ECGxAI

ECGxAI, a deep learning algorithm, was used to compress a median beat ECG, summarizing most ECG features into 21 explainable factors (FactorECG). Compared with QRSAREA and guideline ECG criteria, FactorECG was superior in predicting clinical outcomes in pre-implantation ECGs of CRT patients (c-statistic 0.69). The addition of 13 clinical variables had a limited effect on the FactorECG model (Δ c-statistic 0.03). Inferolateral T-wave inversion, smaller right precordial S- and T-wave amplitude, ventricular rate, and increased PR interval and P-wave duration were significant predictors of poor outcomes.

Topic

Surgery;Cardiology;Machine learning

Detail

  • Operation: Visualisation;Standardisation and normalisation;Regression analysis

  • Software interface: Library, Web user interface

  • Language: Python

  • License: GNU Affero General Public License v 3.0

  • Cost: Free

  • Version name: -

  • Credit: The Dutch Heart Foundation, The Netherlands Organisation for Health Research and Development.

  • Input: -

  • Output: -

  • Contact: Philippe C Wouters p.wouters@umcutrecht.nl

  • Collection: -

  • Maturity: -

Publications

  • Electrocardiogram-based deep learning improves outcome prediction following cardiac resynchronization therapy.
  • Wouters PC, et al. Electrocardiogram-based deep learning improves outcome prediction following cardiac resynchronization therapy. Electrocardiogram-based deep learning improves outcome prediction following cardiac resynchronization therapy. 2023; 44:680-692. doi: 10.1093/eurheartj/ehac617
  • https://doi.org/10.1093/EURHEARTJ/EHAC617
  • PMID: 36342291
  • PMC: PMC9940988

Download and documentation

< Back to DB search