updates

Author: msneubauer

_sources/projects/Project_02_Radiotherapy.ipynb

@@ -12,7 +12,7 @@
    "metadata": {},
    "source": [
     "<div>\n",
-    "<img src=\"https://raw.githubusercontent.com/illinois-mlp/MachineLearningForPhysics/main/img/Project_Radiotherapy-overview.jpg\" width=600></img>\n",
+    "<img src=\"..\\img\\Radiotherapy-overview.jpg\" width=600>\n",
     "</div>"
    ]
   },
@@ -53,6 +53,14 @@
     "* https://github.com/ababier/open-kbp"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Link to data\n",
+    "* https://github.com/ababier/open-kbp"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -74,10 +82,9 @@
     "X-rays attenuate exponentially upon entering the body by the Beer-Lambert Law. \n",
     "\n",
     "$$\n",
-    "\n",
-    "I(x) = I_0 \\, e^{-\\mu x}\n",
-    "\n",
+    "I(x) = I_0 e^{-\\mu x}\n",
     "$$\n",
+    "\n",
     "$$\n",
     "\\begin{aligned}\n",
     "x   & : \\text{Depth in the body} \\\\\n",
@@ -140,7 +147,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "Download `openkbp_patient_data.zip` from File URLS above. Unzip it and upload the folder to your Google Colab (`/content/..`) or local directory. Note that you will need to adjust `base_path` to lead to your uploaded folder.\n",
+    "Run the cell below and unzip the contents of `openkbp_patient_data` into a folder. It should be present in Colab's (`/content/..`) or local directory. Note that you will need to adjust `base_path` to lead to this folder.\n",
     "\n",
     "Each patient hosts the following three files:  \n",
     "`ct.csv` - 3D grayscale CT scan of patient anatomy composed of 2D slices. Each voxel represents how much X-ray is absorbed by the tissue in Hounsfield Units (HU)  \n",
@@ -150,6 +157,16 @@
     "These are currently flattened 1D rows. The cell below converts the csv files into 3D volumes [64, 128, 128] where each point represents an intensity (for the ct scan) or binary value (for ptv & spinal cord). Add a line to the for loop that stacks `ct`, `ptv`, and `spine` into a single tensor. What's the final shape of this tensor? Draw an analogy to an RGB image."
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "!wget https://github.com/florilegium7/Physics-informed-DQN-Radiotherapy/releases/download/v1.0/openkbp_patient_data.zip\n",
+    "!unzip patient_data.zip"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -160,7 +177,7 @@
     "import pandas as pd\n",
     "import os\n",
     "\n",
-    "base_path = 'openkbp_patient_data' #adjust this (e.g. /content/openkbp_patient_data)\n",
+    "base_path = 'openkbp_patient_data' #you may adjust this (e.g. /content/openkbp_patient_data)\n",
     "\n",
     "patient_ids = [\"patient_1\", \"patient_2\", \"patient_3\",\"patient_5\",\"patient_7\",\"patient_9\", \"patient_10\", \"patient_12\" , \"patient_14\", \"patient_16\"]\n",
     "patient_tensors = []\n",
@@ -368,12 +385,12 @@
    "source": [
     "We'd like to inject a physics loss that encodes the actual attenuation and radiation dose delivery of the beam, rewarding correct dose distribution between the spine and tumor. We need to upgrade our simple Q-learning method into a Deep-Q Network (DQN). This should only require simple structural changes (refer to our reinforcement learning [notebook](https://illinois-mlp.github.io/MachineLearningForPhysics/_sources/lectures/ReinforcementLearning.html)). Typically, these would use a Bellman loss. With our new physics-loss it should resemble:\n",
     "$$\n",
-    "\\mathcal{L}_{\\text{Total}} = \\mathcal{L}_{\\text{Bellman}} + \\lambda_{\\text{Physics}} \\mathcal{L}_{\\text{Physics}}\n",
+    "L_{\\text{Total}} = L_{\\text{Bellman}} + \\lambda_{\\text{Physics}} L_{\\text{Physics}}\n",
     "$$\n",
     "\n",
     "The physics loss takes the following form: \n",
     "$$\n",
-    "\\mathcal{L}_{\\text{Physics}} =\n",
+    "L_{\\text{Physics}} =\n",
     "\\lambda_{\\text{PTV}} \\cdot \\text{Underdose}_{\\text{PTV}} +\n",
     "\\lambda_{\\text{Spine}} \\cdot \\text{Overdose}_{\\text{Spine}}\n",
     "$$\n",