diff --git a/notebooks/external_augmentation.ipynb b/notebooks/external_augmentation.ipynb
index 4ff452e8..3e61493b 100644
--- a/notebooks/external_augmentation.ipynb
+++ b/notebooks/external_augmentation.ipynb
@@ -3,39 +3,320 @@
   {
    "cell_type": "code",
    "execution_count": null,
+   "metadata": {
+    "collapsed": false,
+    "jupyter": {
+     "outputs_hidden": false
+    },
+    "pycharm": {
+     "is_executing": false,
+     "name": "#%%\n"
+    }
+   },
    "outputs": [],
    "source": [
     "%reload_ext autoreload\n",
     "%autoreload 2\n",
     "%matplotlib inline"
-   ],
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Integration External Frameworks into a `rising` Augmentation Pipeline"
+   ]
+  },
+  {
+   "cell_type": "markdown",
    "metadata": {
-    "collapsed": false,
     "pycharm": {
-     "name": "#%%\n",
-     "is_executing": false
+     "name": "#%% md\n"
     }
-   }
+   },
+   "source": [
+    "### Using transformation from external libraries inside `rising`\n",
+    "> Note: Some external augmentation libraries are only supported at the beginning of\n",
+    "the transformation pipeline. In general, please consider creating an issue in `rising` \n",
+    "and there will be a high chance we (or if you prefer you :) ) will add the transformation in the future :) "
+   ]
   },
   {
    "cell_type": "markdown",
+   "metadata": {},
    "source": [
-    "## Using transformation from external libraries inside `rising`\n",
-    "> Note: Some external augmentation libraries are only supported at the beginning of\n",
-    "the transformation pipeline. Generally speaking, if you need to resort to an\n",
-    "external library for augmentations, consider creating an issue in `rising` \n",
-    "and there is a high chance we will add the transformation in the future :) "
-   ],
+    "## 3D (Volumetric) Augmentation\n",
+    "The first part of this notebook will focus on frameworks which support volumetric transformations (like rising also does). This mean data with a shape of [C, D, H, W] (C=Channels, DHW sptial dimensions). "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "!pip install --quiet --upgrade SimpleITK\n",
+    "!git clone https://github.com/PhoenixDL/rising.git\n",
+    "!pip install --quiet --upgrade ./rising"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# download some volumetric data (here MRI Data)\n",
+    "from io import BytesIO\n",
+    "from zipfile import ZipFile\n",
+    "from urllib.request import urlopen\n",
+    "\n",
+    "resp = urlopen(\"http://www.fmrib.ox.ac.uk/primers/intro_primer/ExBox3/ExBox3.zip\")\n",
+    "zipfile = ZipFile(BytesIO(resp.read()))\n",
+    "\n",
+    "img_file = zipfile.extract(\"ExBox3/T1_brain.nii.gz\")\n",
+    "mask_file = zipfile.extract(\"ExBox3/T1_brain_seg.nii.gz\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import SimpleITK as sitk\n",
+    "import numpy as np\n",
+    "\n",
+    "img = sitk.GetArrayFromImage(sitk.ReadImage(img_file))\n",
+    "img = img.astype(np.float32)\n",
+    "# sitk.WriteImage(sitk.GetImageFromArray(img), img_file)\n",
+    "mask = mask = sitk.GetArrayFromImage(sitk.ReadImage(mask_file))\n",
+    "mask = mask.astype(np.float32)\n",
+    "# sitk.WriteImage(sitk.GetImageFromArray(mask), mask_file)\n",
+    "\n",
+    "assert mask.shape == img.shape\n",
+    "print(f\"Image shape {img.shape}\")\n",
+    "print(f\"Mask shape {mask.shape}\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
    "metadata": {
-    "collapsed": false,
     "pycharm": {
      "name": "#%% md\n"
     }
-   }
+   },
+   "source": [
+    "### Integration of `batchgenerators` transformations\n",
+    "Note: when batchgenerator transformations are integrated, gradients can not be propagated through its\n",
+    "transformations.\n",
+    "\n",
+    "`batchgenerators` transformations are based on numpy to be framework agnostic. They are also based\n",
+    "on dictionaries which are modified through the transformations.\n",
+    "\n",
+    "There are two steps which need to be integrated into your pipelin in order to the \n",
+    "`batchgenerators` transforms\n",
+    "\n",
+    "1. Exchange the `default_collate` function inside the dataloder with `numpy_collate`\n",
+    "2. When switching from `batchgenerators` transformations to `rising` transformations, insdert `ToTensor` transformation"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "collapsed": false,
+    "jupyter": {
+     "outputs_hidden": false
+    },
+    "pycharm": {
+     "is_executing": false,
+     "name": "#%%\n"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "# setup transforms\n",
+    "from rising.transforms import *\n",
+    "from batchgenerators.transforms import ZeroMeanUnitVarianceTransform\n",
+    "\n",
+    "transforms = []\n",
+    "# convert tuple into dict\n",
+    "transforms.append(SeqToMap(\"data\", \"label\"))\n",
+    "\n",
+    "# batchgenerators transforms\n",
+    "transforms.append(ZeroMeanUnitVarianceTransform())\n",
+    "# ... additional batchgenerator transforms\n",
+    "\n",
+    "# convert to tensor\n",
+    "transforms.append(ToTensor())\n",
+    "\n",
+    "# rising transforms\n",
+    "transforms.append(Rot90((0, 1)))\n",
+    "transforms.append(Mirror(dims=(0, 1)))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "collapsed": false,
+    "jupyter": {
+     "outputs_hidden": false
+    },
+    "pycharm": {
+     "is_executing": false,
+     "name": "#%%\n"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "from rising.loading import DataLoader, default_transform_call, numpy_collate\n",
+    "from rising.transforms import Compose\n",
+    "\n",
+    "composed = Compose(transforms, transform_call=default_transform_call)\n",
+    "dataloader = DataLoader(dataset, batch_size=8, batch_transforms=composed,\n",
+    "                        num_workers=0, collate_fn=numpy_collate)\n",
+    "_iter = iter(dataloader)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "batch = next(_iter)\n",
+    "show_batch(batch[\"data\"])"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Integration of `tochio`"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "!pip install --quiet --upgrade torchio"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "dataset = None\n",
+    "\n",
+    "import torchio\n",
+    "\n",
+    "subject_a = torchio.Subject(\n",
+    "    t1=torchio.Image('./ExBox3/T1_brain.nii.gz', torchio.INTENSITY),\n",
+    ")\n",
+    "rescale = torchio.transforms.RescaleIntensity((0, 1))\n",
+    "transform = torchio.transforms.Compose([rescale])\n",
+    "\n",
+    "# ImagesDataset is a subclass of torch.data.utils.Dataset\n",
+    "dataset = torchio.ImagesDataset([subject_a], transform=transform)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "dataset[0]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# setup any additional rising transformations\n",
+    "from rising.transforms import *\n",
+    "\n",
+    "class SelectKeys(AbstractTransform):\n",
+    "    def __init__(self, keys=[\"t1\"]):\n",
+    "        super().__init__(grad=False)\n",
+    "        self.keys = keys\n",
+    "    \n",
+    "    def forward(self, **batch):\n",
+    "        for _key in self.keys:\n",
+    "            batch[_key] = batch[_key][\"data\"]\n",
+    "        return batch"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "rising_transforms = [\n",
+    "    SelectKeys(keys=[\"t1\"]),\n",
+    "    Rot90(keys=(\"t1\",), dims=(0, 1)),\n",
+    "    Mirror(keys=(\"t1\",), dims=(0, 1)),\n",
+    "]\n",
+    "batch_transforms = Compose(rising_transforms)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Instead of using the native pytorch dataloader we exchange it for the dataloder from rising \n",
+    "from rising.loading import DataLoader\n",
+    "\n",
+    "dataloader = DataLoader(dataset, batch_size=1, batch_transforms=batch_transforms, num_workers=4)\n",
+    "_iter = iter(dataloader)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "batch = next(_iter)\n",
+    "print(batch)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 2D Augmentation"
+   ]
   },
   {
    "cell_type": "code",
    "execution_count": null,
+   "metadata": {
+    "collapsed": false,
+    "jupyter": {
+     "outputs_hidden": false
+    },
+    "pycharm": {
+     "is_executing": false,
+     "name": "#%%\n"
+    }
+   },
    "outputs": [],
    "source": [
     "# lets prepare a basic dataset (e.g. one from `torchvision`)\n",
@@ -65,18 +346,21 @@
     "\n",
     "dataset = torchvision.datasets.MNIST(\n",
     "    os.getcwd(), train=True, download=True, transform=to_array)"
-   ],
-   "metadata": {
-    "collapsed": false,
-    "pycharm": {
-     "name": "#%%\n",
-     "is_executing": false
-    }
-   }
+   ]
   },
   {
    "cell_type": "code",
    "execution_count": null,
+   "metadata": {
+    "collapsed": false,
+    "jupyter": {
+     "outputs_hidden": false
+    },
+    "pycharm": {
+     "is_executing": false,
+     "name": "#%%\n"
+    }
+   },
    "outputs": [],
    "source": [
     "#  plot shape\n",
@@ -87,18 +371,21 @@
     "plt.imshow(dataset[0][0][0], cmap='gray')\n",
     "plt.colorbar()\n",
     "plt.show()"
-   ],
-   "metadata": {
-    "collapsed": false,
-    "pycharm": {
-     "name": "#%%\n",
-     "is_executing": false
-    }
-   }
+   ]
   },
   {
    "cell_type": "code",
    "execution_count": null,
+   "metadata": {
+    "collapsed": false,
+    "jupyter": {
+     "outputs_hidden": false
+    },
+    "pycharm": {
+     "is_executing": false,
+     "name": "#%%\n"
+    }
+   },
    "outputs": [],
    "source": [
     "# helper function to visualize batches of images\n",
@@ -109,101 +396,117 @@
     "    plt.imshow(grid[0], cmap='gray')\n",
     "    # plt.colorbar()\n",
     "    plt.show()"
-   ],
-   "metadata": {
-    "collapsed": false,
-    "pycharm": {
-     "name": "#%%\n",
-     "is_executing": false
-    }
-   }
+   ]
   },
   {
    "cell_type": "markdown",
+   "metadata": {},
    "source": [
-    "### Integration of `batchgenerators` transformations into the augmentation pipeline.\n",
-    "Note: when batchgenerator transformations are integrated, gradients can not be propagated through\n",
-    "the transformation pipeline.\n",
-    "\n",
-    "`batchgenerators` transformations are based on numpy to be framework agnostic. They are also based\n",
-    "on dictionaries which are modified through the transformations.\n",
+    "### Integration of `albumentation`"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "!pip install --quiet --upgrade albumentations"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from albumentations import RandomRotate90, Flip, Compose\n",
     "\n",
-    "There are two steps which need to be integrated into your pipelin in order to the \n",
-    "`batchgenerators` transforms\n",
+    "def aug(p=0.5):\n",
+    "    return Compose([\n",
+    "        RandomRotate90(),\n",
+    "        Flip(),\n",
+    "    ], p=p)\n",
     "\n",
-    "1. Exchange the `default_collate` function inside the dataloder with `numpy_collate`\n",
-    "2. When switching from `batchgenerators` transformations to `rising` transformations, insdert `ToTensor` transformation"
-   ],
-   "metadata": {
-    "collapsed": false,
-    "pycharm": {
-     "name": "#%% md\n"
-    }
-   }
+    "augmentation = aug(p=0.9)"
+   ]
   },
   {
    "cell_type": "code",
    "execution_count": null,
+   "metadata": {},
    "outputs": [],
    "source": [
-    "# setup transforms\n",
-    "from rising.transforms import *\n",
-    "from batchgenerators.transforms import ZeroMeanUnitVarianceTransform\n",
-    "\n",
-    "transforms = []\n",
-    "# convert tuple into dict\n",
-    "transforms.append(SeqToMap(\"data\", \"label\"))\n",
-    "# batchgenerators transforms\n",
-    "transforms.append(ZeroMeanUnitVarianceTransform())\n",
-    "# convert to tensor\n",
-    "transforms.append(ToTensor())\n",
-    "# rising transforms\n",
-    "transforms.append(Rot90((0, 1)))\n",
-    "transforms.append(Mirror(dims=(0, 1)))"
-   ],
-   "metadata": {
-    "collapsed": false,
-    "pycharm": {
-     "name": "#%%\n",
-     "is_executing": false
-    }
-   }
+    "rising_transforms = [\n",
+    "    SelectKeys(keys=[\"t1\"]),\n",
+    "    Rot90(keys=(\"t1\",), dims=(0, 1)),\n",
+    "    Mirror(keys=(\"t1\",), dims=(0, 1)),\n",
+    "]\n",
+    "batch_transforms = Compose(rising_transforms)"
+   ]
   },
   {
    "cell_type": "code",
    "execution_count": null,
+   "metadata": {},
    "outputs": [],
    "source": [
-    "from rising.loading import DataLoader, default_transform_call, numpy_collate\n",
-    "from rising.transforms import Compose\n",
-    "\n",
-    "composed = Compose(transforms, transform_call=default_transform_call)\n",
-    "dataloader = DataLoader(dataset, batch_size=8, batch_transforms=composed,\n",
-    "                        num_workers=0, collate_fn=numpy_collate)\n",
+    "# Instead of using the native pytorch dataloader we exchange it for the dataloder from rising \n",
+    "from rising.loading import DataLoader\n",
     "\n",
-    "_iter = iter(dataloader)\n",
-    "batch = next(_iter)\n",
-    "show_batch(batch[\"data\"])"
-   ],
-   "metadata": {
-    "collapsed": false,
-    "pycharm": {
-     "name": "#%%\n",
-     "is_executing": false
-    }
-   }
+    "dataloader = DataLoader(dataset, batch_size=1, batch_transforms=batch_transforms, num_workers=4)\n",
+    "_iter = iter(dataloader)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Integration of `imgaug`"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "!pip install --quiet --upgrade imgaug"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# needs a rename transform"
+   ]
   },
   {
    "cell_type": "markdown",
+   "metadata": {},
    "source": [
-    "### More libraries will be added in the future :) \n"
-   ],
+    "### Integration of `torchvision`"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "markdown",
    "metadata": {
-    "collapsed": false,
     "pycharm": {
      "name": "#%% md\n"
     }
-   }
+   },
+   "source": [
+    "### You want a library which is not listed here? Just open an issue [here](https://github.com/PhoenixDL/rising/issues).\n"
+   ]
   }
  ],
  "metadata": {
@@ -215,25 +518,25 @@
   "language_info": {
    "codemirror_mode": {
     "name": "ipython",
-    "version": 2
+    "version": 3
    },
    "file_extension": ".py",
    "mimetype": "text/x-python",
    "name": "python",
    "nbconvert_exporter": "python",
-   "pygments_lexer": "ipython2",
-   "version": "2.7.6"
+   "pygments_lexer": "ipython3",
+   "version": "3.7.7"
   },
   "pycharm": {
    "stem_cell": {
     "cell_type": "raw",
-    "source": [],
     "metadata": {
      "collapsed": false
-    }
+    },
+    "source": []
    }
   }
  },
  "nbformat": 4,
- "nbformat_minor": 0
+ "nbformat_minor": 4
 }