Article
Unsupervised feature extraction using deep learning empowers discovery of genetic determinants of the electrocardiogram.
Genome medicine - 9 Oct 2025
Sieliwonczyk Ewa, Sau Arunashis, Patlatzoglou Konstantinos, McGurk Kathryn A, Pastika Libor, Thami Prisca K, Mangino Massimo, Zheng Sean L, Powell George, Curran Lara, Buchan Rachel J, Theotokis Pantazis, Peters Nicholas S, Loeys Bart, Kramer Daniel B, Waks Jonathan W, Ng Fu Siong, Ware James S
Abstract excerpt
BACKGROUND: Electrocardiograms (ECGs) are widely used to assess cardiac health, but traditional clinical interpretation relies on a limited set of human-defined parameters. While advanced data-driven methods can outperform analyses of conventional ECG features for some tasks, they often lack interpretability. Variational autoencoders (VAEs), a form of unsupervised machine learning, can address this limitation by...
Read the complete abstract on PubMedTopics
Share this publication in a Topic to start or enrich a Post.