From: Hippocampal subfields segmentation in brain MR images using generative adversarial networks
Methods | Hippocampal subfields | Overall | ||||||||
---|---|---|---|---|---|---|---|---|---|---|
CA1 | CA2 | DG | CA3 | Head | Tail | SUB | ERC | PHG | ||
Multi-Atlas [8] | – | – | – | – | – | – | – | – | – | 0.8925 |
Hierarchical Multi-Atlas [10] | – | – | – | – | – | – | – | – | – | 0.8850 |
Dictionary learning and sparse representation [27] | 0.804 | 0.538 | 0.807 | 0.548 | 0.890 | 0.772 | 0.751 | 0.743 | 0.620 | – |
CNN [28] | 0.836 | 0.603 | 0.843 | 0.641 | 0.894 | 0.865 | 0.810 | 0.731 | 0.773 | – |
UG-net | 0.820 | 0.631 | 0.805 | 0.666 | 0.915 | 0.838 | 0.811 | 0.743 | 0.792 | 0.849 |
Our method | 0.919 | 0.648 | 0.903 | 0.673 | 0.929 | 0.913 | 0.906 | 0.884 | 0.899 | 0.916 |