FREE Virtual Study Fair | 1 - 2 March | REGISTER NOW FREE Virtual Study Fair | 1 - 2 March | REGISTER NOW

New methods for enhanced brain activity mapping through multi-modal data-fusion and deep learning

   Cardiff School of Psychology

This project is no longer listed on and may not be available.

Click here to search for PhD studentship opportunities
  Dr Marco Palombo, Prof David Marshall, Prof S Rushton, Prof Derek Jones  No more applications being accepted  Funded PhD Project (European/UK Students Only)

About the Project

The School of Psychology and the School of Computer Science and Informatics at Cardiff University are delighted to offer this joint fully funded EPSRC studentships starting in April 2023.

 Project Summary:

Characterizing the human brain’s function and structure is crucial to understand its physiology in health and pathology.

 There are complementary ways of measuring brain activity: magnetoencephalography (MEG), electroencephalography (EEG), and functional magnetic resonance imaging (fMRI). The first two are characterised by high temporal but poor spatial resolution, the latter by low temporal but high spatial resolution. Thus, fusing MEG/EEG and fMRI would clearly be advantageous, providing both high temporal and spatial resolution. Furthermore, structural information derived from diffusion MRI (dMRI) via tractography methods can be coupled with brain activity to allow detailed modelling of brain function.

This PhD will focus on the challenge of fusing the complementary information from MEG, fMRI and dMRI to predict the human brain physiological activity with unprecedented spatial and temporal resolution. Towards this goal, the PhD student will collect a unique dataset by measuring brain activity while the same participants watch the same movies in both MRI and MEG scanners. When observers view film clips their brains show correlated responses in specific areas, making the information-fusion easier. 

The PhD student will

1.      Analyse an existing MEG/fMRI dataset we already collected to identify features of brain activity using independent component analysis (ICA) and similar techniques (e.g. principal component analysis). 

2.      Use standard correlational approach to match these features across modalities and establish a baseline.

3.      Develop the initial fusion super-resolution model (FSM), MEG/fMRI.

4.      Investigate more sophisticated FSM based on temporal deep learning (e.g. Transformers, RNN, LSTM)

5.      Collect (or participate in collection of) first-of-its-kind dataset(s) using state-of-the-art Connectom dMRI, 7T fMRI and MEG data.

6.      Extend the initial FSM by integrating information on structural connectivity from dMRI

From the outset, following initial guidance, and with oversight from the supervisors, the student will be encouraged to organise and lead the supervisory meeting, setting the schedule and agenda. During the PhD, the student will be encouraged and supported to take on responsibility for the direction of the PhD and the lead in paper writing. 

The student will develop technical knowledge and skills in machine learning, brain imaging and neuroscience; complemented by personal, organisational and vocational skills training provided by the doctoral academy, host schools and supervisors. 

Towards the second year of the PhD, the student will have the opportunity to take a 3-6 months industrial secondment at AInostics ( This will provide an opportunity to learn about how science works outside of an academic setting, learn new skills, make contacts and explore a potential career option.

Funding Notes

Full awards are open to UK Nationals and EU students who can satisfy UK residency requirements (students must have been in the UK for >3 years before start of course).
Search Suggestions
Search suggestions

Based on your current searches we recommend the following search filters.

PhD saved successfully
View saved PhDs