Department
The School of Computer Science and Electronic Engineering has a strong, internationally recognised research profile spanning artificial intelligence, machine learning and biomedical signal processing. The School hosts active research in neurotechnology and brain-computer interfacing, and provides access to substantial computing resources, including a GPU-enabled high-performance computing cluster. This post sits within PaiNeuro, an interdisciplinary project working towards objective pain measurement using neurophysiological signals, in collaboration with Vestre Viken Hospital Trust in Norway.
Duties of the Role
Working under the direction of the Principal Investigator, Dr Sebastian Halder, the post holder will carry out the advanced computational analysis for the project, developing and evaluating machine-learning methods that distinguish pain-related from pain-free states in clinical EEG, with particular emphasis on models that generalise to previously unseen participants. Specifically, the post holder will:
- Develop and maintain reproducible EEG preprocessing, artefact-handling and feature-extraction workflows.
- Implement and compare deep-learning approaches for EEG-based pain prediction, including architectures that capture spatial and temporal structure in the data.
- Develop Riemannian geometry-based methods using EEG covariance representations.
- Investigate cross-participant generalisation, including transfer learning, domain alignment and one-class classification approaches based on pain-free EEG.
- Validate models rigorously using held-out participant data and report performance using appropriate classification metrics.
- Contribute to scientific publications, open-source software, data documentation and the public release of code and curated data.
- Work closely with the Principal Investigator and the clinical collaborators in Norway, contributing to regular interdisciplinary project meetings.
Skills and qualifications required
Applicants will have, or be close to completing, a PhD in computer science, electronic engineering, biomedical engineering, neuroscience, machine learning or a closely related discipline. The following are essential:
- Strong Python programming skills and experience with scientific-computing and machine-learning libraries.
- Experience applying machine learning, including deep learning, to EEG or other high-dimensional biosignals.
- A sound understanding of robust model evaluation, including participant-level cross-validation and testing on independent data.
- The ability to work independently while contributing effectively to a multidisciplinary research team.
- Strong academic writing and spoken communication skills.
The following are desirable:
- Experience with deep-learning architectures for time series or neural signals, such as convolutional or recurrent networks.
- Experience with Riemannian geometry methods for covariance-based signal classification.
- Knowledge of transfer learning, domain adaptation or one-class classification and anomaly detection.
- Familiarity with tools such as MNE-Python, PyTorch, Braindecode or pyRiemann.
- Experience working with clinical data, research governance and open, reproducible research practices.
Please use the 'Apply' button to read further information about this role including the full job description and person specification which outlines the full duties, skills, qualifications and experience needed for this role. You will also find details of how to make your application here.
Our website http://www.essex.ac.uk contains more information about the University of Essex. If you have a disability and would like information in a different format, please email [contact details available in the full listing].
£38,784 per annum
