We are searching for a talented PhD candidate to work on a translational project bridging the gap between machine learning, neuroscience and clinical psychiatry. The project focuses on developing advanced statistical and machine learning methods for the analysis of clinical neuroimaging data. The project will be hosted jointly between the Cognitive Neuroscience and Psychiatry departments at the Donders Institute for Brain, Cognition and Behavior at Radboud University Medical Center under the supervision of Dr. Andre Marquand and Dr. Eric Ruhé.
The successful candidate will be embedded within the Statistical Imaging Neuroscience research group, headed by Prof. Christian Beckmann and the Stress-Related Disorders research theme at Radboudumc as well as being part of a wider academic research environment at the Donders Institute.
The primary focus of this project is to develop statistical and machine learning approaches for understanding inter-individual variation in biomarkers derived from magnetic resonance imaging (MRI) data, then use these biomarkers to predict resilience and risk in recurrent major depressive disorder (rMDD). You will employ and extend the normative modelling ('brain growth charting') approaches we have pioneered for this purpose which rely on state-of-the art Bayesian machine learning and deep learning techniques. You will first develop the statistical machinery necessary to apply these methods to 'big data' neuroimaging samples derived from tens of thousands of multi-modal MRI scans (functional MRI, structural MRI and connectivity) e.g. the ABCD study, the Healthy Brain Network and the UK Biobank (N>40k).
You will then apply these models to richly phenotyped clinical samples of patients with rMDD (not depressed at the time of scanning) derived from samples acquired in the Netherlands and via our international collaborations (e.g. with Dr. Roland Zahn at King's College London). The goal is to make predictions about the future course of the disorder (e.g. which subjects will remain in remission and which subjects are likely to have a future depressive episode). The project is highly interdisciplinary, has a clear translational focus and integrates machine learning and statistics with cognitive neuroimaging and clinical neuroscience.
Tasks and responsibilities
- Discuss, plan and perform research in a stimulating environment.
- Develop statistical approaches for data analysis from fundamental principles.
- Apply these statistical models to large-scale population cohorts.
- Interpret findings in the light of clinical knowledge.
- Publish findings in peer-reviewed journals and present at international scientific conferences.
- Produce software tools to enable for the use of the wider scientific community.
- Finalize PhD training and project within the four year contract .
- Work in an interdisciplinary team of international scientists.