Skip to main content

Table 1 Summary of data (total number of cardiovascular magnetic resonance (CMR) images) used during the training and cross-validation phase of the development of the deep learning (DL) algorithms described herein

From: Assessment of right ventricular size and function from cardiovascular magnetic resonance images using artificial intelligence

 

Training datasets

Cross-validation datasets

Datasets

UKBB

ToF

Clinical

UKBB

ToF

Clinical

DL1

67,907

513

1900

16,977

0

0

DL2

67,907

513

2033

26,179

460

9612

  1. Three datasets were used during algorithm training consisting of images from the UK Biobank (UKBB), tetralogy of Fallot (ToF), and clinical dataset consisting of various pathologies including hypertrophic cardiomyopathy, dilated cardiomyopathy, pulmonary hypertension, and left ventricular non-compaction cardiomyopathy