About the project
Biomedical signals such as electroencephalography (EEG) are widely used to human cognition, physiology, and brain-computer interfaces. Yet their temporal structure can unintentionally bias analyses and inflate performance estimates. This project will investigate these effects and develop practical approaches to improve the reliability and reproducibility of biomedical data analysis.
Biomedical recordings contain noise, variability, and temporal dependencies arising from both physiological processes and measurement systems. These characteristics can influence statistical analyses and machine learning models throughout the analysis pipeline, leading to results that may not generalise to real-world applications. Although temporal dependencies are well understood in signal processing, their impact is not always systematically assessed across experimental design, data analysis, and machine learning workflows.
In this project, you will combine signal processing, statistical modelling, and machine learning methods to investigate how temporal structure influences conclusions drawn from biomedical data, with a particular focus on EEG. Using real-world datasets, you will evaluate the effects of temporal leakage on statistical inference and predictive modelling, compare approaches for mitigating these effects, and develop practical guidelines for robust and reproducible analysis. The project offers opportunities to work at the interface of biomedical engineering, data science, and human-centred technologies, with potential collaborations across academia and industry.
You will:
- characterise temporal dependencies and autocorrelation in biomedical signals, with a particular emphasis on EEG
- quantify how temporal structure influences statistical inference, predictive modelling, and the identification of human cognitive and physiological states
- evaluate and compare approaches for mitigating temporal effects across:
- experimental design (for example task timing and trial structure)
- signal processing (for example filtering, segmentation, and decorrelation)
- statistical and machine learning models that explicitly account for temporal dependencies;
- assess current practices in academia and industry and develop practical guidelines for robust and reproducible biomedical time-series analysis
You will gain experience in:
- biomedical signal processing and EEG analysis
- time-series modelling and statistical inference
- machine learning and reproducible data analysis
- experimental research using real-world biomedical datasets
The project provides interdisciplinary training across data science, signal processing, neuroscience, and human-centred technologies. You will have opportunities to collaborate with academic and industrial partners and to disseminate your work through international conferences and journal publications.
The School of Electronics and Computer Science is committed to promoting equality, diversity inclusivity as demonstrated by our Athena SWAN award. We welcome all applicants regardless of their gender, ethnicity, disability, sexual orientation or age, and will give full consideration to applicants seeking flexible working patterns and those who have taken a career break. The University has a generous maternity policy, onsite childcare facilities, and offers a range of benefits to help ensure employees’ well-being and work-life balance. The University of Southampton is committed to sustainability and has been awarded the Platinum EcoAward.
