Combining radiomics and machine learning for enhanced localization of premature ventricular contractions

1

Premature Ventricular Contractions (PVC) can lead to serious cardiac issues, including left ventricular enlargement and increased risk of lethal arrhythmias.
2

ECG is commonly used for PVC localization but struggles with differentiating between closely located origins like LVOT and RVOT.
3

CCTA-based radiomics can identify microstructural changes associated with PVC, offering insights beyond what standard ECG can provide.
4

A hierarchical machine learning framework was developed to integrate CCTA radiomics and clinical data for improved PVC localization.
5

The study involved 304 patients undergoing RFCA, focusing on developing a machine learning model for accurate PVC origin localization.

Digital Health

by Jingjie Liu, Shiyu Dai, Lingxuan Hou, Boyang Zang, Yang Liu, Chongfu Jia, Xiaomeng Yin
May 15, 2026

1