Cardiovascular disease (CVD) remains the leading cause of death worldwide, responsible for more than 170,000 deaths each year in the UK. Although prevention and treatment have improved, detecting early signs of heart disease and identifying those most at risk remain major challenges.
This PhD project will use two world-leading health resources—the UK Biobank and the Danish Nationwide ECG Cohort—to analyse millions of electrocardiograms (ECGs) linked with information on genetics, imaging, lifestyle, and clinical history. By applying advanced artificial intelligence (AI) methods, the project aims to reveal early warning signals in ECGs, predict cardiovascular risk years before symptoms appear, and uncover new biological patterns that could guide personalised prevention.
The novelty of this work lies in using AI to extract subtle patterns from ECGs that are invisible to traditional interpretation. Combining large-scale data from two national cohorts will produce robust, generalisable insights with international relevance, accelerating progress toward data-driven and equitable cardiovascular care.
Aim:
– To develop and evaluate AI methods for large-scale ECG analysis and multimodal integration to improve prediction and understanding of cardiovascular disease risk.
Objectives:
– Develop and validate scalable AI pipelines for automated ECG analysis across large datasets.
– Integrate ECG features with clinical, imaging, genetic, and lifestyle data to enhance prediction of arrhythmia, heart failure, and related outcomes.
– Discover novel ECG biomarkers using unsupervised and self-supervised learning.
– Evaluate model performance, interpretability, and generalisability across populations.
– Build interdisciplinary expertise across cardiovascular medicine, AI, and data science.
-Through this work, the project aims to advance early diagnosis, improve risk prediction, and strengthen international collaboration in digital cardiovascular health.