Merge branch 'MIC-DKFZ:master' into master

documentation/dataset_format.md

@@ -26,7 +26,8 @@ T2 MRI, …) and FILE_ENDING is the file extension used by your image format (.p
 The dataset.json file connects channel names with the channel identifiers in the 'channel_names' key (see below for details).
 
 Side note: Typically, each channel/modality needs to be stored in a separate file and is accessed with the XXXX channel identifier. 
-Exception are natural images (RGB; .png) where the three color channels can all be stored in one file (see the [road segmentation](../nnunetv2/dataset_conversion/Dataset120_RoadSegmentation.py) dataset as an example). 
+Exception are natural images (RGB; .png) where the three color channels can all be stored in one file (see the 
+[road segmentation](../nnunetv2/dataset_conversion/Dataset120_RoadSegmentation.py) dataset as an example). 
 
 **Segmentations** must share the same geometry with their corresponding images (same shape etc.). Segmentations are 
 integer maps with each value representing a semantic class. The background must be 0. If there is no background, then 
@@ -57,14 +58,14 @@ of what the raw data was provided in! This is for performance reasons.
 
 
 By default, the following file formats are supported:
+
 - NaturalImage2DIO: .png, .bmp, .tif
 - NibabelIO: .nii.gz, .nrrd, .mha
 - NibabelIOWithReorient: .nii.gz, .nrrd, .mha. This reader will reorient images to RAS!
 - SimpleITKIO: .nii.gz, .nrrd, .mha
 - Tiff3DIO: .tif, .tiff. 3D tif images! Since TIF does not have a standardized way of storing spacing information, 
-nnU-Net expects each TIF file to be accompanied by an identically named .json file that contains three numbers 
-(no units, no comma. Just separated by whitespace), one for each dimension.
-
+nnU-Net expects each TIF file to be accompanied by an identically named .json file that contains this information (see
+[here](#datasetjson)).
 
 The file extension lists are not exhaustive and depend on what the backend supports. For example, nibabel and SimpleITK 
 support more than the three given here. The file endings given here are just the ones we tested!
@@ -200,6 +201,27 @@ There is a utility with which you can generate the dataset.json automatically. Y
 [here](../nnunetv2/dataset_conversion/generate_dataset_json.py). 
 See our examples in [dataset_conversion](../nnunetv2/dataset_conversion) for how to use it. And read its documentation!
 
+As described above, a json file that contains spacing information is required for TIFF files.
+An example for a 3D TIFF stack with units corresponding to 7.6 in x and y, 80 in z is:
+
+```
+{
+    "spacing": [7.6, 7.6, 80.0]
+}
+```
+
+Within the dataset folder, this file (named `cell6.json` in this example) would be placed in the following folders:
+
+    nnUNet_raw/Dataset123_Foo/
+    ├── dataset.json
+    ├── imagesTr
+    │   ├── cell6.json
+    │   └── cell6_0000.tif
+    └── labelsTr
+        ├── cell6.json
+        └── cell6.tif
+
+
 ## How to use nnU-Net v1 Tasks
 If you are migrating from the old nnU-Net, convert your existing datasets with `nnUNetv2_convert_old_nnUNet_dataset`!
 

documentation/installation_instructions.md

@@ -82,6 +82,6 @@ easy identification.
 
 Note that these commands simply execute python scripts. If you installed nnU-Net in a virtual environment, this
 environment must be activated when executing the commands. You can see what scripts/functions are executed by 
-checking the entry_points in the setup.py file.
+checking the project.scripts in the [pyproject.toml](../pyproject.toml) file.
 
 All nnU-Net commands have a `-h` option which gives information on how to use them.

nnunetv2/experiment_planning/experiment_planners/default_experiment_planner.py

@@ -1,4 +1,3 @@
-import os.path
 import shutil
 from copy import deepcopy
 from functools import lru_cache
@@ -81,6 +80,10 @@ def __init__(self, dataset_name_or_id: Union[str, int],
 
         self.plans = None
 
+        if isfile(join(self.raw_dataset_folder, 'splits_final.json')):
+            _maybe_copy_splits_file(join(self.raw_dataset_folder, 'splits_final.json'),
+                                    join(preprocessed_folder, 'splits_final.json'))
+
     def determine_reader_writer(self):
         example_image = self.dataset[self.dataset.keys().__iter__().__next__()]['images'][0]
         return determine_reader_writer_from_dataset_json(self.dataset_json, example_image)
@@ -642,5 +645,23 @@ def load_plans(self, fname: str):
         self.plans = load_json(fname)
 
 
+def _maybe_copy_splits_file(splits_file: str, target_fname: str):
+    if not isfile(target_fname):
+        shutil.copy(splits_file, target_fname)
+    else:
+        # split already exists, do not copy, but check that the splits match.
+        # This code allows target_fname to contain more splits than splits_file. This is OK.
+        splits_source = load_json(splits_file)
+        splits_target = load_json(target_fname)
+        # all folds in the source file must match the target file
+        for i in range(len(splits_source)):
+            train_source = set(splits_source[i]['train'])
+            train_target = set(splits_target[i]['train'])
+            assert train_target == train_source
+            val_source = set(splits_source[i]['val'])
+            val_target = set(splits_target[i]['val'])
+            assert val_source == val_target
+
+
 if __name__ == '__main__':
     ExperimentPlanner(2, 8).plan_experiment()

nnunetv2/experiment_planning/plan_and_preprocess_api.py

@@ -1,17 +1,16 @@
-import shutil
 from typing import List, Type, Optional, Tuple, Union
 
-import nnunetv2
-from batchgenerators.utilities.file_and_folder_operations import join, maybe_mkdir_p, subfiles, load_json
+from batchgenerators.utilities.file_and_folder_operations import join, maybe_mkdir_p, load_json
 
+import nnunetv2
+from nnunetv2.configuration import default_num_processes
 from nnunetv2.experiment_planning.dataset_fingerprint.fingerprint_extractor import DatasetFingerprintExtractor
 from nnunetv2.experiment_planning.experiment_planners.default_experiment_planner import ExperimentPlanner
 from nnunetv2.experiment_planning.verify_dataset_integrity import verify_dataset_integrity
 from nnunetv2.paths import nnUNet_raw, nnUNet_preprocessed
-from nnunetv2.utilities.dataset_name_id_conversion import convert_id_to_dataset_name, maybe_convert_to_dataset_name
+from nnunetv2.utilities.dataset_name_id_conversion import convert_id_to_dataset_name
 from nnunetv2.utilities.find_class_by_name import recursive_find_python_class
 from nnunetv2.utilities.plans_handling.plans_handler import PlansManager
-from nnunetv2.configuration import default_num_processes
 from nnunetv2.utilities.utils import get_filenames_of_train_images_and_targets
 
 

nnunetv2/experiment_planning/verify_dataset_integrity.py

@@ -76,7 +76,7 @@ def check_cases(image_files: List[str], label_file: str, expected_num_channels:
     if not np.allclose(spacing_seg, spacing_images):
         print('Error: Spacing mismatch between segmentation and corresponding images. \nSpacing images: %s. '
               '\nSpacing seg: %s. \nImage files: %s. \nSeg file: %s\n' %
-              (shape_image, shape_seg, image_files, label_file))
+              (spacing_images, spacing_seg, image_files, label_file))
         ret = False
 
     # check modalities

nnunetv2/imageio/reader_writer_registry.py

@@ -4,7 +4,7 @@
 from batchgenerators.utilities.file_and_folder_operations import join
 
 import nnunetv2
-from nnunetv2.imageio.natural_image_reager_writer import NaturalImage2DIO
+from nnunetv2.imageio.natural_image_reader_writer import NaturalImage2DIO
 from nnunetv2.imageio.nibabel_reader_writer import NibabelIO, NibabelIOWithReorient
 from nnunetv2.imageio.simpleitk_reader_writer import SimpleITKIO
 from nnunetv2.imageio.tif_reader_writer import Tiff3DIO

nnunetv2/inference/examples.py

@@ -12,7 +12,7 @@
         tile_step_size=0.5,
         use_gaussian=True,
         use_mirroring=True,
-        perform_everything_on_gpu=True,
+        perform_everything_on_device=True,
         device=torch.device('cuda', 0),
         verbose=False,
         verbose_preprocessing=False,