Table 5.
Comparison of the proposed approach with U-Net and its variants using BRATS2015 dataset.
| Authors, Year and Citation | Model | Dataset | DSS |
|---|---|---|---|
| Daimary et al. [42] | U-SegNet | BRATS2015 | 0.73 |
| Zhou et al., 2019 | OM-Net + CGAp | BRATS2015 | 0.87 |
| Kayalibay et al., 2017 | CNN + 3D filters | BRATS2015 | 0.85 |
| Isensee et al., 2018 | U-Net + more filters | BRATS2015 | 0.85 |
| + data augmentation | |||
| + dice-loss | |||
| Kamnitsas et al., 2016 | 3D CNN + CRF | BRATS2015 | 0.85 |
| Qin et al., 2018 | AFN-6 | BRATS2015 | 0.84 |
| Havaei et al. [43] | CNN(whole) | BRATS2015 | 0.88 |
| Havaei et al. [43] | CNN(core) | BRATS2015 | 0.79 |
| Havaei et al. [43] | CNN(enhanced) | BRATS2015 | 0.73 |
| Pereira et al. [44] | CNN(whole) | BRATS2015 | 0.87 |
| Pereira et al. [44] | CNN(core) | BRATS2015 | 0.73 |
| Pereira et al. [44] | CNN(enhanced) | BRATS2015 | 0.68 |
| Malmi et al. [45] | CNN(whole) | BRATS2015 | 0.80 |
| Malmi et al. [45] | CNN(core) | BRATS2015 | 0.71 |
| Malmi et al. [45] | CNN(enhanced) | BRATS2015 | 0.64 |
| Taye et al., 2018 [46] | MAKM | BRATS2015 | 0.68 |
| Re-implemented | U-Net | BRATS2015 | 0.75 |
| Erena et al., 2020 | Case-1:Proposed Approach (15 randomly selected images) | BRATS2015 | 0.89 |
| Erena et al., 2020 | Case-2:Proposed Approach (12 randomly selected images) | BRATS2015 | 0.90 |
| Erena et al., 2020 | Case-3:Proposed Approach (800 brain images) | BRATS2015 | 0.80 |
| Erena et al., 2020 | Average:Proposed Approach | BRATS2015 | 0.86 |