Nonlinear System Identification and Deep Learning for Neurodegenerative Disease Detection at Coventry University

Neurodegenerative diseases, such as Alzheimer’s disease (AD) and Parkinson’s disease (PD), have a profound global impact, affecting millions of people and placing a significant economic burden on society. To address these challenges, early and accurate diagnosis, as well as understanding disease progression, are crucial for effective treatment and improving the quality of life for patients. However, current diagnostic methods rely on expensive and time-consuming mental status examinations and neuroimaging scans, which can be prone to inaccuracies. 

There is an urgent need for cost-effective and precise diagnostic tools that enable early detection of neurodegenerative diseases at the individual level. Electroencephalography (EEG) has emerged as a promising non-invasive and economical alternative for studying these diseases. However, current approaches primarily focus on linear single-channel or pairwise connectivity analysis, often limited to specific brain regions. It is increasingly evident that neurodegenerative diseases impact a wide network of brain regions, and these interconnections exhibit nonlinear patterns. Therefore, to enhance diagnostic accuracy, it is crucial to explore how neural activity is coordinated across different spatial and temporal scales. 

In this PhD project, our goal is to advance the field by developing novel techniques that integrate nonlinear signal processing, systems engineering principles, and cutting-edge deep learning methods, including graph neural networks. These approaches will enable us to capture the spatial and temporal features present in EEG signals and effectively characterize neurodegenerative diseases. Building upon the existing work of the principal investigator’s research groups, we aim to push the boundaries of understanding and diagnosing these diseases, ultimately leading to improved patient outcomes and healthcare practices. 

Application Details:

Successful candidates will have at least a minimum of a 2:1 first degree in Mathematics/Statistics, Computer Science, Engineering, Computational Neuroscience with a minimum 60% mark in the project element or equivalent with a minimum 60% overall module average. Candidates are expected to have strong competent programming skills (in Matlab, Python, R or Julia) and experienced in mathematics/statistics and numerical analysis.

To find out more about the project please contact:  Dr Fei He at fei.he@coventry.ac.uk.