Cross-modal contrastive learning for severe stenosis classification

Arxiv CS.LG (machine learning)

The paper proposes a cross-modal contrastive learning approach that jointly learns representations from ECG and X-ray angiography data to classify severe coronary artery stenosis. The method aims to improve diagnostic accuracy by combining non-invasive and invasive modalities.

ecgcontrastive-learning

Arxiv CS.LG (machine learning)·Nikola Cenikj\"Ozg\"un TurgutAlexander M\"ullerAlexander StegerJan KehrerMarcus BruggerDaniel RueckertPhilip M\"uller

8 days ago