135_CinCFinalPDF.pdf (693.42 kB)

Diagnostic of multiple cardiac disorders from 12-lead ECGs using Graph Convolutional Network based multi-label classification

Download (693.42 kB)
conference contribution
posted on 09.10.2020, 12:52 by Z Jiang, TP Almeida, FS Schlindwein, GA Ng, Huiyu Zhou, X Li
Automated detection and classification of clinical elec-trocardiogram (ECG) play a critical role in the analysisof cardiac disorders. Deep learning is effective for auto-mated feature extraction and has shown promising resultsin ECG classification. Most of these methods, however,assume that multiple cardiac disorders are mutually exclu-sive. In this work, we have created and trained a noveldeep learning architecture for addressing the multi-labelclassification of 12-lead ECGs. It contains an ECG rep-resentation work for extracting features from raw ECGrecordings and a Graph Convolutional Network (GCN) formodelling and capturing label dependencies. In the Phys-ioNet/Computing in Cardiology Challenge 2020, our team,Leicester-Fox, reached a score of0.627±0.054using 5-fold cross-validation on the full training data.

History

Citation

Computing in cardiology, 2020; Vol 47

Source

Computing in cardiology (CinC) 2020, Rimini, Italy, 13th-16th September 2020.

Version

AM (Accepted Manuscript)

Published in

Computing in cardiology

Volume

47

Publisher

CINC

issn

2325-887X

Acceptance date

22/09/2020

Copyright date

2020

Available date

22/09/2020

Temporal coverage: start date

13/09/2020

Temporal coverage: end date

16/09/2020

Language

en

Usage metrics

Categories

Exports