Cardiovascular diseases are the leading cause of death globally, accounting for 17.9 million deaths annually (WHO). Early detection of arrhythmias can save lives, yet it remains challenging due to signal variability. This project addresses the critical gap between analytical methods and AI-driven solutions for analysing biomedical signals such as ECG.
By combining these approaches, the project aims to develop interpretable AI tools, ensuring safer and more reliable decision-making—an essential factor in healthcare applications.
The successful applicant will gain advanced expertise in exponential analysis, inverse problems, AI/ML techniques, signal processing, and mathematical modelling, along with invaluable industry collaboration experience (several months) with international partners, equipping them with skills highly sought after in academia and industry.
Objectives:
- Develop a hybrid methodology that integrates classical analytical methods and AI/ML tools to improve the classification of biomedical signals, such as identifying ECG patterns and detecting the onset of arrhythmias.
- Ensure AI decisions are interpretable and safe, addressing the critical need for transparency in healthcare applications.
- Leverage industry collaboration with industry collaborators to validate and implement algorithms in virtual models and medical devices.
- Achieve integration of these tools into real-world healthcare systems, bridging the gap between research and practical application.
Research Question:
- How can the integration of analytical modelling and AI/ML techniques enhance the reliability and safety of biomedical signal classification, such as ECG or EEG, and how can these methods be validated and implemented for real-world medical applications?
