Best Paper Award at IEEE Journal of Biomedical and Health Informatics

We are delighted to announce that the paper “Uncertainty-Inspired Multi-Task Learning in Arbitrary Scenarios of ECG Monitoring” has received the Best Paper Award from the IEEE Journal of Biomedical and Health Informatics (JBHI). The award recognises outstanding contributions to the advancement of biomedical and health informatics research, highlighting the paper’s innovation and impact in the field of intelligent cardiovascular monitoring.

This work introduces UI-Beat, an uncertainty-inspired model for beat-level ECG diagnosis, designed to address the challenges of noise, artefacts, and variability in real-world and wearable ECG data. The model disentangles epistemic and aleatoric uncertainty within a deterministic neural network to improve both noise detection and heartbeat classification across heterogeneous data sources. UI-Beat achieves state-of-the-art performance in heartbeat localisation and classification, offering a robust and generalisable framework for ECG monitoring in diverse clinical and wearable scenarios. For more information, please refer to the paper.

The paper represents a collaborative effort between our lab and Southeast University, supported by funding from the Royal Society and the National Natural Science Foundation of China.