Developing Multi-Modal Neuroimaging Biomarkers for the Detection of Early-Stage Psychosis

Keywords:Psychosis, Neuroimaging, Magnetoencephalography, Machine-Learning, Early Intervention

Project Summary: The project will develop new predictive algorithms for mental health outcomes in young people at high-risk for Schizophrenia (ScZ) that will synthesize expertise from cognitive neuroimaging, clinical psychology and data-analytics. This framework will utilize a unique multi-modal neuroimaging data-set that has been collected as part of the MRC-funded Youth Mental Health Risk and Resilience (YouR) study. The YouR-study is the only study worldwide to employ Magnetoencephalography (MEG) in combination with Magnetic Resonance Imaging (MRI) and Magnetic Resonance Spectroscopy (MRS)-data to develop predictors for ScZ in at-risk individuals. Currently, n=150 participants meeting clinical high-risk (CHR) criteria have been comprehensively assessed with a multi-modal imaging battery in combination with follow-up data for up to three years. In addition, n=40 participants with affective disorders and substance abuse (CHR-negative) as well as n=50 controls have been recruited.

The project will employ cutting-edge machine learning techniques to identify multi-modal biomarkers for the prediction of psychosis in CHR-participants and related mental health outcomes. We expect that the resulting predictors for mental health outcomes in at-risk individuals are significantly more powerful than biomarkers from single modalities. Moreover, we expect to identify subgroups of CHR-participants based on multi-modal neuroimaging data that will reveal insights into pathophysiology processes underlying emerging psychosis that will likely be important for the development of novel psychological and pharmacological therapies.

Project Team: The project team consists of experts in neuroimaging, psychosis research and advanced data-analytics. Professor Peter Uhlhaas (primary supervisor) is based at the Institute of Psychology and Neuroscience (INP), University of Glasgow (UofG). He is an international expert in in cognitive neuroscience and ScZ-research with a particular focus on neural oscillations. He has published 80 articles in internationally high-ranking journals (Nature Rev Neuroscience, Neuron, PNAS, Brain). Professor Roderik Murray-Smith is head of the the “Inference, Dynamics and Interaction group” in the School of Computing Science who will contribute his expertise for machine learning tools.  Drs. Fracasso and Ince are both based the INP will contribute their expertise in data-analytics and neuroimaging.