Table 2 Comparisons of the proposed methods against state-of-the-art methods on the intracranial blood vessel dataset
From: DNL-Net: deformed non-local neural network for blood vessel segmentation
Method | Dice coefficient (%) | Mean IoU (%) |
|---|---|---|
U-Net | 87.32 | 86.48 |
SegNet | 88.40 | 81.63 |
FCN16s | 76.14 | 66.53 |
DenseASPP | 84.38 | 81.80 |
Deeplab V3 +  | 90.70 | 87.83 |
RefineNet | 91.68 | 76.72 |
ENet | 85.97 | 81.72 |
BiSeNet | 92.92 | 89.33 |
SA-Net | 95.89 | 91.58 |
SSCA-Net | 96.20 | 92.70 |
DNL-Net | 96.63 | 92.93 |