Table 3 Grading of AI outcomes from single and combination of two models on validation dataset
From: A clinical evaluation study of cardiothoracic ratio measurement using artificial intelligence
Excellent grade | Good grade | |
|---|---|---|
Single-model mode | ||
AlbuNet | 4295 (57.1%) | 3222 (42.9%) |
SegNet | 4655 (61.9%) | 2862 (38.1%) |
VGG-11 | 3971 (52.8%) | 3546 (47.2%) |
VGG-16 | 5183 (68.9%) | 2334 (31.1%) |
Combined-model mode | ||
AlbuNet + VGG-11 | 6220 (82.7%) | 1297 (17.3%) |
AlbuNet + VGG-16 | 6121 (81.4%) | 1396 (18.6%) |
SegNet + VGG-11 | 5664 (75.3%) | 1853 (24.7%) |