From: DCAU-Net: dense convolutional attention U-Net for segmentation of intracranial aneurysm images
Method | Parameters | Train time (s) | Test time (s) | ||
---|---|---|---|---|---|
 < 3 mm | 3–7 mm |  > 7 mm | |||
  HeadXNet [29] | 101.59 M | 423.67 | 2.01 | 1.85 | 1.92 |
  DeepMedic [32] | 2.51 M | 396.14 | 1.41 | 1.25 | 1.33 |
  GLIA-Net [33] | 49.95 M | 379.31 | 2.09 | 1.86 | 1.92 |
  DAResUNet [30] | 71.97 M | 425.85 | 1.52 | 1.54 | 1.51 |
  U-Net [36] | 5.01 M | 367.04 | 1.41 | 1.37 | 1.41 |
  Ours | 38.35 M | 400.74 | 1.38 | 1.26 | 1.36 |