Table 4 Segmentation performance in Dice Coefficient (DC) obtained on the \(MICCAI \ iSEG\) dataset achieved by our model (with and without BIM), compared to the state-of-the-art models
From: Learning to detect boundary information for brain image segmentation
Model | Dice Coefficient (DC) accuracy | ||
|---|---|---|---|
CSF (%) | GM (%) | WM (%) | |
Özgün et al. [50] | 91.2 | 86.1 | 84.1 |
Dong et al. [51] | 83.5 | 85.2 | 86.4 |
Konstantinos et al. [51] | 90.3 | 86.8 | 84.3 |
Mahbod et al. [52] | 85.5 | 87.3 | 88.7 |
3D, FCNÂ +Â MILÂ +Â GÂ +Â K [17] | 94.1 | 90.2 | 89.7 |
Multi-stage [38] | 95.0 | 94.0 | 92.0 |
Ours (with BIM) | 94.0 | 94.3 | 91.0 |
Ours (without BIM) | 90.0 | 89.0 | 86.0 |