Results will be saved under `test_results` by default.
## Model performance across intra-domain and public out-of-domain data
- For cardiac segmentation on short-axis images, we provide the model trained on UKBB (~4000 training subjects), with extensive data augmentations. This model achieves extradinary performance across various unseen challenge datasets, [ACDC](),[M&Ms](). We report the segmentation performance in terms of Dice score.
- For cardiac segmentation on short-axis images, we provide the model trained on UKBB (~4000 training subjects), with extensive data augmentations. This model achieves extradinary performance across various unseen challenge datasets, [ACDC](https://www.creatis.insa-lyon.fr/Challenge/acdc/databases.html), [M&Ms](https://www.ub.edu/mnms/). We report the segmentation performance on these datasets in terms of Dice score.
