Merge pull request #140 from pyplati/more-fixes-to-cardiac-docs

A few minor fixes

examples/cardiac_segmentation.ipynb

@@ -6,6 +6,9 @@
    "source": [
     "# Cardiac Sub-Structure Segmentation Example\n",
     "\n",
+    "> Important: When running on Google Colab, ensure you change your runtime to use a GPU before\n",
+    "proceeding (Runtime->Change runtime type, select GPU as Hardware accelerator and press Save)\n",
+    "\n",
     "## Import Modules"
    ]
   },
@@ -19,14 +22,13 @@
     "    import platipy\n",
     "except:\n",
     "    !pip install git+https://github.com/pyplati/platipy.git\n",
+    "    !pip install nnunet\n",
     "    import platipy\n",
     "\n",
     "from matplotlib import pyplot as plt\n",
     "\n",
     "import SimpleITK as sitk\n",
     "\n",
-    "%matplotlib inline\n",
-    "\n",
     "from platipy.imaging.tests.data import get_lung_nifti\n",
     "from platipy.imaging.projects.cardiac.run import run_hybrid_segmentation\n",
     "from platipy.imaging import ImageVisualiser\n",
@@ -113,8 +115,7 @@
     "for struct_name in list(auto_structures.keys()):\n",
     "    sitk.WriteImage(auto_structures[struct_name], str(output_directory.joinpath(f\"{struct_name}.nii.gz\")))\n",
     "\n",
-    "print(f\"Segmentations saved to: {output_directory}\")\n",
-    "\n"
+    "print(f\"Segmentations saved to: {output_directory}\")"
    ]
   },
   {
@@ -134,17 +135,10 @@
    "source": [
     "vis = ImageVisualiser(test_image, cut=get_com(auto_structures[\"Heart\"]))\n",
     "vis.add_contour({struct: auto_structures[struct] for struct in auto_structures.keys()})\n",
-    "fig = vis.show()\n",
-    "plt.savefig(output_directory.joinpath(f\"snapshot.png\"))"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "fig"
+    "vis.show()\n",
+    "snapshot_path = output_directory.joinpath(f\"snapshot.png\")\n",
+    "plt.savefig(snapshot_path)\n",
+    "print(f\"Snapshot saved to: {snapshot_path}\")"
    ]
   },
   {

platipy/imaging/projects/cardiac/README.md

@@ -1,12 +1,12 @@
-# Cardiac sub-structure auto-segmentation
+# Cardiac Segmentation
 
 This project consists of an algorithm to automatically segment the heart plus 17 cardiac 
 sub-structures in typical Radiotherapy CT scans. The hybrid algorithm [[1]](#references) first segments the whole
 heart using a deep learning model (nnUNet [[2]](#references)) followed by a multi-atlas based mapping of the
-cardiac sub-structure [[3]](#references) and geometric modelling of smaller cardiac structures. The following
+cardiac sub-structures [[3]](#references) and finally geometric modelling of smaller cardiac structures. The following
 cardiac structures are generated by the auto-segmentation algorithm. By default, structure names
 are formatted to conform to the [TG-263](https://www.aapm.org/pubs/reports/RPT_263.pdf)
-guildelines:
+guidelines:
 
 | Structure                        | Name                |
 | -------------------------------- | ------------------- |
@@ -33,13 +33,18 @@ guildelines:
 
 ## Quickstart
 
-These steps have been tested on an Ubuntu 20.04 OS with Python 3.8. Other Linux Operating Systems
-and Python versions 3.7-3.9 are expected to work. This tool is untest and likely won't run on a
+These steps have been tested on an Ubuntu 20.04 operating system using Python 3.8. Other Linux operating systems
+and Python versions 3.7-3.9 are expected to work. This tool is untested and likely won't run on a
 Windows operating system at this time (contributions are welcome).
 
 ### System Requirements
 
-On a system with the following specification, one case  takes approximately 5 minutes process.
+On a system with the following specifications, one case takes approximately 5 minutes process:
+- 24 core Intel i9 processor
+- 64GB RAM
+- NVIDIA 3090 GPU
+
+The tool will run on a system with lower specifications, however a GPU is recommended.
 
 > Note: The nnUNet model as well as the cardiac atlas will be downloaded which will increase the
 runtime on the first run.
@@ -100,7 +105,7 @@ plt.savefig(output_directory.joinpath(f"snapshot.png"))
 
 ### Run from Command Line Interface (CLI)
 
-This command assumes you have a thorax CT image stored in Nifti format at a location
+This command assumes you have a thorax CT image stored in Nifti format at the location
 `/path/to/ct.nii.gz`:
 
 ```bash
@@ -136,7 +141,7 @@ library:
 !pip install platipy
 ```
 
-If you continute to experience issues, restart the Colab runtime (Runtime->Restart Runtime) and try
+If you continue to experience issues, restart the Colab runtime (Runtime->Restart Runtime) and try
 again.
 
 > If you are experiencing other issues, please consider reporting these in the [PlatiPy GitHub

platipy/imaging/projects/cardiac/run.py

@@ -1015,7 +1015,7 @@ def run_cardiac_segmentation(img, guide_structure=None, settings=CARDIAC_SETTING
         )
 
         # 5 - SINOATRIAL NODE
-        san_name = "CN_Sinoatrialz"
+        san_name = "CN_Sinoatrial"
         results[san_name] = geometric_sinoatrialnode(
             label_svc=results[geom_atlas_names["atlas_superior_vena_cava"]],
             label_ra=results[geom_atlas_names["atlas_right_atrium"]],