Modelling circular inference, from the general population to psychotic experience

Research project description

PhD will investigate discovering computational markers of vertical and horizontal propagation of belief in perception, and their possible dysregulation in psychosis spectrum disorders such as schizophrenia (SZ).  

Vertical propagation of belief between low/sensory and high/associative cortical areas was shown to be affected in SZ, leading to circular inference (“see what you expect to see”) that may be at the origin of the psychotic experience (Jardi et al Nature Communications, 2017). In this project, we want to provide new computational markers for horizontal propagation of belief (i.e. within the same cortical area) which may also lead to circular inference and measure their possible association with traits associated to the schizophrenia spectrum. 

The project will build on a novel theoretical framework developed in the lab explaining how perception is shaped by belief propagation between populations of neurons coding for different parts of the visual field. In particular, this framework explains the emergence of bistable perception as a typical case of circular inference. We would first like to confirm preliminary experimental evidence and validate this framework in a series of novel visual, auditory and linguistic tasks. Computational markers of horizontal and vertical belief propagation will be obtained by fitting our computational model to the behavior of each participant. We will also look for signatures of this different types of information propagation in neural data collected in EEG/MEG. Finally, we will look for correlations between psychosis spectrum traits of each individual and the model-inferred markers of belief propagation across participants. 

This ambitious transdisciplinary project will benefit from the complementary expertise of the partnering teams in computational neuroscience, modelling of behavioural data, neuroimaging, individual differences, and clinical psychology. 

Academic background / Skills

Candidates must hold a degree that allows admission to the official doctoral programme at UAB.  

Additional requirements for a stronger application are: 

• MSc or equivalent in Cognitive Science, Applied Mathematics, Computer Science or a related field. 
• Proficiency in English at a minimum B2 level. 
• Interest in learning computational science methods. 
• Strong communication skills to effectively disseminate research results to diverse audiences. 

Other skills:  

• Ability to work in a team, critical thinking, and a commitment to diversity and inclusion. 
• Theoretical background in Cognitive psychology, Statistics and Computer Science. 

Research group/s description

The computational neuroscience unit at the CRM is made up of the groups of three Principal Investigators (Alex Roxin, Klaus Wimmer and Alex Hyafil). The unit is an active member of a larger, Barcelona-wide Neuroscience community which includes theoretical, experimental and clinical groups. Broadly speaking, we investigate how large assemblies of interacting neurons give rise to animal and human behaviour. Our approach combines computational modeling, behavioural experiments, neuroimaging and data analysis. 

The Person-Environment Interaction in Risk and Resilience for Mental Health integrates clinical practitioners from public mental health centers in Barcelona and world-leading clinical, developmental and genetics researchers. Our approach is characterized by a multi-level approach, integrating the study of person-factors, environmental exposures and genetic variability that moderates the impact of environmental factors on the person.