The datasets are divided by species (human and mouse data).
For a given CDA, the dataset was divided by placing a random sample of 80% of the data in a first subset, and the remaining 20% in a second subset.
Because the dataset was divided by test-day records, the same animals could appear in both sub-datasets and they could even appear in both datasets in the same lactation.
LV datasets were divided into 16 segments, excluding the apex, as recommended by the American Society of Echocadiography [ 9].
The manually clustered websites were sorted by the date each website was identified, and then both datasets were divided into training and testing populations of 80%and20%0%, respectively.
Datasets were divided into modeling and validation datasets.
Both saturated and undersaturated datasets were divided into three segments.
The datasets were divided into two subsets.
The datasets were divided randomly into a development dataset and a test dataset.
Total counts from each reconstructed AC SC SPECT dataset series were divided by T to yield the observed AC SC SPECT counting rate (RSo).
