Skip to main content

Funding by Royal Society to enable international exchange for developing open-source platform for cardiac monitoring

Royal Academy of Engineering Research fellow, Dr Tingting Zhu receives funding to lead on an international collaboration

Tingting Zhu portrait

Royal Academy of Engineering Research fellow Dr Tingting Zhu

Dr Tingting Zhu and Professor Chengyu Liu (Executive Dean of School of Instrument Science and Engineering, Southeast University, China) have been recently awarded two-year funding support from the Royal Society and the Natural Science Foundation of China for the International Exchange Scheme – Cost Share Programme, to develop a machine learning framework and open-source platform for dynamic cardiac monitoring.

The Royal Society International Exchanges Scheme is designed to provide funding support for visits to stimulate new collaborations between scientists in the UK and overseas. In the Cost Share Programme, the Royal Society partners with different funding bodies, such as the Natural Science Foundation of China, to co-fund the scheme. The joint proposal requires scientists from the UK and overseas to each submit a separate application to be reviewed by the funding organisation in the residing country. Successful funding will only be provided when both applications have been awarded. The scheme is highly competitive, with only 57 awarded projects out of a total of 247 proposals submitted.

A large-scale pandemic like COVID-19 places extraordinary demands on the world’s health systems, and threatens the global communities in an unprecedented way. Other than respiratory symptoms, COVID-19 also affects heart function significantly with possible long-term damage. The electrocardiogram (ECG) is a gold standard and cost-effective diagnosis tool to evaluate cardiac function by recording the rhythm and activity of the heart. Dynamic monitoring of the ECG changes for abnormal rhythms (such as arrhythmia detection) can provide valuable information in screening, diagnosis, and risk assessment of cardiovascular function. Commercial ECG wearables offer a convenient means for clinicians to assess patients’ health statuses, especially in resource- constrained settings (e.g., poor doctor-to-patient ratio, and low hospital resources available). However, the analysis of arrhythmia detection is challenging due to lack of correct labelling, distribution inconsistencies caused by equipment, population, and lead differences.

The proposed project aims to develop an interpretable and personalised continual learning framework to address issues in dynamic ECG monitoring, with the ability to learn dynamically new tasks from different datasets derived from different populations or devices and collected at various times while maintaining the memory of old tasks. This project will be implemented jointly by Dr Tingting Zhu’s team, together with Dr Lei Lu at the Computational Health Informatics lab from the University of Oxford in the UK, and Prof. Chengyu Liu’s team at the Wearable Heart-Sleep-Emotion Intelligent Monitoring Lab from Southeast University in China.

Furthermore, the project aims to provide open-source models resulting from the research, which can then be widely used in any remote setting to assist clinical decision support and lower the health care costs for an individual as well as the healthcare provider. A novel open-source ECG database platform will be built to provide researchers with a large, rich dataset for developing and validating new methods in medical informatics.

Dr Zhu says, "I am pleased to receive this award as well as support from the Department of Engineering Science to enable the proposed collaboration. Long-term wearable ECG monitoring captures dynamic changes of the heart and provides an indication of general wellbeing. Continuous tracking and effective analysis on long-term ECGs may also discover potential risk factors of an individual, which would otherwise be left undiagnosed. The funding support helps to foster a new link between Oxford and Southeast University in jointly developing intelligent decision support tools using wearables for dynamic cardiac monitoring. We are currently working with local hospitals in Nanjing, China to link wearable data with electronic health records to provide a comprehensive tracking of holistic health and wellbeing of individuals".

Professor Liu says "I am really pleased to receive this collaboration funding to work with Dr Tingting Zhu, who is an outstanding scientist in the domain of ECG data processing and AI analysis. Recent advances in wearable and IoT technology have led to an explosion in routinely ECG monitoring. Although there are clear clinical and practical benefits to using AI models, those for ECG data are still in their infancy, and many nominal “intelligence” methods do not produce substantial help with clinical diagnostics. This is why we want to work together to strengthen the development of cardiac monitoring. Tingting’s team at Oxford University has a world reputation in ECG and machine learning research, so the collaboration between Southeast and Oxford University has a bright prospect".