Skip to main content

Table 1 Import data

From: Weakly supervised video-based cardiac detection for hypertensive cardiomyopathy

Imported Data

Original Data: Total = 297 subjects, 1793 videos

Healthy (N)

Hypertensive (HTCM)

Total = 112 (707 videos)

Total = 185 (1086 videos)

Training dataset

Test dataset

Total = 247 (1489 videos)

Total = 50 (304 videos)

  1. A cohort of 297 subjects (1793 videos) were divided into 247 subjects as the training dataset, and 50 subjects as the test set, in an approximately 5:1 ratio (N = normal, HTCM = hypertensive cardiomyopathy). 93 healthy controls and 154 hypertensive subjects were assigned to the training dataset, and the remaining were allocated to the test dataset. Datasets were strictly divided by subjects rather than by videos