Mathematics and Machine Learning of Brain Ageing

About the Project

How do brains grow old and stay healthy, or suffer from diseases such as Alzheimers? Biologists uncovered changes ranging fromthe molecular and cellular level all the way to the whole-brain connectome. This PhD will fuse multiscale mathematics with modern ML to link cellular features (e.g., organelle shapes and function) to network-level dynamics (including higher-order dynamics) and build interpretable multi-scale brain-age models. You’ll develop theory (PDEs, dynamical systems, topology), and algorithms (hypergraph/higher-order connectomics, GNNs, Bayesian ML) to integrate microscopy/omics readouts with fMRI/MEG connectomes and test whether cellular-scale features forecast macro-scale ageing signatures and resilience.

You’ll work on:

• Multiscale modelling that ties metabolism and organelle morphology to tissue features and ageing phenotypes.
• Higher-order connectomics and connectome fingerprinting to track ageing and Alzheimer’s heterogeneity.
• Deployable, interpretable brain-age predictors with uncertainty—validated on CHBH and public datasets.

You are: quantitative (maths/CS/physics or similar), curious, and eager to move between theory, code, and data. You are passionate about collaborating with experimental or clinical groups.

Funding Notes

A scholarship for home students may be available. It is awarded in competition with other projects so if you are interested in a scholarship please contact f.spill@bham.ac.uk as soon as possible.

Strong applicants with external funding will also be considered.