86 tutorial updates (#183)

* update nifti read notebook

* revise notebook unet_segmentation_3d.ipynb

* revise transformations

        - revise Spacing transforms to rely on the affines only
        - rectified a few typos

* unit testing spacing transformations

* update docs

* update demo with spacing transform

* fixes padding mode spacing

* fixes load nifti

* 3d transform tutorial

* update notebooks

* revised spacing, load nifti for readability

* update example readme to include 3d transforms

* fixes tensor contiguous in spatial transforms, fixes docs

* update codecov xml for all branches

* [DLMED] update spleen notebook

Co-authored-by: Nic Ma <nma@nvidia.com>

Author: wyli

codecov.yml

@@ -5,7 +5,6 @@ coverage:
         target: 70%
         threshold: 10
         base: parent
-        branches: null
         if_no_uploads: error
         if_not_found: success
         if_ci_failed: error
@@ -18,7 +17,6 @@ coverage:
         # Allows PRs without tests, overall stats count
         threshold: 100
         base: auto
-        branches: null
         if_no_uploads: error
         if_not_found: success
         if_ci_failed: error

examples/README.md

@@ -10,17 +10,17 @@ python -m pip install -U notebook
 ```
 
 ### 2. List of examples
-#### 1. [classification_3d](https://github.com/Project-MONAI/MONAI/tree/master/examples/classification_3d)
+#### [classification_3d](./classification_3d)
 Training and evaluation examples of 3D classification based on DenseNet3D and [IXI dataset](https://brain-development.org/ixi-dataset):
 The examples are standard PyTorch programs and have both dictionary-based and array-based transformation versions.
-#### 2. [classification_3d_ignite](https://github.com/Project-MONAI/MONAI/tree/master/examples/classification_3d_ignite)
+#### [classification_3d_ignite](./classification_3d_ignite)
 Training and evaluation examples of 3D classification based on DenseNet3D and [IXI dataset](https://brain-development.org/ixi-dataset):
 The examples are PyTorch ignite programs and have both dictionary-based and array-based transformation versions.
-#### 3. [notebooks/multi_gpu_test](https://github.com/Project-MONAI/MONAI/blob/master/examples/notebooks/multi_gpu_test.ipynb)
+#### [notebooks/multi_gpu_test](./notebooks/multi_gpu_test.ipynb)
 This notebook is a quick demo for devices, run the Ignite trainer engine on CPU, GPU and multiple GPUs.
-#### 4. [notebooks/nifti_read_example](https://github.com/Project-MONAI/MONAI/blob/master/examples/notebooks/nifti_read_example.ipynb)
+#### [notebooks/nifti_read_example](./notebooks/nifti_read_example.ipynb)
 Illustrate reading NIfTI files and iterating over image patches of the volumes loaded from them.
-#### 5. [notebooks/spleen_segmentation_3d](https://github.com/Project-MONAI/MONAI/blob/master/examples/notebooks/spleen_segmentation_3d.ipynb)
+#### [notebooks/spleen_segmentation_3d](./notebooks/spleen_segmentation_3d.ipynb)
 This notebook is an end-to-end training and evaluation example of 3D segmentation based on [MSD Spleen dataset](http://medicaldecathlon.com):
 The example shows the flexibility of MONAI modules in a PyTorch-based program:
 - Transforms for dictionary-based training data structure.
@@ -30,17 +30,20 @@ The example shows the flexibility of MONAI modules in a PyTorch-based program:
 - 3D UNet, Dice loss function, Mean Dice metric for 3D segmentation task.
 - Sliding window inference.
 - Deterministic training for reproducibility.
-#### 6. [notebooks/transform_speed](https://github.com/Project-MONAI/MONAI/blob/master/examples/notebooks/transform_speed.ipynb)
+#### [notebooks/transform_speed](./notebooks/transform_speed.ipynb)
 Illustrate reading NIfTI files and test speed of different transforms on different devices.
-#### 7. [notebooks/transforms_demo_2d](https://github.com/Project-MONAI/MONAI/blob/master/examples/notebooks/transforms_demo_2d.ipynb)
-This notebook demonstrates the medical domain specific transforms on 2D medical images.
-#### 8. [notebooks/unet_segmentation_3d_ignite](https://github.com/Project-MONAI/MONAI/blob/master/examples/notebooks/unet_segmentation_3d_ignite.ipynb)
+#### [notebooks/transforms_demo_2d](./notebooks/transforms_demo_2d.ipynb)
+This notebook demonstrates the image transformations on histology images using
+[the GlaS Contest dataset](https://warwick.ac.uk/fac/sci/dcs/research/tia/glascontest/download/).
+#### [notebooks/3d_image_transforms](./notebooks/3D_image_transforms.ipynb)
+This notebook demonstrates the transformations on volumetric images.
+#### [notebooks/unet_segmentation_3d_ignite](./notebooks/unet_segmentation_3d_ignite.ipynb)
 This notebook is an end-to-end training & evaluation example of 3D segmentation based on synthetic dataset.
 The example is a PyTorch Ignite program and shows several key features of MONAI,
 especially with medical domain specific transforms and event handlers.
-#### 9. [segmentation_3d](https://github.com/Project-MONAI/MONAI/tree/master/examples/segmentation_3d)
+#### [segmentation_3d](./examples/segmentation_3d)
 Training and evaluation examples of 3D segmentation based on UNet3D and synthetic dataset.
 The examples are standard PyTorch programs and have both dictionary-based and array-based versions.
-#### 10. [segmentation_3d_ignite](https://github.com/Project-MONAI/MONAI/tree/master/examples/segmentation_3d_ignite)
+#### [segmentation_3d_ignite](./examples/segmentation_3d_ignite)
 Training and evaluation examples of 3D segmentation based on UNet3D and synthetic dataset.
 The examples are PyTorch Ignite programs and have both dictionary-base and array-based transformations.

examples/notebooks/3D_image_transforms.ipynb

@@ -17,7 +17,13 @@
     {
      "name": "stdout",
      "output_type": "stream",
-     "text": "MONAI version: 0.0.1\nPython version: 3.8.1 (default, Jan  8 2020, 22:29:32)  [GCC 7.3.0]\nNumpy version: 1.18.1\nPytorch version: 1.4.0\nIgnite version: 0.3.0\n"
+     "text": [
+      "MONAI version: 0.0.1\n",
+      "Python version: 3.5.6 |Anaconda, Inc.| (default, Aug 26 2018, 16:30:03)  [GCC 4.2.1 Compatible Clang 4.0.1 (tags/RELEASE_401/final)]\n",
+      "Numpy version: 1.18.2\n",
+      "Pytorch version: 1.4.0\n",
+      "Ignite version: 0.3.0\n"
+     ]
     }
    ],
    "source": [
@@ -35,56 +41,18 @@
     "\n",
     "import torch\n",
     "from torch.utils.data import DataLoader\n",
-    "import monai.transforms.compose as transforms\n",
+    "from monai.transforms.compose import Compose\n",
     "\n",
     "import monai\n",
     "\n",
-    "from monai.transforms.utils import rescale_array\n",
     "from monai.data.nifti_reader import NiftiDataset\n",
     "from monai.transforms import AddChannel, Transpose, Rescale, ToTensor, RandUniformPatch\n",
     "from monai.data.grid_dataset import GridPatchDataset\n",
+    "from monai.data.synthetic import create_test_image_3d\n",
     "\n",
     "monai.config.print_config()"
    ]
   },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "Define a function for creating test images and segmentations:"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 2,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "def create_test_image_3d(height, width, depth, numObjs=12, radMax=30, noiseMax=0.0, numSegClasses=5):\n",
-    "    '''Return a noisy 3D image and segmentation.'''\n",
-    "    image = np.zeros((width, height,depth))\n",
-    "\n",
-    "    for i in range(numObjs):\n",
-    "        x = np.random.randint(radMax, width - radMax)\n",
-    "        y = np.random.randint(radMax, height - radMax)\n",
-    "        z = np.random.randint(radMax, depth - radMax)\n",
-    "        rad = np.random.randint(5, radMax)\n",
-    "        spy, spx, spz = np.ogrid[-x:width - x, -y:height - y, -z:depth - z]\n",
-    "        circle = (spx * spx + spy * spy + spz * spz) <= rad * rad\n",
-    "\n",
-    "        if numSegClasses > 1:\n",
-    "            image[circle] = np.ceil(np.random.random() * numSegClasses)\n",
-    "        else:\n",
-    "            image[circle] = np.random.random() * 0.5 + 0.5\n",
-    "\n",
-    "    labels = np.ceil(image).astype(np.int32)\n",
-    "\n",
-    "    norm = np.random.uniform(0, numSegClasses * noiseMax, size=image.shape)\n",
-    "    noisyimage = rescale_array(np.maximum(image, norm))\n",
-    "\n",
-    "    return noisyimage, labels"
-   ]
-  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -94,14 +62,14 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": 2,
    "metadata": {},
    "outputs": [],
    "source": [
     "tempdir = tempfile.mkdtemp()\n",
     "\n",
     "for i in range(5):\n",
-    "    im, seg = create_test_image_3d(256,256,256)\n",
+    "    im, seg = create_test_image_3d(128, 128, 128)\n",
     "    \n",
     "    n = nib.Nifti1Image(im, np.eye(4))\n",
     "    nib.save(n, os.path.join(tempdir, 'im%i.nii.gz'%i))\n",
@@ -119,33 +87,35 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 3,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
-     "text": "torch.Size([5, 1, 64, 64, 64]) torch.Size([5, 256, 256, 256])\n"
+     "text": [
+      "torch.Size([5, 1, 64, 64, 64]) torch.Size([5, 1, 64, 64, 64])\n"
+     ]
     }
    ],
    "source": [
     "images = sorted(glob(os.path.join(tempdir,'im*.nii.gz')))\n",
     "segs = sorted(glob(os.path.join(tempdir,'seg*.nii.gz')))\n",
     "\n",
-    "imtrans=transforms.Compose([\n",
+    "imtrans = Compose([\n",
     "    Rescale(),\n",
     "    AddChannel(),\n",
     "    RandUniformPatch((64, 64, 64)),\n",
     "    ToTensor()\n",
     "])    \n",
     "\n",
-    "segtrans=transforms.Compose([\n",
+    "segtrans = Compose([\n",
     "    AddChannel(),\n",
     "    RandUniformPatch((64, 64, 64)),\n",
     "    ToTensor()\n",
     "])    \n",
     "    \n",
-    "ds = NiftiDataset(images, segs, imtrans, segtrans)\n",
+    "ds = NiftiDataset(images, segs, transform=imtrans, seg_transform=segtrans)\n",
     "\n",
     "loader = DataLoader(ds, batch_size=10, num_workers=2, pin_memory=torch.cuda.is_available())\n",
     "im, seg = monai.utils.misc.first(loader)\n",
@@ -161,28 +131,30 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 4,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
-     "text": "torch.Size([10, 1, 64, 64, 64]) torch.Size([10, 256, 64, 64])\n"
+     "text": [
+      "torch.Size([10, 1, 64, 64, 64]) torch.Size([10, 1, 64, 64, 64])\n"
+     ]
     }
    ],
    "source": [
-    "imtrans=transforms.Compose([\n",
+    "imtrans = Compose([\n",
     "    Rescale(),\n",
     "    AddChannel(),\n",
     "    ToTensor()\n",
     "])    \n",
     "\n",
-    "segtrans=transforms.Compose([\n",
+    "segtrans = Compose([\n",
     "    AddChannel(),\n",
     "    ToTensor()\n",
     "])    \n",
     "    \n",
-    "ds = NiftiDataset(images, segs, imtrans, segtrans)\n",
+    "ds = NiftiDataset(images, segs, transform=imtrans, seg_transform=segtrans)\n",
     "ds = GridPatchDataset(ds, (64, 64, 64))\n",
     "\n",
     "loader = DataLoader(ds, batch_size=10, num_workers=2, pin_memory=torch.cuda.is_available())\n",
@@ -192,7 +164,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": 5,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -223,9 +195,9 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.8.1-final"
+   "version": "3.5.6"
   }
  },
  "nbformat": 4,
  "nbformat_minor": 4
-}
+}

examples/notebooks/nifti_read_example.ipynb

@@ -70,7 +70,8 @@
     "import monai\n",
     "import monai.transforms.compose as transforms\n",
     "from monai.transforms.composables import \\\n",
-    "    LoadNiftid, AddChanneld, ScaleIntensityRanged, RandCropByPosNegLabeld, RandAffined\n",
+    "    LoadNiftid, AddChanneld, ScaleIntensityRanged, RandCropByPosNegLabeld, \\\n",
+    "    RandAffined, Spacingd, Orientationd\n",
     "from monai.data.utils import list_data_collate\n",
     "from monai.utils.sliding_window_inference import sliding_window_inference\n",
     "from monai.metrics.compute_meandice import compute_meandice\n",
@@ -115,6 +116,8 @@
     "train_transforms = transforms.Compose([\n",
     "    LoadNiftid(keys=['image', 'label']),\n",
     "    AddChanneld(keys=['image', 'label']),\n",
+    "    Spacingd(keys=['image', 'label'], pixdim=(1.5, 1.5, 2.), interp_order=(3, 0)),\n",
+    "    Orientationd(keys=['image', 'label'], axcodes='RAS'),\n",
     "    ScaleIntensityRanged(keys=['image'], a_min=-57, a_max=164, b_min=0.0, b_max=1.0, clip=True),\n",
     "    # randomly crop out patch samples from big image based on pos / neg ratio\n",
     "    # the image centers of negative samples must be in valid image area\n",
@@ -127,6 +130,8 @@
     "val_transforms = transforms.Compose([\n",
     "    LoadNiftid(keys=['image', 'label']),\n",
     "    AddChanneld(keys=['image', 'label']),\n",
+    "    Spacingd(keys=['image', 'label'], pixdim=(1.5, 1.5, 2.), interp_order=(3, 0)),\n",
+    "    Orientationd(keys=['image', 'label'], axcodes='RAS'),\n",
     "    ScaleIntensityRanged(keys=['image'], a_min=-57, a_max=164, b_min=0.0, b_max=1.0, clip=True)\n",
     "])"
    ]
@@ -516,4 +521,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 4
-}
+}