CEREBRIS Research Associate (Generative AI for Multimodal Neuroimaging) at the University of Bath

CEREBRIS is a European Innovation Council (EIC) Pathfinder Open 2025 project involving an innovative, data-driven approach to managing neurological diseases. It aims to develop a secure, federated and explainable ecosystem- starting with stroke care. Its AI models will be designed from multiple types of patient data- such as brain scans, motion patterns and brain signals- to help doctors understand and predict more objectively how an individual is progressing and will recover after a brain injury. Using privacy-preserving technologies, the system will allow hospitals and clinicians to work together without ever sharing raw data, ensuring trust and compliance with global regulations.

Driven by a multidisciplinary European consortium with deep expertise in neurology, AI, neurotechnology, and clinical translation, CEREBRIS will transform the entire stroke pathway- from early diagnosis to recovery. By delivering precise diagnostics and targeted rehabilitation, CEREBRIS will significantly reduce long term disability, substantially reduce healthcare expenses, and set a new standard for brain health globally.

This is a research-intensive role, with the expectation of significant contributions to research excellence, translational medicine, and innovation.

Working alongside Prof Michael Yang, Prof Damien Coyle and the CEREBRIS team associated with the Bath Institute for the Augmented Human and Centre for Spatial Intelligence, the Research Associate will develop novel artificial intelligence methods for the analysis and integration of multimodal biomedical data, as described in the CEREBRIS work plan. The research will focus on developing advanced generative and foundation AI models that learn shared representations from neuroimaging, wearable physiological signals, movement analysis and clinical data to improve the understanding, assessment and prediction of post-stroke motor function and recovery.

The postholder will design, implement and evaluate state-of-the-art machine learning algorithms for multimodal data fusion, representation learning and predictive modelling, including the development of scalable AI pipelines and synthetic data generation approaches. The role will also contribute to privacy-preserving and federated learning methodologies to enable secure analysis of distributed healthcare datasets and support the translation of AI technologies into clinically relevant tools for stroke prognosis and personalised rehabilitation.

The post-holder will be expected to publish the results of the research in high impact journal publications (e.g. T-PAMI, IJCV) and top-tier AI conferences (e.g. NeurIPS, ICLR, CVPR, ICCV, ECCV). Experience of AI research (e.g. diffusion models, agentic AI), developing and applying technologies including deep learning using platforms such as Keras, Pytorch and/or experience of optimising the performance of software for high-performance computing on GPU clusters is requirement for this role.