Dr. Tong Xia

Postdoctoral Research Associate

University of Cambridge

mHealth, short for mobile health, denotes the utilization of mobile devices such as smartphones, tablets, and wearable gadgets in healthcare provision and health-related services. It encompasses a broad spectrum of applications and services leveraging mobile technology to enhance health outcomes, improve healthcare accessibility, and enable individuals to manage their health more effectively. My research in sustainable mHealth applications intersects various disciplines, necessitating expertise in medicine, machine learning, and signal processing. I am dedicated to addressing a fundamental query: How can we effectively utilize mobile sensing and AI to extend healthcare services to a wider population in a reliable manner? To tackle this question, I summarize a few key highlights of my research below.

 

• Algorithm for electrocardiogram (ECG) data, particularly those collected by mobile devices, for detecting heart abnormalities.
• Data mining and machine learning algorithms to model complex human behaviours like individual and collective mobility patterns.
• Sequential deep learning and graph neural networks for app usage behaviour modelling and prediction.
• Mobility intervention strategies for pandemic control and devising state-of-the-art intelligent interventions.
• Data-efficient transformer neural network for human mobility data generation.

• Uncertainty quantification for safety-critical health applications.
• Decentralized and privacy-preserving machine learning models for real-world applications.
• Toward generalizable foundation models for respiratory audio.