Skip to main content
. 2022 Nov 22;5:1059007. doi: 10.3389/frai.2022.1059007

Table 2.

Computational time comparison for the proposed DeepHeartCT frameworks using All and strong label (SL) dataset.

Dataset Training samples (cases) GPU training time (mins) GPU inference time (secs) CPU inference time (secs)
All 1,064 90 20.6 ± 4.80 32.0 ± 7.79
SL 110 50 21.9 ± 4.81 34.2 ± 7.04