Skip to main content

Table 1 AI outcomes from single and combination of two models on validation dataset

From: A clinical evaluation study of cardiothoracic ratio measurement using artificial intelligence

 

CTR

CTRdiff (%)

Single-model mode

AlbuNet

0.489 ± 0.074

− 0.69 ± 2.64

SegNet

0.491 ± 0.071

− 0.22 ± 3.17

VGG-11

0.502 ± 0.075

1.96 ± 3.20

VGG-16

0.494 ± 0.073

0.48 ± 2.91

Combined-model mode

AlbuNet + SegNet

0.494 ± 0.072

0.39 ± 1.99

AlbuNet + VGG-11

0.491 ± 0.071

− 0.18 ± 1.92

AlbuNet + VGG-16

0.492 ± 0.072

− 0.01 ± 1.98

SegNet + VGG-11

0.495 ± 0.071

0.63 ± 2.10

SegNet + VGG-16

0.493 ± 0.072

0.20 ± 2.09

VGG-11 + VGG-16

0.496 ± 0.073

0.83 ± 2.38

  1. The bold-data rows indicate CTR values of combination-model modes that were significantly different (P < 0.01) from each individual model before the combination
Back to article page

BMC Medical Imaging

ISSN: 1471-2342

Contact us