Merge pull request #35 from andreped/andreped-patch-2
Browse files
README.md
CHANGED
|
@@ -125,7 +125,9 @@ The segmentation models were evaluated on an internal dataset against manual ann
|
|
| 125 |
| Parenchyma | 0.946±0.046 | 10.122±11.032 |
|
| 126 |
| Vessels | 0.355±0.090 | 24.872±5.161 |
|
| 127 |
|
| 128 |
-
The parenchyma segmentation model was trained on the LITS dataset, whereas the vessel model was trained on a local dataset. The LITS dataset is openly accessible and can be downloaded from [here](https://competitions.codalab.org/competitions/17094).
|
|
|
|
|
|
|
| 129 |
|
| 130 |
## Acknowledgements
|
| 131 |
If you found this tool helpful in your research, please, consider citing it (see [here](https://zenodo.org/badge/latestdoi/238680374) for more information on how to cite):
|
|
|
|
| 125 |
| Parenchyma | 0.946±0.046 | 10.122±11.032 |
|
| 126 |
| Vessels | 0.355±0.090 | 24.872±5.161 |
|
| 127 |
|
| 128 |
+
The parenchyma segmentation model was trained on the LITS dataset, whereas the vessel model was trained on a local dataset (Oslo-CoMet). The LITS dataset is openly accessible and can be downloaded from [here](https://competitions.codalab.org/competitions/17094).
|
| 129 |
+
|
| 130 |
+
The Oslo-CoMet included 60 patients, of which 11 representative patients were used as hold out sample for the performance metrics assessment.
|
| 131 |
|
| 132 |
## Acknowledgements
|
| 133 |
If you found this tool helpful in your research, please, consider citing it (see [here](https://zenodo.org/badge/latestdoi/238680374) for more information on how to cite):
|