Login / Signup

The effect of confounding data features on a deep learning algorithm to predict complete coronary occlusion in a retrospective observational setting.

Rob BriskRaymond R BondDewar FinlayJames A D McLaughlinAlicja PiadloStephen J LeslieDavid E GossmanIan B A MenownD J McEneaneyS Warren
Published in: European heart journal. Digital health (2021)
The dataset was too small for the second model to achieve meaningful performance, despite the use of transfer learning. However, 'data leakage' during the first iteration of the experiment led to falsely high results. This study highlights the risk of DL models leveraging data leaks to produce spurious results.
Keyphrases
  • deep learning
  • electronic health record
  • big data
  • machine learning
  • coronary artery
  • coronary artery disease
  • artificial intelligence
  • heart failure
  • cross sectional
  • aortic valve