Skip to main content
European Radiology Experimental

Table 3 Performance analysis (mean ± standard deviation) of all within-patient pair combinations (2,225) of hepatobiliary phase images in the internal and external datasets (438 image pairs)

From: Fully automated convolutional neural network-based affine algorithm improves liver registration and lesion co-localization on hepatobiliary phase T1-weighted MR images

Dataset

Liver overlap score

Image correlation

Internal dataset (intra-exam)

0.88 ± 0.14

0.44 ± 0.15

Internal dataset (inter-exam)

0.81 ± 0.14

0.41 ± 0.14

External dataset (intra-exam)

0.91 ± 0.06

0.41 ± 0.13

  1. Statistical comparisons between internal datasets were performed using unpaired t tests
Back to article page