Meta menu:

From here, you can access the Emergencies page, Contact Us page, Accessibility Settings, Language Selection, and Search page.

Open Menu

Research Projects - Visual Perception Lab

You are here:

Bistable perception and predictive coding

Perception aims at identifying the causes of sensory stimuli. However, the sensory data available for this inferential process are inherently noisy and ambiguous. According to the notion of “predictive coding”, the brain therefore uses prior beliefs to enable an unambiguous and stable perceptual representation of the world. These prior beliefs are constantly updated in accordance with continuous changes of the statistical properties of the environment. In this project, we investigate these inferential processes using the intriguing phenomenon of bistable perception, which is typically evoked by perceptually ambiguous visual information such as reversible figures, ambiguous motion stimuli, or binocular rivalry. We use behavioural experiments, fMRI, TMS, and computational modelling to examine the inferential neural processes underlying bistable perception, with a particular focus on the role of prefrontal cortex. Moreover, within the context of a predictive-account of schizophrenia we use bistable phenomena to investigate alterations of perceptual inference in relation to psychotic symptoms such as delusions and hallucinations.

This project is funded by Berlin Institute of Health (Clinician Scientist Programme), Berlin School of Mind and Brain and DFG.

The role of prior beliefs in perceptual decisions

Perceptual decision-making is influenced by prior beliefs, as they determine the selection of sensory data. Within the framework of predictive coding, such selective sampling of sensory data can be modelled as Bayesian inference, through which the states of the world are inferred from the sensory data, using prior beliefs. In this project, we investigate the neuronal mechanisms underlying the influence of prior beliefs on selective sampling, and the modulation of these processes by higher-level beliefs regarding environmental volatility. Moreover, within the context of a predictive-account of schizophrenia we investigate the relationship between alterations in selective sampling mechanisms and psychotic symptoms such as delusions and hallucination.

This project is funded by DFG.

The role of dopamine in the emergence of perceptual abnormalities in psychosis

This project aims to characterize the role dopamine in the emergence of perceptual alterations that typically occur in psychotic states. We are using pharmacological modulation of the dopaminergic system to examine how increases and decreases in dopamine availability affect visual perception of noisy stimuli and its neural correlates using fMRI. We hope to gain a refined understanding of the relationship between specific transmitter changes and psychotic symptoms, which will be a prerequisite for the development of individualized and symptom-oriented therapeutic interventions.

This project is funded by BMBF and Berlin Institute of Health (Clinician Scientist Programme).

The influence of external feedback on perceptual inference

According to the 'Bayesian Brain Hypothesis', our perception is governed by probabilistic inferences, whereby learned prior beliefs are used to infer the states of the world from the data registered by our sensory organs. This project uses behavioural and fMRI experiments to explore how such perceptual inference is influenced by external feedback regarding our perceptual decisions. In particular, we focus on the question how the reliability of external feedback information affects the weighting of sensory data vs. prior beliefs in perceptual inference.

This project is funded by DFG.

Reducing stigma by explaining psychosis – a new therapeutic model of schizophrenia

The diagnosis of schizophrenia is still highly stigmatizing. This is likely due to the severity of the disorder and its chronicity, but may also be related to the symptoms characteristic of schizophrenia, including so-called positive symptoms such as delusions and hallucinations and negative symptoms such as affective flattening and apathy. The goal of this project is to develop a new therapeutic model that bridges the explanatory gap between neurobiological findings and subjective experience in psychosis, thereby helping to further the acceptance of psychotic disorders by depathologizing and destigmatizing its symptoms. Rather than investigating the mechanisms underlying the disorder, this project focusses on the development of a plausible and easy-to-explain model that is indeed based on the latest scientific developments but is intended for practical therapeutic use.

This project is funded by Berlin Institute of Health (Clinical Fellow Programme).

Confidence-based learning (lead by Dr. Matthias Guggenmos)

Reinforcement learning can explain fundamental aspects of human learning and behaviour, and in addition, how deficient learning mechanisms may give rise to dysfunctional behaviour and psychiatric disorders. A key limitation of this theory, however, is that many forms of learning occur in the absence of external feedback. Moreover, important aspects of psychiatric disorders are characterized by self-reinforcing processes. In this project, we investigate both at the behavioural and neural levels the role of subjective confidence in driving learning in the absence of feedback. Furthermore, individual differences in such confidence-based learning are related to symptoms of depression, such as anhedonia.

This project is funded by DFG.