Merge branch 'main' into evamariie-patch-2-1

Author: MarcelRosier

BraTS-Toolkit/glioma_segmentation_with_BraTS_Toolkit.ipynb

@@ -6,7 +6,7 @@
    "source": [
     "# Warning\n",
     "\n",
-    "This tutorial is a work in progress, it does not cover all the functionalities yet. However, it will give you a good overview how you can generate tumor segmentations with BraTS Toolkit.\n"
+    "This tutorial is a work in progress, it does not cover all the functionalities yet. However, it will give you a good overview of how you can generate tumor segmentations with BraTS Toolkit.\n"
    ]
   },
   {
@@ -28,13 +28,13 @@
     "2. **Import** of raw data (cMRI)\n",
     "    \n",
     "        \n",
-    "3. **Preprocessing via BrainLes preprocessig package** instead of vanilla preprocessing pipeline from BraTS Toolkit. BraTS challenge algorithms expect preprocessed files (co-registered, skullstripped in SRI-24 space)\n",
+    "3. **Preprocessing via BrainLes preprocessing package** instead of vanilla preprocessing pipeline from BraTS Toolkit. BraTS challenge algorithms expect preprocessed files (co-registered, skullstripped in SRI-24 space)\n",
     "     \n",
     "          \n",
     "4. **Segmentation with BraTS Toolkit** \n",
     "      \n",
     "            \n",
-    "5. **Fusion** of generated segementations \n",
+    "5. **Fusion** of generated segmentations \n",
     "\n"
    ]
   },
@@ -55,10 +55,10 @@
     "optional (but recommended):\n",
     "\n",
     "- CUDA drivers\n",
-    "- a GPU that is supported (each algorithms supports different set of GPUs)\n",
+    "- a GPU that is supported (each algorithm supports different set of GPUs)\n",
     "<!-- TODO specify -->\n",
     "\n",
-    "Run the subsequent cell to check CUDA availability, your python and docker version.  "
+    "Run the subsequent cell to check CUDA availability, your Python and docker version.  "
    ]
   },
   {
@@ -206,14 +206,14 @@
    "source": [
     "## 2. Import raw input data\n",
     "\n",
-    "Raw imput data require\n",
+    "Raw input data require\n",
     "- exam with skull and 4 sequences (T1, T1c, T2, T2-FLAIR), only one file each (in total 4 files) \n",
     "    - file names must end with \"*t1.nii.gz\", \"*t1c.nii.gz\", \"*t2.nii.gz\" and \"*fla.nii.gz\"\n",
     "    - [Nifti \".nii.gz\"](https://brainder.org/2012/09/23/the-nifti-file-format/) file type \n",
     "- each exam (4 files) needs to be places in an individual folder within the data folder \n",
     "```<this repository>/BraTS-Toolkit/data/```. \n",
     "\n",
-    "The structure is shown using two example exams:\n",
+    "The structure is shown using two examples exams:\n",
     "```\n",
     "BraTS-Toolkit\n",
     "├── data\n",
@@ -297,7 +297,7 @@
    "source": [
     "## 3. Preprocessing\n",
     "\n",
-    "BraTS challenge algorithms expect co-registered, skullstripped files in SRI-24 space, to achieve this preprocessing is required.\n",
+    "BraTS challenge algorithms expect co-registered, skull-stripped files in SRI-24 space, to achieve this preprocessing is required.\n",
     "Instead of using the vanilla preprocessing pipeline from BraTS Toolkit, we recommend using the new [BrainLes preprocessing package](https://github.com/BrainLesion/preprocessing/tree/main/brainles_preprocessing).\n"
    ]
   },

preprocessing/preprocessing_tutorial.ipynb

@@ -235,7 +235,13 @@
     "from brainles_preprocessing.brain_extraction import HDBetExtractor\n",
     "from brainles_preprocessing.modality import Modality\n",
     "from brainles_preprocessing.preprocessor import Preprocessor\n",
-    "from brainles_preprocessing.registration import NiftyRegRegistrator\n",
+    "from brainles_preprocessing.registration import (\n",
+    "    ANTsRegistrator,\n",
+    "    NiftyRegRegistrator,\n",
+    "    eRegRegistrator,\n",
+    ")\n",
+    "\n",
+    "# ANTSRegistrator (default), NiftyRegRegistrator (Linux), eRegRegistrator (experimental)\n",
     "\n",
     "\n",
     "def preprocess_exam_in_brats_style(inputDir: str) -> None:\n",
@@ -326,7 +332,7 @@
     "        preprocessor = Preprocessor(\n",
     "            center_modality=center,\n",
     "            moving_modalities=moving_modalities,\n",
-    "            registrator=NiftyRegRegistrator(),\n",
+    "            registrator=ANTsRegistrator(),\n",
     "            brain_extractor=HDBetExtractor(),\n",
     "            # optional: we provide a temporary directory as a sandbox for the preprocessin\n",
     "            temp_folder=\"temporary_directory\",\n",