PCPs: Patient Cardiac Prototypes

Dani Kiyasseh, Tingting Zhu, David A. Clifton

Keywords: Contrastive learning, dataset distillation, patient-similarity, physiological signals, healthcare

Abstract: Existing deep learning methodologies within the medical domain are typically population-based and difficult to interpret. This limits their clinical utility as population-based findings may not generalize to the individual patient. To overcome these obstacles, we propose to learn patient-specific representations, entitled patient cardiac prototypes (PCPs), that efficiently summarize the cardiac state of a patient. We show that PCPs, learned in an end-to-end manner via contrastive learning, allow for the discovery of similar patients both within and across datasets, and can be exploited for dataset distillation as a compact substitute for the original dataset.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Reviewed Version (pdf): https://openreview.net/references/pdf?id=3mpvZ7vGYU