Code/Image-Processing-and-Quantification/msd_pancreas_2D_3D_metrics_extraction.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "gather": {
          "logged": 1697511974920
        }
      },
      "outputs": [],
      "source": [
        "import os\n",
        "import nibabel as nib\n",
        "import numpy as np\n",
        "import skimage\n",
        "import SimpleITK as sitk\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "import cv2, pickle"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "gather": {
          "logged": 1697511976242
        },
        "jupyter": {
          "outputs_hidden": false,
          "source_hidden": false
        },
        "nteract": {
          "transient": {
            "deleting": false
          }
        }
      },
      "outputs": [],
      "source": [
        "# Prototype with one subject segmentation first\n",
        "print(os.getcwd())\n",
        "image_path = r'/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/imagesTr'\n",
        "tumor_path = r'/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr'\n",
        "\n",
        "# Organ segmentation files obtained from TotalSegmentator v1.5.7\n",
        "# Results are generated from the run_total_segmentator.py script\n",
        "label_organs_path = r'/home/azureuser/cloudfiles/code/Users/data/msd_segmentation/Task07_Pancreas/imagesTr'\n",
        "\n",
        "# List all the files in the image path and label path, exclude files that start with ._ \n",
        "image_list = [f for f in os.listdir(image_path) if not f.startswith('._') and  'nii.gz' in f]\n",
        "tumor_label_value = 2\n",
        "number_of_organs = 104 \n",
        "# Sort the list\n",
        "image_list.sort()\n",
        "print(len(image_list))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Saved file:  /home/azureuser/cloudfiles/code/Users/data/msd_quantification/msd_pancreas_metrics_2D_per_organ.pkl\n"
          ]
        }
      ],
      "source": [
        "image_id_dict ={}\n",
        "for image_name in image_list:\n",
        "    label_organ_file = os.path.join(label_organs_path, image_name)\n",
        "    label_organ = sitk.ReadImage(label_organ_file)\n",
        "    img_emb_list = []\n",
        "\n",
        "    # Get 3D matrix from the image\n",
        "    label_organ_data = sitk.GetArrayFromImage(label_organ)\n",
        "\n",
        "    for i in range(label_organ_data.shape[0]):\n",
        "        label_slice = label_organ_data[i,:,:]\n",
        "        # For label slice, create a dictionary where key is the pixel value and value is the number of pixels different from 0\n",
        "        label_dict = dict(zip(*np.unique(label_slice, return_counts=True)))\n",
        "        label_vector = np.zeros(104, dtype=int)\n",
        "        for key, value in label_dict.items():\n",
        "            if key > 0:\n",
        "                label_vector[key-1] = value\n",
        "\n",
        "        img_emb_list.append(label_vector)\n",
        "\n",
        "    # concatenate the vectors into a matrix\n",
        "    full_emb = np.stack(img_emb_list, axis=0)\n",
        "    image_id_dict[image_name] = full_emb.copy()\n",
        "\n",
        "output_file = '/home/azureuser/cloudfiles/code/Users/data/msd_quantification/msd_pancreas_metrics_2D_per_organ.pkl'\n",
        "with open(output_file, \"wb\") as f:\n",
        "    pickle.dump(image_id_dict, f)\n",
        "print(\"Saved file: \", output_file)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "gather": {
          "logged": 1697511981439
        }
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "dict_keys(['pancreas_001.nii.gz', 'pancreas_004.nii.gz', 'pancreas_005.nii.gz', 'pancreas_006.nii.gz', 'pancreas_010.nii.gz', 'pancreas_012.nii.gz', 'pancreas_015.nii.gz', 'pancreas_016.nii.gz', 'pancreas_018.nii.gz', 'pancreas_019.nii.gz', 'pancreas_021.nii.gz', 'pancreas_024.nii.gz', 'pancreas_025.nii.gz', 'pancreas_028.nii.gz', 'pancreas_029.nii.gz', 'pancreas_032.nii.gz', 'pancreas_035.nii.gz', 'pancreas_037.nii.gz', 'pancreas_040.nii.gz', 'pancreas_041.nii.gz', 'pancreas_042.nii.gz', 'pancreas_043.nii.gz', 'pancreas_045.nii.gz', 'pancreas_046.nii.gz', 'pancreas_048.nii.gz', 'pancreas_049.nii.gz', 'pancreas_050.nii.gz', 'pancreas_051.nii.gz', 'pancreas_052.nii.gz', 'pancreas_055.nii.gz', 'pancreas_056.nii.gz', 'pancreas_058.nii.gz', 'pancreas_061.nii.gz', 'pancreas_064.nii.gz', 'pancreas_066.nii.gz', 'pancreas_067.nii.gz', 'pancreas_069.nii.gz', 'pancreas_070.nii.gz', 'pancreas_071.nii.gz', 'pancreas_074.nii.gz', 'pancreas_075.nii.gz', 'pancreas_077.nii.gz', 'pancreas_078.nii.gz', 'pancreas_080.nii.gz', 'pancreas_081.nii.gz', 'pancreas_083.nii.gz', 'pancreas_084.nii.gz', 'pancreas_086.nii.gz', 'pancreas_087.nii.gz', 'pancreas_088.nii.gz', 'pancreas_089.nii.gz', 'pancreas_091.nii.gz', 'pancreas_092.nii.gz', 'pancreas_093.nii.gz', 'pancreas_094.nii.gz', 'pancreas_095.nii.gz', 'pancreas_096.nii.gz', 'pancreas_098.nii.gz', 'pancreas_099.nii.gz', 'pancreas_100.nii.gz', 'pancreas_101.nii.gz', 'pancreas_102.nii.gz', 'pancreas_103.nii.gz', 'pancreas_104.nii.gz', 'pancreas_105.nii.gz', 'pancreas_106.nii.gz', 'pancreas_107.nii.gz', 'pancreas_109.nii.gz', 'pancreas_110.nii.gz', 'pancreas_111.nii.gz', 'pancreas_113.nii.gz', 'pancreas_114.nii.gz', 'pancreas_117.nii.gz', 'pancreas_119.nii.gz', 'pancreas_120.nii.gz', 'pancreas_122.nii.gz', 'pancreas_124.nii.gz', 'pancreas_125.nii.gz', 'pancreas_126.nii.gz', 'pancreas_127.nii.gz', 'pancreas_129.nii.gz', 'pancreas_130.nii.gz', 'pancreas_131.nii.gz', 'pancreas_135.nii.gz', 'pancreas_137.nii.gz', 'pancreas_138.nii.gz', 'pancreas_140.nii.gz', 'pancreas_145.nii.gz', 'pancreas_147.nii.gz', 'pancreas_148.nii.gz', 'pancreas_149.nii.gz', 'pancreas_155.nii.gz', 'pancreas_157.nii.gz', 'pancreas_158.nii.gz', 'pancreas_159.nii.gz', 'pancreas_160.nii.gz', 'pancreas_165.nii.gz', 'pancreas_166.nii.gz', 'pancreas_167.nii.gz', 'pancreas_169.nii.gz', 'pancreas_170.nii.gz', 'pancreas_172.nii.gz', 'pancreas_173.nii.gz', 'pancreas_175.nii.gz', 'pancreas_178.nii.gz', 'pancreas_179.nii.gz', 'pancreas_180.nii.gz', 'pancreas_181.nii.gz', 'pancreas_182.nii.gz', 'pancreas_183.nii.gz', 'pancreas_186.nii.gz', 'pancreas_187.nii.gz', 'pancreas_191.nii.gz', 'pancreas_193.nii.gz', 'pancreas_194.nii.gz', 'pancreas_196.nii.gz', 'pancreas_197.nii.gz', 'pancreas_198.nii.gz', 'pancreas_199.nii.gz', 'pancreas_200.nii.gz', 'pancreas_201.nii.gz', 'pancreas_203.nii.gz', 'pancreas_204.nii.gz', 'pancreas_207.nii.gz', 'pancreas_209.nii.gz', 'pancreas_210.nii.gz', 'pancreas_211.nii.gz', 'pancreas_212.nii.gz', 'pancreas_213.nii.gz', 'pancreas_214.nii.gz', 'pancreas_215.nii.gz', 'pancreas_217.nii.gz', 'pancreas_218.nii.gz', 'pancreas_219.nii.gz', 'pancreas_222.nii.gz', 'pancreas_224.nii.gz', 'pancreas_225.nii.gz', 'pancreas_226.nii.gz', 'pancreas_227.nii.gz', 'pancreas_228.nii.gz', 'pancreas_229.nii.gz', 'pancreas_230.nii.gz', 'pancreas_231.nii.gz', 'pancreas_234.nii.gz', 'pancreas_235.nii.gz', 'pancreas_236.nii.gz', 'pancreas_239.nii.gz', 'pancreas_241.nii.gz', 'pancreas_242.nii.gz', 'pancreas_243.nii.gz', 'pancreas_244.nii.gz', 'pancreas_246.nii.gz', 'pancreas_247.nii.gz', 'pancreas_249.nii.gz', 'pancreas_253.nii.gz', 'pancreas_254.nii.gz', 'pancreas_255.nii.gz', 'pancreas_256.nii.gz', 'pancreas_258.nii.gz', 'pancreas_259.nii.gz', 'pancreas_261.nii.gz', 'pancreas_262.nii.gz', 'pancreas_264.nii.gz', 'pancreas_265.nii.gz', 'pancreas_266.nii.gz', 'pancreas_267.nii.gz', 'pancreas_268.nii.gz', 'pancreas_269.nii.gz', 'pancreas_270.nii.gz', 'pancreas_274.nii.gz', 'pancreas_275.nii.gz', 'pancreas_276.nii.gz', 'pancreas_277.nii.gz', 'pancreas_278.nii.gz', 'pancreas_279.nii.gz', 'pancreas_280.nii.gz', 'pancreas_283.nii.gz', 'pancreas_284.nii.gz', 'pancreas_285.nii.gz', 'pancreas_286.nii.gz', 'pancreas_287.nii.gz', 'pancreas_289.nii.gz', 'pancreas_290.nii.gz', 'pancreas_291.nii.gz', 'pancreas_292.nii.gz', 'pancreas_293.nii.gz', 'pancreas_294.nii.gz', 'pancreas_295.nii.gz', 'pancreas_296.nii.gz', 'pancreas_297.nii.gz', 'pancreas_298.nii.gz', 'pancreas_299.nii.gz', 'pancreas_300.nii.gz', 'pancreas_301.nii.gz', 'pancreas_302.nii.gz', 'pancreas_303.nii.gz', 'pancreas_304.nii.gz', 'pancreas_305.nii.gz', 'pancreas_308.nii.gz', 'pancreas_309.nii.gz', 'pancreas_310.nii.gz', 'pancreas_311.nii.gz', 'pancreas_312.nii.gz', 'pancreas_313.nii.gz', 'pancreas_315.nii.gz', 'pancreas_316.nii.gz', 'pancreas_318.nii.gz', 'pancreas_320.nii.gz', 'pancreas_321.nii.gz', 'pancreas_323.nii.gz', 'pancreas_325.nii.gz', 'pancreas_326.nii.gz', 'pancreas_327.nii.gz', 'pancreas_328.nii.gz', 'pancreas_329.nii.gz', 'pancreas_330.nii.gz', 'pancreas_331.nii.gz', 'pancreas_333.nii.gz', 'pancreas_334.nii.gz', 'pancreas_336.nii.gz', 'pancreas_339.nii.gz', 'pancreas_342.nii.gz', 'pancreas_343.nii.gz', 'pancreas_344.nii.gz', 'pancreas_345.nii.gz', 'pancreas_346.nii.gz', 'pancreas_347.nii.gz', 'pancreas_348.nii.gz', 'pancreas_350.nii.gz', 'pancreas_351.nii.gz', 'pancreas_354.nii.gz', 'pancreas_355.nii.gz', 'pancreas_356.nii.gz', 'pancreas_357.nii.gz', 'pancreas_358.nii.gz', 'pancreas_360.nii.gz', 'pancreas_361.nii.gz', 'pancreas_362.nii.gz', 'pancreas_364.nii.gz', 'pancreas_365.nii.gz', 'pancreas_366.nii.gz', 'pancreas_367.nii.gz', 'pancreas_369.nii.gz', 'pancreas_370.nii.gz', 'pancreas_372.nii.gz', 'pancreas_374.nii.gz', 'pancreas_375.nii.gz', 'pancreas_376.nii.gz', 'pancreas_377.nii.gz', 'pancreas_378.nii.gz', 'pancreas_379.nii.gz', 'pancreas_380.nii.gz', 'pancreas_382.nii.gz', 'pancreas_385.nii.gz', 'pancreas_386.nii.gz', 'pancreas_387.nii.gz', 'pancreas_388.nii.gz', 'pancreas_389.nii.gz', 'pancreas_391.nii.gz', 'pancreas_392.nii.gz', 'pancreas_393.nii.gz', 'pancreas_395.nii.gz', 'pancreas_398.nii.gz', 'pancreas_399.nii.gz', 'pancreas_400.nii.gz', 'pancreas_401.nii.gz', 'pancreas_402.nii.gz', 'pancreas_404.nii.gz', 'pancreas_405.nii.gz', 'pancreas_406.nii.gz', 'pancreas_409.nii.gz', 'pancreas_410.nii.gz', 'pancreas_411.nii.gz', 'pancreas_412.nii.gz', 'pancreas_413.nii.gz', 'pancreas_414.nii.gz', 'pancreas_415.nii.gz', 'pancreas_416.nii.gz', 'pancreas_418.nii.gz', 'pancreas_419.nii.gz', 'pancreas_421.nii.gz'])"
            ]
          },
          "execution_count": 4,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "output_file = '/home/azureuser/cloudfiles/code/Users/data/msd_quantification/msd_pancreas_metrics_2D_per_organ.pkl'\n",
        "# Load pickle file\n",
        "with open(output_file, \"rb\") as f:\n",
        "    data_dict = pickle.load(f)\n",
        "data_dict.keys()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "gather": {
          "logged": 1697505031703
        }
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "pancreas_001.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_001.nii.gz 1\n",
            "pancreas_004.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_004.nii.gz 1\n",
            "pancreas_005.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_005.nii.gz 1\n",
            "pancreas_006.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_006.nii.gz 1\n",
            "pancreas_010.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_010.nii.gz 1\n",
            "pancreas_012.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_012.nii.gz 1\n",
            "pancreas_015.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_015.nii.gz 1\n",
            "pancreas_016.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_016.nii.gz 1\n",
            "pancreas_018.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_018.nii.gz 1\n",
            "pancreas_019.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_019.nii.gz 1\n",
            "pancreas_021.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_021.nii.gz 2\n",
            "pancreas_024.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_024.nii.gz 1\n",
            "pancreas_025.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_025.nii.gz 1\n",
            "pancreas_028.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_028.nii.gz 1\n",
            "pancreas_029.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_029.nii.gz 1\n",
            "pancreas_032.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_032.nii.gz 1\n",
            "pancreas_035.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_035.nii.gz 1\n",
            "pancreas_037.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_037.nii.gz 1\n",
            "pancreas_040.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_040.nii.gz 1\n",
            "pancreas_041.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_041.nii.gz 1\n",
            "pancreas_042.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_042.nii.gz 1\n",
            "pancreas_043.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_043.nii.gz 1\n",
            "pancreas_045.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_045.nii.gz 1\n",
            "pancreas_046.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_046.nii.gz 1\n",
            "pancreas_048.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_048.nii.gz 1\n",
            "pancreas_049.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_049.nii.gz 1\n",
            "pancreas_050.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_050.nii.gz 1\n",
            "pancreas_051.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_051.nii.gz 1\n",
            "pancreas_052.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_052.nii.gz 1\n",
            "pancreas_055.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_055.nii.gz 1\n",
            "pancreas_056.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_056.nii.gz 1\n",
            "pancreas_058.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_058.nii.gz 1\n",
            "pancreas_061.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_061.nii.gz 1\n",
            "pancreas_064.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_064.nii.gz 1\n",
            "pancreas_066.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_066.nii.gz 1\n",
            "pancreas_067.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_067.nii.gz 1\n",
            "pancreas_069.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_069.nii.gz 1\n",
            "pancreas_070.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_070.nii.gz 1\n",
            "pancreas_071.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_071.nii.gz 1\n",
            "pancreas_074.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_074.nii.gz 1\n",
            "pancreas_075.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_075.nii.gz 1\n",
            "pancreas_077.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_077.nii.gz 1\n",
            "pancreas_078.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_078.nii.gz 1\n",
            "pancreas_080.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_080.nii.gz 1\n",
            "pancreas_081.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_081.nii.gz 1\n",
            "pancreas_083.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_083.nii.gz 1\n",
            "pancreas_084.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_084.nii.gz 1\n",
            "pancreas_086.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_086.nii.gz 1\n",
            "pancreas_087.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_087.nii.gz 1\n",
            "pancreas_088.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_088.nii.gz 1\n",
            "pancreas_089.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_089.nii.gz 1\n",
            "pancreas_091.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_091.nii.gz 1\n",
            "pancreas_092.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_092.nii.gz 1\n",
            "pancreas_093.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_093.nii.gz 1\n",
            "pancreas_094.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_094.nii.gz 1\n",
            "pancreas_095.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_095.nii.gz 1\n",
            "pancreas_096.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_096.nii.gz 1\n",
            "pancreas_098.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_098.nii.gz 1\n",
            "pancreas_099.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_099.nii.gz 1\n",
            "pancreas_100.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_100.nii.gz 1\n",
            "pancreas_101.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_101.nii.gz 1\n",
            "pancreas_102.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_102.nii.gz 1\n",
            "pancreas_103.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_103.nii.gz 1\n",
            "pancreas_104.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_104.nii.gz 1\n",
            "pancreas_105.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_105.nii.gz 1\n",
            "pancreas_106.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_106.nii.gz 1\n",
            "pancreas_107.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_107.nii.gz 1\n",
            "pancreas_109.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_109.nii.gz 1\n",
            "pancreas_110.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_110.nii.gz 1\n",
            "pancreas_111.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_111.nii.gz 1\n",
            "pancreas_113.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_113.nii.gz 1\n",
            "pancreas_114.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_114.nii.gz 1\n",
            "pancreas_117.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_117.nii.gz 1\n",
            "pancreas_119.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_119.nii.gz 1\n",
            "pancreas_120.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_120.nii.gz 1\n",
            "pancreas_122.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_122.nii.gz 1\n",
            "pancreas_124.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_124.nii.gz 1\n",
            "pancreas_125.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_125.nii.gz 1\n",
            "pancreas_126.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_126.nii.gz 1\n",
            "pancreas_127.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_127.nii.gz 1\n",
            "pancreas_129.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_129.nii.gz 1\n",
            "pancreas_130.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_130.nii.gz 1\n",
            "pancreas_131.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_131.nii.gz 1\n",
            "pancreas_135.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_135.nii.gz 1\n",
            "pancreas_137.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_137.nii.gz 1\n",
            "pancreas_138.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_138.nii.gz 1\n",
            "pancreas_140.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_140.nii.gz 1\n",
            "pancreas_145.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_145.nii.gz 1\n",
            "pancreas_147.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_147.nii.gz 1\n",
            "pancreas_148.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_148.nii.gz 1\n",
            "pancreas_149.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_149.nii.gz 1\n",
            "pancreas_155.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_155.nii.gz 1\n",
            "pancreas_157.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_157.nii.gz 1\n",
            "pancreas_158.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_158.nii.gz 1\n",
            "pancreas_159.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_159.nii.gz 1\n",
            "pancreas_160.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_160.nii.gz 1\n",
            "pancreas_165.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_165.nii.gz 1\n",
            "pancreas_166.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_166.nii.gz 1\n",
            "pancreas_167.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_167.nii.gz 1\n",
            "pancreas_169.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_169.nii.gz 1\n",
            "pancreas_170.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_170.nii.gz 1\n",
            "pancreas_172.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_172.nii.gz 1\n",
            "pancreas_173.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_173.nii.gz 1\n",
            "pancreas_175.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_175.nii.gz 1\n",
            "pancreas_178.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_178.nii.gz 1\n",
            "pancreas_179.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_179.nii.gz 1\n",
            "pancreas_180.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_180.nii.gz 1\n",
            "pancreas_181.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_181.nii.gz 1\n",
            "pancreas_182.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_182.nii.gz 1\n",
            "pancreas_183.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_183.nii.gz 1\n",
            "pancreas_186.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_186.nii.gz 1\n",
            "pancreas_187.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_187.nii.gz 1\n",
            "pancreas_191.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_191.nii.gz 1\n",
            "pancreas_193.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_193.nii.gz 1\n",
            "pancreas_194.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_194.nii.gz 1\n",
            "pancreas_196.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_196.nii.gz 1\n",
            "pancreas_197.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_197.nii.gz 1\n",
            "pancreas_198.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_198.nii.gz 1\n",
            "pancreas_199.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_199.nii.gz 1\n",
            "pancreas_200.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_200.nii.gz 1\n",
            "pancreas_201.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_201.nii.gz 1\n",
            "pancreas_203.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_203.nii.gz 1\n",
            "pancreas_204.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_204.nii.gz 1\n",
            "pancreas_207.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_207.nii.gz 1\n",
            "pancreas_209.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_209.nii.gz 1\n",
            "pancreas_210.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_210.nii.gz 1\n",
            "pancreas_211.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_211.nii.gz 1\n",
            "pancreas_212.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_212.nii.gz 1\n",
            "pancreas_213.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_213.nii.gz 1\n",
            "pancreas_214.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_214.nii.gz 1\n",
            "pancreas_215.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_215.nii.gz 1\n",
            "pancreas_217.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_217.nii.gz 1\n",
            "pancreas_218.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_218.nii.gz 1\n",
            "pancreas_219.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_219.nii.gz 1\n",
            "pancreas_222.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_222.nii.gz 1\n",
            "pancreas_224.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_224.nii.gz 1\n",
            "pancreas_225.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_225.nii.gz 1\n",
            "pancreas_226.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_226.nii.gz 1\n",
            "pancreas_227.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_227.nii.gz 1\n",
            "pancreas_228.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_228.nii.gz 1\n",
            "pancreas_229.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_229.nii.gz 1\n",
            "pancreas_230.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_230.nii.gz 1\n",
            "pancreas_231.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_231.nii.gz 1\n",
            "pancreas_234.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_234.nii.gz 1\n",
            "pancreas_235.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_235.nii.gz 1\n",
            "pancreas_236.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_236.nii.gz 1\n",
            "pancreas_239.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_239.nii.gz 1\n",
            "pancreas_241.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_241.nii.gz 1\n",
            "pancreas_242.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_242.nii.gz 1\n",
            "pancreas_243.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_243.nii.gz 1\n",
            "pancreas_244.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_244.nii.gz 1\n",
            "pancreas_246.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_246.nii.gz 1\n",
            "pancreas_247.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_247.nii.gz 1\n",
            "pancreas_249.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_249.nii.gz 1\n",
            "pancreas_253.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_253.nii.gz 1\n",
            "pancreas_254.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_254.nii.gz 1\n",
            "pancreas_255.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_255.nii.gz 1\n",
            "pancreas_256.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_256.nii.gz 1\n",
            "pancreas_258.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_258.nii.gz 1\n",
            "pancreas_259.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_259.nii.gz 1\n",
            "pancreas_261.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_261.nii.gz 1\n",
            "pancreas_262.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_262.nii.gz 1\n",
            "pancreas_264.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_264.nii.gz 1\n",
            "pancreas_265.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_265.nii.gz 1\n",
            "pancreas_266.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_266.nii.gz 1\n",
            "pancreas_267.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_267.nii.gz 1\n",
            "pancreas_268.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_268.nii.gz 1\n",
            "pancreas_269.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_269.nii.gz 1\n",
            "pancreas_270.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_270.nii.gz 1\n",
            "pancreas_274.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_274.nii.gz 2\n",
            "pancreas_275.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_275.nii.gz 1\n",
            "pancreas_276.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_276.nii.gz 1\n",
            "pancreas_277.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_277.nii.gz 1\n",
            "pancreas_278.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_278.nii.gz 1\n",
            "pancreas_279.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_279.nii.gz 1\n",
            "pancreas_280.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_280.nii.gz 1\n",
            "pancreas_283.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_283.nii.gz 1\n",
            "pancreas_284.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_284.nii.gz 1\n",
            "pancreas_285.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_285.nii.gz 1\n",
            "pancreas_286.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_286.nii.gz 1\n",
            "pancreas_287.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_287.nii.gz 1\n",
            "pancreas_289.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_289.nii.gz 1\n",
            "pancreas_290.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_290.nii.gz 1\n",
            "pancreas_291.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_291.nii.gz 1\n",
            "pancreas_292.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_292.nii.gz 1\n",
            "pancreas_293.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_293.nii.gz 1\n",
            "pancreas_294.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_294.nii.gz 1\n",
            "pancreas_295.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_295.nii.gz 1\n",
            "pancreas_296.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_296.nii.gz 1\n",
            "pancreas_297.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_297.nii.gz 1\n",
            "pancreas_298.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_298.nii.gz 1\n",
            "pancreas_299.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_299.nii.gz 1\n",
            "pancreas_300.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_300.nii.gz 1\n",
            "pancreas_301.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_301.nii.gz 1\n",
            "pancreas_302.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_302.nii.gz 1\n",
            "pancreas_303.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_303.nii.gz 1\n",
            "pancreas_304.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_304.nii.gz 1\n",
            "pancreas_305.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_305.nii.gz 1\n",
            "pancreas_308.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_308.nii.gz 1\n",
            "pancreas_309.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_309.nii.gz 1\n",
            "pancreas_310.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_310.nii.gz 1\n",
            "pancreas_311.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_311.nii.gz 1\n",
            "pancreas_312.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_312.nii.gz 1\n",
            "pancreas_313.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_313.nii.gz 1\n",
            "pancreas_315.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_315.nii.gz 1\n",
            "pancreas_316.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_316.nii.gz 1\n",
            "pancreas_318.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_318.nii.gz 1\n",
            "pancreas_320.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_320.nii.gz 1\n",
            "pancreas_321.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_321.nii.gz 1\n",
            "pancreas_323.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_323.nii.gz 1\n",
            "pancreas_325.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_325.nii.gz 1\n",
            "pancreas_326.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_326.nii.gz 1\n",
            "pancreas_327.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_327.nii.gz 1\n",
            "pancreas_328.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_328.nii.gz 1\n",
            "pancreas_329.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_329.nii.gz 1\n",
            "pancreas_330.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_330.nii.gz 1\n",
            "pancreas_331.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_331.nii.gz 1\n",
            "pancreas_333.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_333.nii.gz 1\n",
            "pancreas_334.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_334.nii.gz 1\n",
            "pancreas_336.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_336.nii.gz 1\n",
            "pancreas_339.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_339.nii.gz 1\n",
            "pancreas_342.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_342.nii.gz 1\n",
            "pancreas_343.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_343.nii.gz 1\n",
            "pancreas_344.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_344.nii.gz 1\n",
            "pancreas_345.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_345.nii.gz 1\n",
            "pancreas_346.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_346.nii.gz 1\n",
            "pancreas_347.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_347.nii.gz 1\n",
            "pancreas_348.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_348.nii.gz 1\n",
            "pancreas_350.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_350.nii.gz 1\n",
            "pancreas_351.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_351.nii.gz 1\n",
            "pancreas_354.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_354.nii.gz 1\n",
            "pancreas_355.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_355.nii.gz 1\n",
            "pancreas_356.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_356.nii.gz 1\n",
            "pancreas_357.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_357.nii.gz 1\n",
            "pancreas_358.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_358.nii.gz 1\n",
            "pancreas_360.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_360.nii.gz 1\n",
            "pancreas_361.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_361.nii.gz 1\n",
            "pancreas_362.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_362.nii.gz 1\n",
            "pancreas_364.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_364.nii.gz 1\n",
            "pancreas_365.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_365.nii.gz 1\n",
            "pancreas_366.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_366.nii.gz 1\n",
            "pancreas_367.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_367.nii.gz 1\n",
            "pancreas_369.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_369.nii.gz 1\n",
            "pancreas_370.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_370.nii.gz 1\n",
            "pancreas_372.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_372.nii.gz 1\n",
            "pancreas_374.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_374.nii.gz 1\n",
            "pancreas_375.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_375.nii.gz 1\n",
            "pancreas_376.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_376.nii.gz 1\n",
            "pancreas_377.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_377.nii.gz 1\n",
            "pancreas_378.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_378.nii.gz 1\n",
            "pancreas_379.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_379.nii.gz 1\n",
            "pancreas_380.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_380.nii.gz 1\n",
            "pancreas_382.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_382.nii.gz 1\n",
            "pancreas_385.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_385.nii.gz 1\n",
            "pancreas_386.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_386.nii.gz 1\n",
            "pancreas_387.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_387.nii.gz 1\n",
            "pancreas_388.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_388.nii.gz 1\n",
            "pancreas_389.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_389.nii.gz 1\n",
            "pancreas_391.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_391.nii.gz 1\n",
            "pancreas_392.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_392.nii.gz 1\n",
            "pancreas_393.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_393.nii.gz 1\n",
            "pancreas_395.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_395.nii.gz 1\n",
            "pancreas_398.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_398.nii.gz 1\n",
            "pancreas_399.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_399.nii.gz 1\n",
            "pancreas_400.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_400.nii.gz 1\n",
            "pancreas_401.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_401.nii.gz 1\n",
            "pancreas_402.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_402.nii.gz 1\n",
            "pancreas_404.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_404.nii.gz 1\n",
            "pancreas_405.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_405.nii.gz 1\n",
            "pancreas_406.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_406.nii.gz 1\n",
            "pancreas_409.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_409.nii.gz 1\n",
            "pancreas_410.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_410.nii.gz 1\n",
            "pancreas_411.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_411.nii.gz 1\n",
            "pancreas_412.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_412.nii.gz 1\n",
            "pancreas_413.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_413.nii.gz 1\n",
            "pancreas_414.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_414.nii.gz 1\n",
            "pancreas_415.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_415.nii.gz 1\n",
            "pancreas_416.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_416.nii.gz 1\n",
            "pancreas_418.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_418.nii.gz 1\n",
            "pancreas_419.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_419.nii.gz 1\n",
            "pancreas_421.nii.gz\n",
            "/home/azureuser/cloudfiles/code/Users/data/msd/Task07_Pancreas/labelsTr/pancreas_421.nii.gz 1\n"
          ]
        }
      ],
      "source": [
        "import tqdm\n",
        "\n",
        "# Create a dataframe to store the metrics per file and 2D image\n",
        "df2d = pd.DataFrame(columns=['filename', 'slice', 'cancer_stage', 'tumor_area'] )\n",
        "df3d = pd.DataFrame(columns=['filename','cancer_stage'])\n",
        "\n",
        "# #iterate through the list of files image_list and label_list\n",
        "# for image_name, label_name in tqdm.tqdm(zip(image_list, label_list)):\n",
        "(stats_list_2d, stats_list_3d) = ([], [])\n",
        "for image_name in image_list: \n",
        "#for image_name in tqdm.tqdm(image_list):\n",
        "    print(image_name)\n",
        "\n",
        "    (image_file, tumor_file, organs_file) = (os.path.join(image_path, image_name), os.path.join(tumor_path, image_name), os.path.join(label_organs_path, image_name))\n",
        "    (image, tumor, organs) = (sitk.ReadImage(image_file), sitk.ReadImage(tumor_file), sitk.ReadImage(organs_file))\n",
        "\n",
        "    # Cast tumor to int8 using CastImageFilter to uint8\n",
        "    castFilter = sitk.CastImageFilter()\n",
        "    castFilter.SetOutputPixelType(sitk.sitkUInt8)\n",
        "    tumor = castFilter.Execute(tumor)\n",
        "\n",
        "    (image_data, tumor_data, organs_data) = (sitk.GetArrayFromImage(image), sitk.GetArrayFromImage(tumor), sitk.GetArrayFromImage(organs))\n",
        "    spacing = image.GetSpacing()\n",
        "\n",
        "    # For tumor image make zero voxels except for tumor label, pancreas seems the exception from other msd cohorts\n",
        "    tumor_data[tumor_data != tumor_label_value] = 0\n",
        "    tumor_data[tumor_data == tumor_label_value] = 1\n",
        "\n",
        "    #Get pixel type for organs\n",
        "    pixel_type = tumor.GetPixelIDTypeAsString()\n",
        "    tumor = sitk.GetImageFromArray(tumor_data)\n",
        "\n",
        "    cc = sitk.ConnectedComponent(tumor)\n",
        "    stats = sitk.LabelIntensityStatisticsImageFilter()\n",
        "    stats.Execute(cc, tumor)\n",
        "    print(tumor_file, stats.GetNumberOfLabels())\n",
        "    # Get array of connected component numbers\n",
        "    cc_tumor_array = sitk.GetArrayFromImage(cc)\n",
        "\n",
        "    # Creates dictionary where key is tuple consisting of subject ID and label number\n",
        "    #tumor_size_list = []\n",
        "    for l in stats.GetLabels():\n",
        "        # Ignore if number of voxels is less than 6\n",
        "\n",
        "        # Remove spurius or flat objects objects \n",
        "        if stats.GetNumberOfPixels(l) < 3 or stats.GetBoundingBox(l)[-1] == 1:\n",
        "            cc_tumor_array[cc_tumor_array == l] = 0\n",
        "            continue\n",
        "\n",
        "        # Count number of voxels in this connected component\n",
        "        voxel_count = np.sum(cc_tumor_array == l)\n",
        "        #Initialize dictionary containing different metrics\n",
        "        cc_dictionary = {}\n",
        "        cc_dictionary['image_name'] = image_name\n",
        "        cc_dictionary['PhysicalSize_cm3'] = stats.GetPhysicalSize(l) * spacing[0] * spacing[1] * spacing[2] / 1000\n",
        "        cc_dictionary['NumberOfVoxels'] = stats.GetNumberOfPixels(l)\n",
        "        cc_dictionary['Roundness'] = stats.GetRoundness(l)\n",
        "        cc_dictionary['Elongation'] = stats.GetElongation(l)\n",
        "        cc_dictionary['Flatness'] = stats.GetFlatness(l)\n",
        "        cc_dictionary['EquivalentSphericalRadius_cm'] = stats.GetEquivalentSphericalRadius(l)/10\n",
        "        cc_dictionary['EquivalentSphericalPerimeter_cm'] = stats.GetEquivalentSphericalPerimeter(l)/10\n",
        "        cc_dictionary['EquivalentEllipsoidDiameter'] = stats.GetEquivalentEllipsoidDiameter(l)\n",
        "        cc_dictionary['major_axis_length_cm'] = max(stats.GetEquivalentEllipsoidDiameter(l)) / 10\n",
        "        cc_dictionary['centroid'] = stats.GetCentroid(l)\n",
        "        cc_dictionary['bounding_box'] = stats.GetBoundingBox(l)\n",
        "        cc_dictionary['spacing']= spacing\n",
        "\n",
        "        # Get the number of voxels in the tumor that are in the organ\n",
        "        tumor_bbox = stats.GetBoundingBox(l)\n",
        "        # Get voxels only in the tumor bounding box\n",
        "        (x0,y0,z0,x1,y1,z1)= tumor_bbox\n",
        "        # Add the width, height and depth to the x0, y0, z0\n",
        "        (x1,y1,z1) = (x1+x0, y1+y0, z1+z0)\n",
        "        # Coordinate systems ITK - Numpy are different:\n",
        "        (organ_bbox_array, tumor_bbox_array)  = (organs_data[z0:z1, y0:y1, x0:x1], cc_tumor_array[z0:z1, y0:y1, x0:x1])\n",
        "\n",
        "        # Quantifies tumor in each organ\n",
        "        for i in range(number_of_organs+1):\n",
        "            cc_dictionary['organ_'+str(i)+'_voxel_count'] = 0\n",
        "            # organ_voxel_count = np.sum(np.logical_and(cc_tumor_array == l, organs_data == i))\n",
        "\n",
        "        (voxel_values, voxel_counts) = np.unique(organ_bbox_array[tumor_bbox_array == l], return_counts=True)\n",
        "        for (v,c) in zip(voxel_values, voxel_counts):\n",
        "            cc_dictionary['organ_'+str(v)+'_voxel_count'] = c\n",
        "        stats_list_3d.append(cc_dictionary.copy())\n",
        "\n",
        "    # 2D Report -- Iterate through the slices\n",
        "    for i in range(organs_data.shape[0]):\n",
        "        tumor_slice = cc_tumor_array[i,:,:]\n",
        "        organs_slice = organs_data[i,:,:]\n",
        "\n",
        "        if not np.any(tumor_slice):\n",
        "            continue\n",
        "\n",
        "        number_of_tumors = 0\n",
        "        # OpenCV returns number of objects + 1\n",
        "        number_of_tumors = cv2.connectedComponents(tumor_slice.astype('uint8'),connectivity=8)[0] - 1\n",
        "\n",
        "        # Calculate area of the tumors\n",
        "        tumor_area_cm2 = np.sum(tumor_slice == 1) * spacing[0] * spacing[1] / 100\n",
        "\n",
        "        # Localize where in the organs the tumor is\n",
        "        organ_voxel_count_2d_dict = {}\n",
        "        organ_voxel_count_2d_dict['filename']= image_name\n",
        "        organ_voxel_count_2d_dict['slice']= i\n",
        "        organ_voxel_count_2d_dict['number_of_tumors']= number_of_tumors\n",
        "        organ_voxel_count_2d_dict['tumor_area_cm2']= tumor_area_cm2\n",
        "        organ_voxel_count_2d_dict['spacing'] = spacing\n",
        "        for i in range(number_of_organs+1):\n",
        "            organ_voxel_count = np.sum(np.logical_and(tumor_slice > 0, organs_slice == i))\n",
        "            organ_voxel_count_2d_dict['organ_'+str(i)+'_voxel_count'] = organ_voxel_count\n",
        "\n",
        "        stats_list_2d.append(organ_voxel_count_2d_dict.copy())\n",
        "\n",
        "df_tumor_report_2d = pd.DataFrame(stats_list_2d)\n",
        "df_tumor_report_3d = pd.DataFrame(stats_list_3d)\n",
        "\n",
        "df_tumor_report_3d.to_csv('/home/azureuser/cloudfiles/code/Users/data/msd_quantification/msd_pancreas_metrics_3D_tumor_per_organ.csv', index=False)\n",
        "df_tumor_report_2d.to_csv('/home/azureuser/cloudfiles/code/Users/data/msd_quantification/msd_pancreas_metrics_2D_tumor_per_organ.csv', index=False)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "gather": {
          "logged": 1697511987091
        }
      },
      "outputs": [],
      "source": [
        "# Read the data frames\n",
        "df3d = pd.read_csv('/home/azureuser/cloudfiles/code/Users/data/msd_quantification/msd_pancreas_metrics_3D_tumor_per_organ.csv')\n",
        "df2d = pd.read_csv('/home/azureuser/cloudfiles/code/Users/data/msd_quantification/msd_pancreas_metrics_2D_tumor_per_organ.csv')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "gather": {
          "logged": 1697512281696
        }
      },
      "outputs": [],
      "source": [
        "image_name_list_with_tumor = df3d.image_name.unique().tolist()\n",
        "image_name_list_all_subjects = [f for f in data_dict.keys()]\n",
        "\n",
        "subject_list = []\n",
        "# Group by image name and iterate through the groups\n",
        "for name, group in df3d.groupby('image_name'):\n",
        "    subject_dict = {}\n",
        "    # Number of tumors\n",
        "    number_tumors_group = len(group)    \n",
        "    max_tumor_size_group = max(group.major_axis_length_cm)\n",
        "    cancer_stage = -1 # No Cancer\n",
        "    # Get cancer stage\n",
        "    # For details see: https://www.cancerresearchuk.org/about-cancer/pancreatic-cancer/stages-types-grades/tnm-staging\n",
        "    # A very rough and oversimplified approximation using only tumor size, consider the following hypothetical rule:\n",
        "    # Tumor size ≤ 2 cm: Approximate as Stage I (T1)\n",
        "    # Tumor size > 2 cm and ≤ 4 cm: Approximate as Stage IIa (T2)\n",
        "    # Tumor size > 4 cm and ≤ 6 cm: Approximate as Stage IIb (T3)\n",
        "    # T3 means the cancer is bigger than 4cm but is still within the pancreas. (T4)\n",
        "    # Please note that T staging also considers depth of invasion into the pancreas wall and surrounding structures, not just diameter size.\n",
        "    # Here we will use fictional staging for 1, 2, 3\n",
        "\n",
        "    # Stage 1\n",
        "    if max_tumor_size_group < 2:\n",
        "        cancer_stage = 1\n",
        "    # Stage 2\n",
        "    elif  max_tumor_size_group >= 2 and max_tumor_size_group < 4:\n",
        "        cancer_stage = 2\n",
        "    # Stage 3\n",
        "    elif  max_tumor_size_group >= 4 and max_tumor_size_group < 6:\n",
        "        cancer_stage = 3\n",
        "    else:\n",
        "    # Stage 4\n",
        "        cancer_stage = 4\n",
        "\n",
        "\n",
        "    subject_dict['name'] = name\n",
        "    subject_dict['max_tumor_size'] = max_tumor_size_group\n",
        "    subject_dict['number_tumors'] = number_tumors_group\n",
        "    subject_dict['cancer_stage'] = cancer_stage\n",
        "    \n",
        "    subject_list.append(subject_dict.copy())\n",
        "\n",
        "df_subject = pd.DataFrame(subject_list)\n",
        "df_subject.to_csv('/home/azureuser/cloudfiles/code/Users/data/msd_quantification/msd_pancreas_metrics_3D_subject.csv', index=False)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "gather": {
          "logged": 1697512575534
        }
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7fa84d4cf5b0>"
            ]
          },
          "execution_count": 13,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "# Distribution of tumor vs organ number where values are non-zero\n",
        "df3d_organ = df3d.loc[:, df3d.columns.str.startswith('organ')]\n",
        "# Remove columns with all zeros\n",
        "df3d_organ = df3d_organ.loc[:, (df3d_organ != 0).any(axis=0)]\n",
        "\n",
        "# # Remove columns with less than 400 voxel counts\n",
        "df3d_organ = df3d_organ.loc[:, df3d_organ.sum() > 40000]\n",
        "\n",
        "# print(df3d_organ.sum())\n",
        "\n",
        "df3d_organ.sum().plot(kind='bar')\n",
        "#10\tpancreas\n",
        "\n",
        "# Rename columns as follows: \n",
        "# organ 0 is backbround\n",
        "# organ 3 is kidney_left\t\n",
        "# organ 6\tstomach\n",
        "# organ 10\tpancreas\n",
        "df3d_organ.columns = ['background', 'kidney left', 'stomach', 'pancreas']\n",
        "df3d_organ.sum().plot(kind='bar')\n",
        "\n",
        "# Normalize the number of voxels as cm3\n",
        "# df3d_organ = df3d_organ.div(df3d_organ.sum(axis=1), axis=0)\n",
        "\n",
        "# Plot histogram of number of voxels per organ normalized by total number of voxels\n",
        "#df3d_organ.sum().plot(kind='bar', title='Percentage of voxel tumor count per organ')\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "gather": {
          "logged": 1697512579558
        }
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "name              pancreas_415.nii.gz\n",
              "max_tumor_size                13.9619\n",
              "number_tumors                       1\n",
              "cancer_stage                        4\n",
              "Name: 276, dtype: object"
            ]
          },
          "execution_count": 14,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjIAAAHHCAYAAACle7JuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAAPYQAAD2EBqD+naQAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOzdeVxU9f4/8NewDfsgMIAjhLjmHmqaoohlcksty67K9RqaZjdNNEuLFteSbNHKzLRSrtS10tT6WrklSq6hguUOgsTV2FR2RITP749+zHVgmM3ZDryej8c8Hsznc+ac9xwOw2vO+ZxzZEIIASIiIiIJcrB1AURERESmYpAhIiIiyWKQISIiIslikCEiIiLJYpAhIiIiyWKQISIiIslikCEiIiLJYpAhIiIiyWKQISIiIslikJGYhQsXQiaTWWVZUVFRiIqKUj/ft28fZDIZNm/ebJXlT5o0CW3btrXKskxVXl6OqVOnIigoCDKZDLNnz7Z1SUSSdOnSJchkMiQmJlp1uRkZGRg+fDgUCgVkMhm2bdtm1eVr07ZtW0yaNMnWZUgGg4wNJSYmQiaTqR+urq5QqVSIjo7Ghx9+iLKyMrMs58qVK1i4cCHS09PNMj9zsufaDLF06VIkJibi2WefRVJSEiZOnGjrkiRn6dKldvHPw9YOHTqEhQsXori42NaltCixsbH4/fff8eabbyIpKQl9+/a12LLqw9q7775rsWW0SIJsZv369QKAWLx4sUhKShLr1q0TS5cuFcOHDxcymUyEhoaKkydParympqZGVFVVGbWc1NRUAUCsX7/eqNdVV1eL6upq9fPk5GQBQGzatMmo+Zha282bN8WNGzfMtixL6N+/v4iIiLB1GZLm4eEhYmNjbV2Gzb3zzjsCgMjOzrZ1KTZRV1cnqqqqxK1bt6y2zMrKSgFAvPrqq1ZZXnZ2tgAg3nnnHZ3T3bhxQ9y8edMqNTUHTjbMUPT/PfTQQxrfAuLj47F3716MHDkSjzzyCM6ePQs3NzcAgJOTE5ycLPtrq6yshLu7O1xcXCy6HH2cnZ1tunxDFBQUoGvXrrYuw2aEELhx44Z6+6SW69atW6irqzP5c6N+r7Q1FRYWAgB8fHzMNs+Kigp4eHjc0TzkcrmZqjHcjRs34OLiAgcHCR6osXWSasnq98ikpqZq7V+6dKkAINauXatuW7BggWj4a9u1a5eIiIgQCoVCeHh4iE6dOon4+HghxP/2ojR81O8BGTJkiOjWrZs4duyYGDx4sHBzcxOzZs1S9w0ZMkS9nPp5ffXVVyI+Pl4EBgYKd3d3MWrUKPHHH39o1BQaGqr1W/bt89RXW2xsrAgNDdV4fXl5uZgzZ44IDg4WLi4uolOnTuKdd94RdXV1GtMBEDNmzBBbt24V3bp1Ey4uLqJr167ip59+0rquG8rPzxdPPfWUCAgIEHK5XPTs2VMkJiY2WhcNH/q+TSclJYl7771XuLm5CR8fHzF48GCxc+dOdf+2bdvEww8/LFq3bi1cXFxEu3btxOLFixt9S63/vZ0+fVpERUUJNzc3oVKpxLJlyxots6qqSixYsEB07NhRyOVyERQUJB577DGRmZmpnqa2tlasWLFCdO3aVcjlchEQECCmTZsmrl27pjGv0NBQMWLECLFjxw7Rp08fIZfLxYoVK5p8vxcuXBCPP/64CAwMFHK5XLRp00aMGzdOFBcXCyGE1nVYv91cunRJPPvss6JTp07C1dVV+Pr6iieeeELrOj558qSIjIwUrq6uok2bNmLJkiVi3bp1Wn8nP/74oxg0aJBwd3cXnp6e4uGHHxanTp1q8j0I8b89h7dvA/V27NghAIj/+7//E0IIUVpaKmbNmiVCQ0OFi4uLUCqVYtiwYeL48eNNzr/+71rb9lT/LV7bXksAYsGCBY3mc/78eTFhwgTh7e0t/P39xWuvvSbq6urEH3/8IR555BHh5eUlAgMDxbvvvttonvq2fSE09yysWLFCtGvXTjg4OIi0tLQm36Ouz6nb51n/Ppv6GwPQ6HPBlN+ptnV++3xPnDgh/va3vwkvLy/h4eEh7r//fnH48GGNedR/hu/bt088++yzQqlUCh8fnyaXaegemds/P43Z9oQQ4r///a+YPHmyCAgIUH/uff755xqvq1+3GzduFK+++qpQqVRCJpOJ69ev66zLXnGPjB2bOHEiXnnlFezatQtPP/201mlOnz6NkSNHomfPnli8eDHkcjkyMzNx8OBBAECXLl2wePFizJ8/H9OmTcPgwYMBAAMHDlTP4+rVq3jooYcwfvx4/POf/0RgYKDOut58803IZDK89NJLKCgowPvvv49hw4YhPT3dqG/mhtR2OyEEHnnkESQnJ2PKlCm45557sHPnTsydOxeXL1/GihUrNKY/cOAAtmzZgunTp8PLywsffvghxowZgz/++AN+fn5N1lVVVYWoqChkZmbiueeeQ1hYGDZt2oRJkyahuLgYs2bNQpcuXZCUlITnn38ewcHBeOGFFwAASqWyyfkuWrQICxcuxMCBA7F48WK4uLjg6NGj2Lt3L4YPHw7gr3FTnp6emDNnDjw9PbF3717Mnz8fpaWleOeddzTmd/36dfztb3/D448/jrFjx2Lz5s146aWX0KNHDzz00EMAgNraWowcORI///wzxo8fj1mzZqGsrAy7d+/GqVOn0L59ewDAM888g8TEREyePBlxcXHIzs7GRx99hLS0NBw8eFBj79j58+cRExODZ555Bk8//TQ6d+6s9f3evHkT0dHRqK6uxsyZMxEUFITLly9j+/btKC4uhkKhQFJSEqZOnYp+/fph2rRpAKCuKTU1FYcOHcL48eMRHByMS5cuYfXq1YiKisKZM2fg7u4OALh8+TKGDh0KmUyG+Ph4eHh44LPPPtP6rTYpKQmxsbGIjo7GsmXLUFlZidWrV2PQoEFIS0trcnB537590a5dO3zzzTeIjY3V6Pv666/RqlUrREdHAwD+9a9/YfPmzXjuuefQtWtXXL16FQcOHMDZs2fRu3dvrfN//PHHceHCBWzcuBErVqyAv78/gL+2p/q9BsYYN24cunTpgrfeegs//PAD3njjDfj6+mLNmjW4//77sWzZMnz55Zd48cUXce+99yIyMhKAYdv+7davX48bN25g2rRpkMvl8PX11VqPvs8pber/xm5XXFyMOXPmICAgQN1m6u/08ccfh4+PD55//nnExMTg4Ycfhqenp7rewYMHw9vbG/PmzYOzszPWrFmDqKgo7N+/H/3799eY1/Tp06FUKjF//nxUVFQ0+Z5MYcy2l5+fj/vuuw8ymQzPPfcclEolfvrpJ0yZMgWlpaWNTkZYsmQJXFxc8OKLL6K6utrme+FNZusk1ZLp2yMjhBAKhUKEh4ernzfcI7NixQoBQBQWFjY5D13jUIYMGSIAiE8++URrn7Y9Mm3atBGlpaXq9m+++UYAEB988IG6zZA9Mvpqa7hHZtu2bQKAeOONNzSme+KJJ4RMJtPYwwBAuLi4aLSdPHlSABArV65stKzbvf/++wKA+OKLL9RtN2/eFAMGDBCenp4a771+D4U+GRkZwsHBQTz22GOitrZWo+/2vUmVlZWNXvvMM88Id3d3jfFC9b+3DRs2qNuqq6tFUFCQGDNmjLqtfq/E8uXLG823frm//PKLACC+/PJLjf76b3u3t4eGhgoAYseOHXrfc1pamkFjqpoaI6NtXRw+fLjR+545c6aQyWQaewOuXr0qfH19NfbIlJWVCR8fH/H0009rzDMvL08oFIpG7Q3Fx8cLZ2dnjb1U1dXVwsfHRzz11FPqNoVCIWbMmKFzXto0NUbGlD0y06ZNU7fdunVLBAcHC5lMJt566y11+/Xr14Wbm5vGujd026+vydvbWxQUFOh9b4Z8Tul6n0L8tb2OHDlSeHp6itOnTwsh7vx32tQektGjRwsXFxdx8eJFdduVK1eEl5eXiIyMVLfVf4YPGjTIoLE9puyREcLwbW/KlCmidevWoqioSGN+48ePFwqFQv03Vf9Z3q5dO61/Z1IjwYNhLYunp6fOs5fqj+1+9913qKurM2kZcrkckydPNnj6J598El5eXurnTzzxBFq3bo0ff/zRpOUb6scff4SjoyPi4uI02l944QUIIfDTTz9ptA8bNkz97R4AevbsCW9vb2RlZeldTlBQEGJiYtRtzs7OiIuLQ3l5Ofbv32907du2bUNdXR3mz5/f6Bj07afT375Hq6ysDEVFRRg8eDAqKytx7tw5jdd5enrin//8p/q5i4sL+vXrp/H+vv32W/j7+2PmzJmNaqpf7qZNm6BQKPDggw+iqKhI/ejTpw88PT2RnJys8bqwsDD1N0BdFAoFAGDnzp2orKzUO31Dt6+LmpoaXL16FR06dICPjw9OnDih7tuxYwcGDBiAe+65R93m6+uLCRMmaMxv9+7dKC4uRkxMjMb7dHR0RP/+/Ru9z4bGjRuHmpoabNmyRd22a9cuFBcXY9y4ceo2Hx8fHD16FFeuXDH6PZvL1KlT1T87Ojqib9++EEJgypQp6nYfHx907txZY3sxdtsfM2aMzr2Qty8LuLPPqSVLlmD79u1ITExUj0u709+pNrW1tdi1axdGjx6Ndu3aqdtbt26Nf/zjHzhw4ABKS0s1XvP000/D0dHRpPdlCEO2PSEEvv32W4waNQpCCI31ER0djZKSEo2/G+CvM7aaw/g2Bhk7V15erhEaGho3bhwiIiIwdepUBAYGYvz48fjmm2+M+rBo06aNUbsUO3bsqPFcJpOhQ4cOuHTpksHzMEVOTg5UKlWj9dGlSxd1/+3uuuuuRvNo1aoVrl+/rnc5HTt2bBQ4mlqOIS5evAgHBwe9A4NPnz6Nxx57DAqFAt7e3lAqleqwUlJSojFtcHBwo2sKNXx/Fy9eROfOnXUOEM/IyEBJSQkCAgKgVCo1HuXl5SgoKNCYPiwszKD3HBYWhjlz5uCzzz6Dv78/oqOjsWrVqkbvoylVVVWYP38+QkJCIJfL4e/vD6VSieLiYo155OTkoEOHDo1e37AtIyMDAHD//fc3ep+7du1q9D4b6tWrF+6++258/fXX6ravv/4a/v7+uP/++9Vtb7/9Nk6dOoWQkBD069cPCxcu1Bueza3htq9QKODq6qo+ZHV7++3bi7HbvqHbwp1+Tu3YsQOLFi1CfHw8xowZo26/09+pNoWFhaisrNR6yLRLly6oq6tDbm6uRruh68FUhmx7hYWFKC4uxtq1axuti/ovqqb+Lds7jpGxY//9739RUlKi9UO6npubG1JSUpCcnIwffvgBO3bswNdff437778fu3btMuhbgiUSeVMX7autrbXoN5fbNbUcIYRVlm+s4uJiDBkyBN7e3li8eDHat28PV1dXnDhxAi+99FKjD31zvb+6ujoEBATgyy+/1Nrf8Bu3MdvLe++9h0mTJuG7777Drl27EBcXh4SEBBw5cgTBwcE6Xztz5kysX78es2fPxoABA9QXLBs/frxJ3+rrX5OUlISgoKBG/YacDThu3Di8+eabKCoqgpeXF77//nvExMRovHbs2LEYPHgwtm7dil27duGdd97BsmXLsGXLFvXYJWPo+ltqirZtwxJ/D4ZuC3fyOZWdnY0JEybgwQcfxBtvvKHRZ47fqTlYY6+Gvm2vfl3885//bDSWpl7Pnj01njeHvTEAg4xdqx/opm83voODAx544AE88MADWL58OZYuXYpXX30VycnJGDZsmNmvBFz/LaieEAKZmZkafyStWrXSemGvnJwcjd21xtQWGhqKPXv2oKysTGOvTP0hl9DQUIPnpW85v/32G+rq6jS+md7Jctq3b4+6ujqcOXNG4xDI7fbt24erV69iy5Yt6sGXwF8f5KZq3749jh49ipqamiZPZ2/fvj327NmDiIgIi3yw9ejRAz169MBrr72GQ4cOISIiAp988on6n1JT28DmzZsRGxuL9957T91248aNRttVaGgoMjMzG72+YVv9YcaAgAAMGzbMpPcybtw4LFq0CN9++y0CAwNRWlqK8ePHN5qudevWmD59OqZPn46CggL07t0bb775ps4g09R6aNWqFQA0et+m7BnUxxLbfj19n1PaVFVVqQflbty4sdGeInP8ThtSKpVwd3fH+fPnG/WdO3cODg4OCAkJMcuyjKFv21MqlfDy8kJtba3Z1oVU8NCSndq7dy+WLFmCsLCwRsf6b3ft2rVGbfX/KKurqwFAfU0Dc10xdMOGDRrjdjZv3ow///xT40O6ffv2OHLkCG7evKlu2759e6NdssbU9vDDD6O2thYfffSRRvuKFSsgk8lM+rbb1HLy8vI0duPeunULK1euhKenJ4YMGWL0PEePHg0HBwcsXry40d6E+m/E9d9Kb/+GfPPmTXz88cemvA0Af41hKCoqarTObl/O2LFjUVtbiyVLljSa5tatWyZvN6Wlpbh165ZGW48ePeDg4KDeNoG/tgFty3B0dGy0t2DlypWN9kRER0fj8OHDGleHvnbtWqM9TNHR0fD29sbSpUtRU1PTaHmGnB3UpUsX9OjRA19//TW+/vprtG7dWiN01tbWNjp0FhAQAJVKpfGetWnqb8Hb2xv+/v5ISUnRaL+T7aIpltj2AcM+p7T517/+hQsXLmDr1q3qQHc7c/xOG3J0dMTw4cPx3XffaRwuz8/Px3/+8x8MGjQI3t7eRs/3Tunb9hwdHTFmzBh8++23OHXqVKPXm7IupIJ7ZOzATz/9hHPnzuHWrVvIz8/H3r17sXv3boSGhuL777/XeZGoxYsXIyUlBSNGjEBoaCgKCgrw8ccfIzg4GIMGDQLwV6jw8fHBJ598Ai8vL3h4eKB///4mHx/19fXFoEGDMHnyZOTn5+P9999Hhw4dNE4Rnzp1KjZv3oy//e1vGDt2LC5evIgvvvhCY/CtsbWNGjUKQ4cOxauvvopLly6hV69e2LVrF7777jvMnj270bxNNW3aNKxZswaTJk3C8ePH0bZtW2zevBkHDx7E+++/r3PMUlM6dOiAV199FUuWLMHgwYPx+OOPQy6XIzU1FSqVCgkJCRg4cCBatWqF2NhYxMXFQSaTISkp6Y52/T/55JPYsGED5syZg19//RWDBw9GRUUF9uzZg+nTp+PRRx/FkCFD8MwzzyAhIQHp6ekYPnw4nJ2dkZGRgU2bNuGDDz7AE088YfSy9+7di+eeew5///vf0alTJ9y6dQtJSUnqD9x6ffr0wZ49e7B8+XKoVCqEhYWhf//+GDlyJJKSkqBQKNC1a1ccPnwYe/bsaXTq/Lx58/DFF1/gwQcfxMyZM9WnX9911124du2aek+Ht7c3Vq9ejYkTJ6J3794YP348lEol/vjjD/zwww+IiIjQGvgaGjduHObPnw9XV1dMmTJFYy9BWVkZgoOD8cQTT6BXr17w9PTEnj17kJqaqrFnSZs+ffoAAF599VWMHz8ezs7OGDVqFDw8PDB16lS89dZbmDp1Kvr27YuUlBRcuHDB4N+FoSyx7QOGfU419MMPP2DDhg0YM2YMfvvtN/z222/qPk9PT4wePdpsv9OG3njjDezevRuDBg3C9OnT4eTkhDVr1qC6uhpvv/22Sevgdj///DNu3LjRqH306NHo3r17k6/Tte0BwFtvvYXk5GT0798fTz/9NLp27Ypr167hxIkT2LNnj9ZA2SzY5FwpEkL879S9+oeLi4sICgoSDz74oPjggw80TvOt1/D0659//lk8+uijQqVSCRcXF6FSqURMTIy4cOGCxuu+++470bVrV+Hk5KT1gnjaNHX69caNG0V8fLwICAgQbm5uYsSIESInJ6fR69977z3Rpk0bIZfLRUREhDh27FijeeqqTdsF8crKysTzzz8vVCqVcHZ2Fh07dtR5QbyGmjotvKH8/HwxefJk4e/vL1xcXESPHj20nhZq6OnX9datWyfCw8OFXC4XrVq1EkOGDBG7d+9W9x88eFDcd9996gvczZs3T+zcuVMAEMnJyerpmvq9aVtnlZWV4tVXXxVhYWHC2dlZBAUFiSeeeELj1FIhhFi7dq3o06ePcHNzE15eXqJHjx5i3rx54sqVKya936ysLPHUU0+J9u3bqy9oN3ToULFnzx6N6c6dOyciIyOFm5ubxgXxrl+/rv4deHp6iujoaHHu3Dmtv8O0tDQxePBgIZfLRXBwsEhISBAffvihACDy8vI0pk1OThbR0dFCoVAIV1dX0b59ezFp0iRx7Ngxg95XRkaG+m/2wIEDGn3V1dVi7ty5olevXuoLqfXq1Ut8/PHHBs17yZIlok2bNsLBwUHjVOzKykoxZcoUoVAohJeXlxg7dqwoKCho8vTrhqc5x8bGCg8Pj0bL07YdGbLtG3oacT1DPqcann7d8PPx9kfDbdzU36mu93HixAkRHR0tPD09hbu7uxg6dKg4dOiQxjSGXEJD2/KaeiQlJQkhmv6c0rXt1cvPzxczZswQISEh6r/3Bx54QOPCqpa43YwtyYSw05GPRER3YPbs2VizZg3Ky8utNsCciKyPY2SISPKqqqo0nl+9ehVJSUkYNGgQQwxRM8cxMkQkeQMGDEBUVBS6dOmC/Px8fP755ygtLcXrr79u69KIyMIYZIhI8h5++GFs3rwZa9euhUwmQ+/evfH5559rnNVBRM0Tx8gQERGRZHGMDBEREUkWgwwRERFJVrMfI1NXV4crV67Ay8vL7JfqJyIiIssQQqCsrAwqlarRxf9u1+yDzJUrV2xyXwwiIiK6c7m5uTpvMtvsg0z9JbVzc3Ntcn8MIiIiMl5paSlCQkL03hqj2QeZ2++zwiBDREQkLfqGhXCwLxEREUkWgwwRERFJFoMMERERSRaDDBEREUkWgwwRERFJFoMMERERSRaDDBEREUkWgwwRERFJFoMMERERSZZNg0xKSgpGjRoFlUoFmUyGbdu2NTntv/71L8hkMrz//vtWrJCIiIjsmU2DTEVFBXr16oVVq1bpnG7r1q04cuQIVCqVlSojIiIifbIKy5F8vgDZRRU2q8Gm91p66KGH8NBDD+mc5vLly5g5cyZ27tyJESNGWKkyIiIiakpx5U3EbUxHSkahui2yoxIrY8KhcHe2ai12PUamrq4OEydOxNy5c9GtWzeDXlNdXY3S0lKNBxEREZlP3MZ0HMws0mg7mFmEmRvTrF6LXQeZZcuWwcnJCXFxcQa/JiEhAQqFQv0ICQmxYIVEREQtS1ZhOVIyClErhEZ7rRBIySi0+mEmuw0yx48fxwcffIDExES9t/C+XXx8PEpKStSP3NxcC1ZJRETUsuRcq9TZf+kqgwwA4JdffkFBQQHuuusuODk5wcnJCTk5OXjhhRfQtm3bJl8nl8vh7e2t8SAiIiLzCPV119nf1s/DSpX8xaaDfXWZOHEihg0bptEWHR2NiRMnYvLkyTaqioiIqGVrp/REZEclDmYWaRxecpTJENHBH2H+LSjIlJeXIzMzU/08Ozsb6enp8PX1xV133QU/Pz+N6Z2dnREUFITOnTtbu1QiIiL6/1bGhGPmxjSNs5YiOvhjZUy41WuxaZA5duwYhg4dqn4+Z84cAEBsbCwSExNtVBURERHponB3xoYp/ZBdVIFLVyvQ1s/D6nti6smEaDDsuJkpLS2FQqFASUkJx8sQERFJhKH/v+12sC8RERGRPgwyREREJFkMMkRERCRZDDJEREQkWQwyREREJFkMMkRERCRZDDJEREQkWQwyREREJFkMMkRERCRZDDJEREQkWQwyREREJFkMMkRERCRZDDJEREQkWQwyREREJFkMMkRERCRZDDJEREQkWQwyREREJFkMMkRERCRZDDJEREQkWQwyREREJFkMMkRERCRZDDJEREQkWQwyREREJFkMMkRERCRZDDJEREQkWQwyREREJFkMMkRERCRZDDJEREQkWQwyREREJFkMMkRERCRZDDJEREQkWQwyREREJFkMMkRERCRZDDJEREQkWQwyREREJFkMMkRERCRZDDJEREQkWQwyREREJFkMMkRERCRZDDJEREQkWQwyREREJFkMMkRERCRZNg0yKSkpGDVqFFQqFWQyGbZt26buq6mpwUsvvYQePXrAw8MDKpUKTz75JK5cuWLDiomIiMie2DTIVFRUoFevXli1alWjvsrKSpw4cQKvv/46Tpw4gS1btuD8+fN45JFHbFApERER2SOZEELYuggAkMlk2Lp1K0aPHt3kNKmpqejXrx9ycnJw1113GTTf0tJSKBQKlJSUwNvb21zlEhERkQUZ+v/byYo13bGSkhLIZDL4+Pg0OU11dTWqq6vVz0tLS61RGhEREdmAZAb73rhxAy+99BJiYmJ0JrOEhAQoFAr1IyQkxIpVEhERkTVJIsjU1NRg7NixEEJg9erVOqeNj49HSUmJ+pGbm2ulKomIiMja7P7QUn2IycnJwd69e/WOc5HL5ZDL5VaqjoiIiGzJroNMfYjJyMhAcnIy/Pz8bF0SERER2RGbBpny8nJkZmaqn2dnZyM9PR2+vr5o3bo1nnjiCZw4cQLbt29HbW0t8vLyAAC+vr5wcXGxVdlERERkJ2x6+vW+ffswdOjQRu2xsbFYuHAhwsLCtL4uOTkZUVFRBi2Dp18TERFJjyROv46KioKuHGUnl7ghIiIiOyWJs5aIiIiItGGQISIiIslikCEiIiLJYpAhIiIiyWKQISIiIslikCEiIiLJYpAhIiIiyWKQISIiIslikCEiIiLJYpAhIiIiyWKQISIiIslikCEiIiLJYpAhIiIiyWKQISIiIslikCEiIiLJYpAhIiIiyWKQISIiIslysnUBREREJE1ZheXIuVaJtn4eCPP3sEkNDDJERERklOLKm4jbmI6UjEJ1W2RHJVbGhEPh7mzVWnhoiYiIiIwStzEdBzOLNNoOZhZh5sY0q9fCIENEREQGyyosR0pGIWqF0GivFQIpGYXILqqwaj0MMkRERGSwnGuVOvsvXWWQISIiIjsV6uuus7+tn3UH/TLIEBERkcHaKT0R2VEJR5lMo91RJkNkR6XVz15ikCEiIiKjrIwJR0QHf422iA7+WBkTbvVaePo1ERERGUXh7owNU/ohu6gCl65W8DoyRFfxo+4AACAASURBVEREJD1h/rYLMPV4aImIiIgki0GGiIiIJItBhoiIiCSLQYaIiIgki0GGiIiIJItBhoiIiCSLQYaIiIgki0GGiIiIJItBhoiIiCSLQYaIiIgki7coICIiIpNkFZYj51ol77VERERE0lFceRNxG9ORklGobovsqMTKmHAo3J2tWgsPLREREZFR4jam42BmkUbbwcwizNyYZvVaGGSIiIjIYFmF5UjJKEStEBrttUIgJaMQ2UUVVq2HQYaIiIgMlnOtUmf/pastKMikpKRg1KhRUKlUkMlk2LZtm0a/EALz589H69at4ebmhmHDhiEjI8NG1RIREVGor7vO/rZ+1h30a9MgU1FRgV69emHVqlVa+99++218+OGH+OSTT3D06FF4eHggOjoaN27csHKlREREBADtlJ6I7KiEo0ym0e4okyGyo9LqZy/JhGhwkMtGZDIZtm7ditGjRwP4a2+MSqXCCy+8gBdffBEAUFJSgsDAQCQmJmL8+PEGzbe0tBQKhQIlJSXw9va2WP1EREQtRUllDWZuTLPoWUuG/v+229Ovs7OzkZeXh2HDhqnbFAoF+vfvj8OHDzcZZKqrq1FdXa1+XlpaavFaiYiIWhKFuzM2TOmH7KIKXLpaYdPryNjtYN+8vDwAQGBgoEZ7YGCguk+bhIQEKBQK9SMkJMSidRIREbVUYf4eGNo5wGYhBrDjIGOq+Ph4lJSUqB+5ubm2LomIiIgsxG6DTFBQEAAgPz9foz0/P1/dp41cLoe3t7fGg4iIiJonuw0yYWFhCAoKws8//6xuKy0txdGjRzFgwAAbVkZERET2wqaDfcvLy5GZmal+np2djfT0dPj6+uKuu+7C7Nmz8cYbb6Bjx44ICwvD66+/DpVKpT6ziYiIiFo2mwaZY8eOYejQoernc+bMAQDExsYiMTER8+bNQ0VFBaZNm4bi4mIMGjQIO3bsgKurq61KJiIiIjtiN9eRsRReR4aIiEh6DP3/bbdjZIiIiIj0YZAhIiIiyTJ6jMzZs2fx1Vdf4ZdffkFOTg4qKyuhVCoRHh6O6OhojBkzBnK53BK1EhEREWkweIzMiRMnMG/ePBw4cAARERHo168fVCoV3NzccO3aNZw6dQq//PILSktLMW/ePMyePdsuAg3HyBAREUmP2e+1NGbMGMydOxebN2+Gj49Pk9MdPnwYH3zwAd577z288sorxlVNREREZASD98jU1NTA2dnwO1oaO72lcI8MERGR9Jj9rCVjQ4k9hBgiIiJq3ky+IF5qaiqSk5NRUFCAuro6jb7ly5ffcWFERERE+pgUZJYuXYrXXnsNnTt3RmBgIGQymbrv9p+JiIiILMmkIPPBBx9g3bp1mDRpkpnLISIiIjKcSRfEc3BwQEREhLlrISIiIjKKSUHm+eefx6pVq8xdCxEREZFRTDq09OKLL2LEiBFo3749unbt2ugMpS1btpilOCIiIiJdTAoycXFxSE5OxtChQ+Hn58cBvkRERGQTJgWZf//73/j2228xYsQIc9dDREREZDCTxsj4+vqiffv25q6FiIiIyCgmBZmFCxdiwYIFqKysNHc9RERERAYz6dDShx9+iIsXLyIwMBBt27ZtNNj3xIkTZimOiIiISBeTgszo0aPNXQcRERGR0Qy++7VU8e7XRERE0mP2u1/fLjU1FUePHm3UfvToURw7dsyUWRIREREZzaQgM2PGDOTm5jZqv3z5MmbMmHHHRREREREZwqQgc+bMGfTu3btRe3h4OM6cOXPHRREREREZwqQgI5fLkZ+f36j9zz//hJOTSeOHiYiIiIxmUpAZPnw44uPjUVJSom4rLi7GK6+8ggcffNBsxRERERHpYtLuk3fffReRkZEIDQ1FeHg4ACA9PR2BgYFISkoya4FERERETTEpyLRp0wa//fYbvvzyS5w8eRJubm6YPHkyYmJiGl0cj4iIiMhSTB7Q4uHhgWnTppmzFiIiIiKjGDxG5siRIwbPtLKyEqdPnzapICIiIiJDGRxkJk6ciOjoaGzatAkVFRVapzlz5gxeeeUVtG/fHsePHzdbkURERETaGHxo6cyZM1i9ejVee+01/OMf/0CnTp2gUqng6uqK69ev49y5cygvL8djjz2GXbt2oUePHpasm4iIiMi0ey0dO3YMBw4cQE5ODqqqquDv74/w8HAMHToUvr6+lqjTZLzXEhERkfQY+v/bpMG+ffv2Rd++fU0ujoiIiMgcTLogHhEREZE94P0EiIiIyCRZheXIuVaJtn4eCPP3sEkNDDJERERklOLKm4jbmI6UjEJ1W2RHJVbGhEPhbt0L4/LQEhERERklbmM6DmYWabQdzCzCzI1pVq/ljoPMjRs3zFEHERERSUBWYTlSMgpR2+Ck51ohkJJRiOwi7deasxSTgkxdXR2WLFmCNm3awNPTE1lZWQCA119/HZ9//rlZCyQiIiL7kXOtUmf/pasSCDJvvPEGEhMT8fbbb8PFxUXd3r17d3z22WdmK46IiIjsS6ivu87+tn7WHfRrUpDZsGED1q5diwkTJsDR0VHd3qtXL5w7d85sxREREZF9aaf0RGRHJRxlMo12R5kMkR2VVj97yaQgc/nyZXTo0KFRe11dHWpqau64KCIiIrJfK2PCEdHBX6MtooM/VsaEW70Wk06/7tq1K3755ReEhoZqtG/evBnh4eZ7E7W1tVi4cCG++OIL5OXlQaVSYdKkSXjttdcga5AEiYiIyDoU7s7YMKUfsosqcOlqhfSuIzN//nzExsbi8uXLqKurw5YtW3D+/Hls2LAB27dvN1txy5Ytw+rVq/Hvf/8b3bp1w7FjxzB58mQoFArExcWZbTlERERkvDB/2wWYeibdNBIAfvnlFyxevBgnT55EeXk5evfujfnz52P48OFmK27kyJEIDAzUOBNqzJgxcHNzwxdffGHQPHjTSCIiIumx6E0jAWDw4MHYvXu3qS83yMCBA7F27VpcuHABnTp1wsmTJ3HgwAEsX768yddUV1ejurpa/by0tNSiNRIREZHtmDTY96mnnsK///3vRu2lpaV46qmn7rioei+//DLGjx+Pu+++G87OzggPD8fs2bMxYcKEJl+TkJAAhUKhfoSEhJitHiIiIrIvJh1acnBwgJubG6ZMmYL3338fDg5/5aH8/HyoVCrU1taapbivvvoKc+fOxTvvvINu3bohPT0ds2fPxvLlyxEbG6v1Ndr2yISEhPDQEhERkYRY/NDSDz/8gKlTp+Ls2bP45ptv0KpVK1Nn1aS5c+eq98oAQI8ePZCTk4OEhIQmg4xcLodcLjd7LURERGR/TL7XUteuXXH06FHU1NSgX79+OHv2rDnrAgBUVlaq9/bUc3R0RF1dndmXRURERNJjUpCpv4aLn58f9uzZgyFDhmDAgAH4/vvvzVrcqFGj8Oabb+KHH37ApUuXsHXrVixfvhyPPfaYWZdDRERE0mTSoaXbh9U4OTnhs88+Q9euXTF9+nSzFQYAK1euxOuvv47p06ejoKAAKpUKzzzzDObPn2/W5RAREZE0mTTYd//+/YiIiICTk2YO2rNnDw4ePIgFCxaYrcA7xevIEBERSY+h/79NviCeVDDIEBERSY/Zz1qaM2cOlixZAg8PD8yZM0fntLouWEdERERkLgYHmbS0NPWdrdPS0pqcjjdzJCIiImvhoSUiIiKyO4b+/zb5OjK3y8nJwZkzZ3h9FyIiIrIqo4LMunXrGo1/mTZtGtq1a4cePXqge/fuyM3NNWuBRERERE0xKsisXbtW41YEO3bswPr167FhwwakpqbCx8cHixYtMnuRRERERNoYdUG8jIwM9O3bV/38u+++w6OPPqq+G/XSpUsxefJk81ZIREREdimrsBw51yrR1s8DYf4eNqnBqCBTVVWlMeDm0KFDmDJlivp5u3btkJeXZ77qiIiIyO4UV95E3MZ0pGQUqtsiOyqxMiYcCndnq9Zi1KGl0NBQHD9+HABQVFSE06dPIyIiQt2fl5cHhUJh3gqJiIjIrsRtTMfBzCKNtoOZRZi5senLs1iKUXtkYmNjMWPGDJw+fRp79+7F3XffjT59+qj7Dx06hO7du5u9SCIiIrIPWYXlGnti6tUKgZSMQmQXVVj1MJNRQWbevHmorKzEli1bEBQUhE2bNmn0Hzx4EDExMWYtkIiIiOxHzrVKnf2Xrlo3yPCCeERERGSwrMJy3P/e/ib7k1+MMkuQseoF8YiIiKhlaKf0xMD2flr7Brb3s/rZSwwyREREZJSmjuXY4hgPgwwREREZLKuwHIezrmrtO5x1FdlFFVath0GGiIiIDGbIYF9rMjrI1NTUoH379jh79qwl6iEiIiI7FurrrrO/rZ+dj5FxdnbGjRs3LFELERER2bl2Sk9EdlTCUSbTaHeUyRDZUSmNwb4zZszAsmXLcOvWLXPXQ0RERHZuZUw4Ijr4a7RFdPDHyphwq9di0nVkHnvsMfz888/w9PREjx494OGhmb62bNlitgLvFK8jQ0REZBnZRRW4dLXCIjeNNPT/t1FX9q3n4+ODMWPGmFwcERERSV+Yv+3uel3PpCCzfv16c9dBREREZDSTgky9wsJCnD9/HgDQuXNnKJVKsxRFREREZAiTBvtWVFTgqaeeQuvWrREZGYnIyEioVCpMmTIFlZW6zy8nIiIiMheTgsycOXOwf/9+/N///R+Ki4tRXFyM7777Dvv378cLL7xg7hqJiIiItDLprCV/f39s3rwZUVFRGu3JyckYO3YsCgsLzVXfHeNZS0RERNJj0bOWKisrERgY2Kg9ICCAh5aIiIhaiKzCcuRcq7TI6deGMinIDBgwAAsWLMCGDRvg6uoKAKiqqsKiRYswYMAAsxZIRERE9qW48ibiNqYjJeN/R2AiOyqxMiYcCndnq9Zi0qGlU6dOITo6GtXV1ejVqxcA4OTJk3B1dcXOnTvRrVs3sxdqKh5aIiIiMq+YtUe03gF7QDs/bJx2n1mWYdFDS927d0dGRga+/PJLnDt3DgAQExODCRMmwM3NzbSKiYiIyO5lFZZrDTEAcDjrKrKLKqx6mMnk68i4u7vj6aefNmctREREZOeOZl/T2X8k66o0gsyVK1dw4MABFBQUoK6uTqMvLi7ujgsjIiIie6R7RIpMZ6/5mRRkEhMT8cwzz8DFxQV+fn6Q3XYrb5lMxiBDRETUTKkUuoeQtGll3SEmJgWZ119/HfPnz0d8fDwcHEy6ph4RERFJUJ2e/lt1Rp9DdEdMSiGVlZUYP348QwwREVELE+rrrrO/rZ91rydjUhKZMmUKNm3aZO5aiIiIiIxi0qGlhIQEjBw5Ejt27ECPHj3g7Kx58Zvly5ebpTgiIiKyLznXdF/B/9JVCZx+nZCQgJ07d6Jz584A0GiwLxERETVP9nZoyaQg895772HdunWYNGmSmcshIiIie9ZO6YkB7fyavLKvte+5ZNIYGblcjoiICHPXQkRERBLQ1MEXWxyUMSnIzJo1CytXrjR3LURERGTnsgrLceii9lsUHLr41y0KrMmkQ0u//vor9u7di+3bt6Nbt26NBvtu2bLFLMUBwOXLl/HSSy/hp59+QmVlJTp06ID169ejb9++ZlsGERERGaZZDPb18fHB448/bu5aGrl+/ToiIiIwdOhQ/PTTT1AqlcjIyECrVq0svmwiouYkq7AcOdcq0dbPw+pjGKh5aRaDfdevX2/uOrRatmwZQkJCNJYXFhZmlWUTETUHxZU3EbcxHSkZheq2yI5KrIwJh8LdWccriaTBri/N+/3336Nv3774+9//joCAAISHh+PTTz/V+Zrq6mqUlpZqPIiIWqrpX57QCDEAkJJRiGe/PG6jikjqDDm0ZE0m7ZEJCwvTeb2YrKwskwtqOJ/Vq1djzpw5eOWVV5Camoq4uDi4uLggNjZW62sSEhKwaNEisyyfiEjKDBmUycNMZKxmcWhp9uzZGs9ramqQlpaGHTt2YO7cuWYpDADq6urQt29fLF26FAAQHh6OU6dO4ZNPPmkyyMTHx2POnDnq56WlpQgJCTFbTUREUnE0W3uIUfdnXWWQIaO1U3oisqMSBzOLUCv+d4NIR5kMER38rb5NmRRkZs2apbV91apVOHbs2B0VdLvWrVuja9euGm1dunTBt99+2+Rr5HI55HK52WogIpIu3Rf1sO49iqk5WRkTjpkb0zQOW0Z08MfKmHCr12JSkGnKQw89hPj4eLMNBo6IiMD58+c12i5cuIDQ0FCzzJ+IqDnrH+ars/++dn5WqoSaG4W7MzZM6Yfsogpculph07PhzBpkNm/eDF9f3X84xnj++ecxcOBALF26FGPHjsWvv/6KtWvXYu3atWZbBhFRc5WrZ1Dmf69X8tAS3ZEwf9ufzm9UkFm8eDFeeOEFDBo0SGOwrxACeXl5KCwsxMcff2y24u69915s3boV8fHxWLx4McLCwvD+++9jwoQJZlsGEVFzlf7fYp39J/64jsEdlVaqhsgyZEIIgw+TOjo64s8//8THH3+sEWQcHBygVCoRFRWFu+++2yKFmqq0tBQKhQIlJSXw9va2dTlERFaz/3wBYtenNtmfNKUfgwzZLUP/fxu1R6Y+8yxcuPCOiiMiIssb0jkATg4y3Kpr/H3VyUHGEEPNgtEXxNN1/RgiIrIfWYXlWkMMANyqE1a/uR+RJRg92LdTp056w8y1a9dMLoiIiMzD3m7uR2QJRgeZRYsWQaFQWKIWIiIyI3u7AiuRJRgdZMaPH4+AgABL1EJERERkFKPGyHB8DBGRdGz/7YrO/h/09BNJgVFBxogztYmIyMb0DebNKiy3UiVElmPUoaW6ujpL1UFERGambyBvO6WnlSohshyjT78mIiJpGNlTpbN/hJ5+IilgkCEiaqbaKT3RK1j7Waa9ghU89ZqaBQYZIqJm7NJV7eNkmmonkhoGGSKiZmr/+QKUVN3S2ldSdQu/ZBRauSIi82OQISJqpgy5+zWR1DHIEBE1U0Ferjr7VQo3K1VCZDkMMkREzZS+C2Y0dUNJIilhkCEiaqYKy27o7L9aXm2lSogsh0GGiKiZuqTnyr76+omkgEGGiKiZKr2h/YylesVVNVaqhMhyGGSIiJqpe4J9dPb3vquVlSohshwGGSKiZqpbE1f1rddF5W2lSogsh0GGiKiZ0vcB7+Qgs0odRJbEIENE1Ezx9GtqCRhkiIiaKe6RoZaAQYaIqJnad173vZT2nSuwUiVElsMgQ0TUTKXpuZcS77VEzQGDDBFRM+Xu4qin38lKlRBZDoMMEVEz1TnIS2f/3Xr6iaSAQYaIqJkqqeKVfan5Y5AhImqmsorKdfcX8l5LJH0MMkREzZS/p1xnv1JPP5EUMMgQETVTE+8L1d0/UHc/kRQwyBARNVNDOgfo7B/cUWmlSogsh0GGiKiZyirUPUYmu4hjZEj6GGSIiJqpj/Zm6uxftTfDSpUQWQ6DDBFRM3XmzxKd/aev6O4nkgIGGSKiZiqklbvO/rt8PaxUCZHlMMgQETVTwa3c7qifSAoYZIiImqliPVf2vV7JK/uS9DHIEBE1Uz5uzjr7W7nr7ieSAgYZIqIWSti6ACIzYJAhImqmhJ6oIphkqBlgkCEiaqb8PfTca8mL91oi6ZNUkHnrrbcgk8kwe/ZsW5dCRGT3avXskampq7NSJUSWI5kgk5qaijVr1qBnz562LoXIorIKy5F8voCXj6c7tv9sgc7+X84XWqkSIstxsnUBhigvL8eECRPw6aef4o033rB1OUQWUVx5E3Eb05GS8b9/LpEdlVgZEw4Fzy4hE2ToCcMXCnTfi4lICiSxR2bGjBkYMWIEhg0bZutSiCzmqfW/aoQYAEjJKMTk9b/aqCKSOoWr7u+qCj2nZxNJgd3vkfnqq69w4sQJpKamGjR9dXU1qqur1c9LS0stVRqR2WQVluNErvb73pzILUZ2UQXC/Hk5eTLO2HtDsHx30zeGHHdviBWrIbIMu94jk5ubi1mzZuHLL7+Eq6urQa9JSEiAQqFQP0JC+IdK9i/p8CWd/V/o6SfSZmRPlc7+EXr6iaTAroPM8ePHUVBQgN69e8PJyQlOTk7Yv38/PvzwQzg5OaG2trbRa+Lj41FSUqJ+5Obm2qByIuOk/aH7LsQn/ii2UiXUnLRTeqJfW1+tff3a+nIvHzULdn1o6YEHHsDvv/+u0TZ58mTcfffdeOmll+Do6NjoNXK5HHI5r41A0iKT6T5NVmalOqj5+fTJvpi5MU3rIHKi5sCug4yXlxe6d++u0ebh4QE/P79G7URS5urcOJTfzs3Frneekh1TuDtjw5R+yC6qwKWrFWjr58E9MdSs2HWQIWopPOW6/xQ99PQT6RPmzwBDzZPkPh337dtn6xKIzO7sn7rPrjv3Z5mVKiEikhburyayA4Xl1Tr7C8puWKkSIiJpYZAhsgP6/hA52JeISDsGGSK7oC+qMMoQEWnDIENkB+R6zkpyddF9VhMRUUvFIENkB1p76b5ytcrbsCtbExG1NAwyRHag7OYtnf2l1br7iYhaKgYZIjtw42bj223crkpP0CEiaqkYZIjsQJXeIKO7n4iopWKQIbIDt+p032tJXz8RUUvFIENkB9z0nJXkzrOWiIi0YpAhsgNODrqvE+Oop5+IqKVikCGyA/oOHdXy0BIRkVYMMkR2oLWe68S0VrhZqRIiImlhkCGyA74ect39ni5WqoSISFoYZIjsQGZhme7+gnIrVUJEJC0MMkR24JbQd/p1nZUqISKSFgYZIjsQ4Kn70FKAJ++1RESkDYMMkR1wdXLS3e/M68gQEWnDIENkB6pqdN9LSV8/EVFLxSBDZAfK9dwUsrya91oiItKGQYbIDlwvv6mz/1qF7n4iopaKQYbIDlTX6j4rqfoW98gQEWnDIENkD3gHAiIikzDIENkBZyfdN4V0ceRNI4mItGGQIbIDcifdp1fr6yciaqkYZIjsgKdc93VkPF2drVQJEZG0MMgQ2QEXJ91/ii6O/FMlItKGn45EduDmLd1nLd3Uc1YTEVFLxSBDZAe4R4aIyDT8dCSyA5V6ruxboaefiKilYpAhsgPX9Vy593pFtZUqISKSFgYZIjugZ4iM3n4iopaKQYbIDui73h2vh0dEpB2DDJEdkMl0JxV9/URELRWDDJEdqKvTfbOlOsGbMRERacMgQ2QHBHe4EBGZhEGGyA64Oeu+l5K+fiKilopBhsgOODro3iWjr5+IqKVikCGyA/ouiFd5s9ZKlRARSQuDDJEdcNK7R8ZKhRARSQw/HonsgJers85+bzcXK1VCRCQtDDJEdsDXXa6nn0GGiEgbuw4yCQkJuPfee+Hl5YWAgACMHj0a58+ft3VZRGZXfUv3GJibvEcBEZFWdh1k9u/fjxkzZuDIkSPYvXs3ampqMHz4cFRUVNi6NHz0cwZi1h7Gx8mZti6FmoGbdbqDSnUtB/sSEWnjZOsCdNmxY4fG88TERAQEBOD48eOIjIy0SU2HMgvxj89+VT8/nHUNb+88j6+fvg/92/vZpCaSvlt69rjU1vLKvkRE2tj1HpmGSkpKAAC+vr42q+H2EHO7cZ8esXIl1JyU3NB9+vX1qhorVUJEJC2SCTJ1dXWYPXs2IiIi0L179yanq66uRmlpqcbDXD76OUNnPw8zkam83XTvHFW46T6riYiopZJMkJkxYwZOnTqFr776Sud0CQkJUCgU6kdISIjZathzNl9n/64zeWZbFrUsE/qH6uz/5326+4mIWipJBJnnnnsO27dvR3JyMoKDg3VOGx8fj5KSEvUjNzfXbHV4uOj+1uwpt+shR2THRvZU6ewfoaefiKilsuv/vEIIzJw5E1u3bsW+ffsQFham9zVyuRxyue5rcpgqQOGqu99Ldz9RU9opPdGvbSv8eul6o75+bVshzN/DBlUREdk/u94jM2PGDHzxxRf4z3/+Ay8vL+Tl5SEvLw9VVVU2qSfM311nfzsl/9mQ6T598l5EdlRqtEV2VOLTJ++1UUVERPbPrvfIrF69GgAQFRWl0b5+/XpMmjTJ6vX0CvbR3R+iu59IF4W7MzZM6YfsogpculqBtn4e3BNDRKSHXQcZIezr2hkn/1uisz89txiDG3yjJjJWmD8DDBGRoez60JK9yS4q191faPsrDhMREbUkDDJGCPP31N3PMTJERERWxSBjhF7BCp3993CMDBERkVUxyBjh5H+Ldffn6u4nIiIi82KQMUJ2ke4xMFkcI0NERGRVDDJGkEGmu193NxEREZkZg4wR9N3Yz9uVN/YjIiKyJgYZIwztHKCz//4uuvuJiIjIvBhkjBDiq/sWBcGtdPcTERGReTHIGOFo9jWd/UeyrlqpEiIiIgIYZIxSWHZDZ//V8morVUJEREQAg4yRdJ+WZF93hiIiImr+GGSMoPRy0d3vKbdSJURERAQwyBilf5if7v52uvuJiIjIvBhkjNBO6YmB7bWHlYHt/RDmz5tGEhERWRODjJFWT+iDyI5KjbbIjkqsntDHRhURERG1XLovVUuNKNydsWFKP2QXVeDS1Qq09fPgnhgiIiIbYZAxUZg/AwwREZGt8dASERERSRaDDBEREUkWgwwRERFJFoMMERERSRaDDBEREUkWgwwRERFJFoMMERERSRaDDBEREUkWgwwRERFJFoMMERERSVazv0WBEAIAUFpaauNKiIiIyFD1/7fr/483pdkHmbKyMgBASEiIjSshIiIiY5WVlUGhUDTZLxP6oo7E1dXV4cqVK/Dy8oJMJjPbfEtLSxESEoLc3Fx4e3ubbb7NFdeX4biuDMd1ZTiuK8NxXRnOkutKCIGysjKoVCo4ODQ9EqbZ75FxcHBAcHCwxebv7e3NDd0IXF+G47oyHNeV4biuDMd1ZThLrStde2LqcbAvERERSRaDDBEREUmW48KFCxfaugipcnR0RFRUFJycmv0ROrPg+jIc15XhuK4MwiSfJwAADq1JREFUx3VlOK4rw9l6XTX7wb5ERETUfPHQEhEREUkWgwwRERFJFoMMERERSRaDDBEREUkWg0wTUlJSMGrUKKhUKshkMmzbtk3va/bt24fevXtDLpejQ4cOSExMtHyhdsDYdbVv3z7IZLJGj7y8PCtVbDsJCQm499574eXlhYCAAIwePRrnz5/X+7pNmzbh7rvvhqurK3r06IEff/zRCtXalinrKjExsdF25erqaqWKbWf16tXo2bOn+qJkAwYMwE8//aTzNS1xmwKMX1ctdZvS5q233oJMJsPs2bN1TmftbYtBpgkVFRXo1asXVq1aZdD02dnZGDFiBIYOHYr09HTMnj0bU6dOxc6dOy1cqe0Zu67qnT9/Hn/++af6ERAQYKEK7cf+/fsxY8YMHDlyBLt370ZNTQ2GDx+OioqKJl9z6NAhxMTEYMqUKUhLS8Po0aMxevRonDp1yoqVW58p6wr46wqjt29XOTk5VqrYdoKDg/HWW2/h+PHjOHbsGO6//348+uijOH36tNbpW+o2BRi/roCWuU01lJqaijVr1qBnz546p7PJtiVILwBi69atOqeZN2+e6Natm0bbuHHjRHR0tCVLszuGrKvk5GQBQFy/ft1KVdmvgoICAUDs37+/yWnGjh0rRowYodHWv39/8cwzz1i6PLtiyLpav369UCgUVqzKfrVq1Up89tlnWvu4TWnSta64TQlRVlYmOnbsKHbv3i2GDBkiZs2a1eS0tti2uEfGTA4fPoxhw4ZptEVHR+Pw4cM2qsj+3XPPPWjdujUefPBBHDx40Nbl2ERJSQkAwNfXt8lpuG39xZB1BQDl5eUIDQ1FSEiI3m/azVFtbS2++uorVFRUYMCAAVqn4Tb1F0PWFcBtasaMGRgxYkSjbUYbW2xbvGShmeTl5SEwMFCjLTAwEKWlpaiqqoKbm5uNKrM/rVu3xieffIK+ffuiuroan332GaKionD06FH07t3b1uVZTV1dHWbPno2IiAh07969yema2rZawpiieoauq86dO2PdunXo2bMnSkpK8O6772LgwIE4ffq0RW8eaw9+//13DBgwADdu3ICnpye2bt2Krl27ap22pW9TxqyrlrxNAcBXX32FEydOIDU11aDpbbFtMciQ1XXu3BmdO3dWPx84cCAuXryIFStWICkpyYaVWdeMGTNw6tQpHDhwwNal2D1D19WAAQM0vlkPHDgQXbp0wZo1a7BkyRJLl2lTnTt3Rnp6OkpKSrB582bExsZi//79Tf6DbsmMWVcteZvKzc3FrFmzsHv3brse4MwgYyZBQUHIz8/XaMvPz4e3tzf3xhigX79+Leof+nPPPYft27cjJSVF77e6pratoKAgS5ZoN4xZVw05OzsjPDwcmZmZFqrOfri4uKBDhw4AgD59+iA1NRUffPAB1qxZ02jalr5NGbOuGmpJ29Tx48dRUFCgsae8trYWKSkp+Oijj1BdXQ1HR0eN19hi2+IYGTMZMGAAfv75Z4223bt36zzuSv+Tnp6O1q1b27oMixNC4LnnnsPWrVuxd+9ehIWF6X1NS922TFlXDdXW1uL3339vEdtWQ3V1daiurtba11K3qaboWlcNtaRt6oEHHsDvv/+O9PR09aNv376YMGEC0tPTG4UYwEbblsWGEUtcWVmZSEtLE2lpaQKAWL58uUhLSxM5OTlCCCFefvllMXHiRPX0WVlZwt3dXcydO1ecPXtWrFq1Sjg6OoodO3bY6i1YjbHrasWKFWLbtm0iIyND/P7772LWrFnCwcFB7Nmzx1ZvwWqeffZZoVAoxL59+8Sff/6pflRWVqqnmThxonj55ZfVzw8ePCicnJzEu+++K86ePSsWLFggnJ2dxe+//26Lt2A1pqyrRYsWiZ07d4qLFy+K48ePi/HjxwtXV1dx+vRpW7wFq3n55ZfF/v37RXZ2tvjtt9/Eyy+/LGQymdi1a5cQgtvU7YxdVy11m2pKw7OW7GHbYpBpQv0pwg0fsbGxQgghYmNjxZAhQxq95p577hEuLi6iXbt2Yv369Vav2xaMXVfLli37f+3daUhUXx8H8K+Vt7SxMZd0pkXL1DDLVsuk1WIqspXqhZCiGOowEtHyosU2ixaLDNGwzcTKVggko7J9wywzywIN88UY86KiJjO1fs+L+N+nSa3secr/0Pfzau4959zzO5eL/jj3HK74+flJly5dxM3NTSZMmCBFRUXtE/wf1tJ9AmDzrIwfP169d/84fvy4BAQEiKIoMnDgQCkoKPizgbeDX7lXS5YskT59+oiiKOLl5SXTp0+X+/fv//ng/7DY2Fjx8fERRVHE09NTIiIi1H/MInymvtbWe/W3PlOt+TaR+Tc8Ww4iIr9vvoeIiIjo9+EaGSIiIrJbTGSIiIjIbjGRISIiIrvFRIaIiIjsFhMZIiIisltMZIiIiMhuMZEhIiIiu8VEhoiIiOwWExkiUr18+RImkwn9+vVD586d0bt3b0RGRjb7doo9yc7ORkhICDQaDVxdXTF06FBs2bJFLY+JicHs2bPbMUIi+l/w69dEBACorq5GeHg4XF1dsX37dgwaNAiNjY04f/48jEYjnj592t4htqqhoQGKojQ7f+DAASxZsgTp6ekYP348Pn78iLKyMpSXl7dDlET0W/zWDyAQkd2YNm2a9OzZU6xWa7Oy169fq7/T0tIkODhYnJ2dpVevXpKYmCjv3r1Tyw8ePCharVYKCwtlwIAB0rVrVzEYDGI2m22uuX//fgkKChJFUcTb21uMRqNNf3FxceLh4SEuLi4yceJEKS0tVctTUlIkJCREsrOzxdfXVxwcHFoc06xZsyQmJqbVMaekpDT7ltPly5dFRGTFihXi7+8vTk5O0rdvX1m9erU0NDTYtN+4caN4enqKRqORuLg4WblypYSEhNjUyc7OlgEDBkjnzp0lMDBQMjIyWo2HiNqOr5aICK9evUJhYSGMRiO6du3arNzV1VX93aFDB6Snp+Px48fIyclBUVERVqxYYVO/rq4OO3bsQG5uLq5du4aamhosW7ZMLc/MzITRaMTixYvx6NEjnD17Fv3791fL58+fD4vFgnPnzqGkpATDhg1DREQEXr16pdaprKzEqVOncPr0aZSWlrY4Lm9vb9y5cwcvXrxosXzZsmVYsGABpk6ditraWtTW1mLMmDEAABcXFxw6dAhPnjzB7t27kZ2djV27dqlt8/LykJqaiq1bt6KkpAR9+vRBZmamzfXz8vKwdu1apKamoqKiAps3b8aaNWuQk5PTYjxE9AvaO5MiovZ39+5dASCnT59uc9sTJ06Iu7u7enzw4EEBIJWVleq5jIwM8fLyUo/1er2sWrWqxetdv35dunXrJvX19Tbn/fz8ZO/evSLyZSbF0dFRLBbLd2Mzm80yevRoASABAQESHR0t+fn58unTJ7VOdHS0zJo164fj3L59uwwfPlw9HjVqlM0skohIeHi4zYyMn5+fHDlyxKbOxo0bJSws7If9EdHP4RoZIoKI/HTdixcvYsuWLXj69Cnevn2LpqYm1NfXo66uDs7OzgAAZ2dn+Pn5qW10Oh0sFgsAwGKxwGw2IyIiosXrP3z4EFarFe7u7jbnP3z4gKqqKvXYx8cHnp6e341Vp9Ph9u3bKC8vx7Vr13Dr1i1ER0dj3759KCwsRIcOrU9K5+fnIz09HVVVVbBarWhqakK3bt3U8mfPniEpKcmmTWhoKIqKigAA79+/R1VVFeLi4hAfH6/WaWpqglar/W7cRPTzmMgQEfz9/eHg4PDDBb3V1dWYMWMGEhMTkZqaCjc3N9y4cQNxcXFoaGhQExlHR0ebdg4ODmqy5OTk9N0+rFYrdDodrly50qzs61dcLb0Ca01wcDCCg4ORlJSEhIQEjB07FlevXsXEiRNbrH/79m1ERUVh/fr1MBgM0Gq1OHbsGNLS0n66T6vVCuDLrqlRo0bZlHXs2PGnr0NE38dEhojg5uYGg8GAjIwMJCcnN0sS3rx5A1dXV5SUlODz589IS0tTZzOOHz/epr5cXFzg6+uLS5cutZhIDBs2DC9fvkSnTp3g6+v7y2NqTVBQEIAvMyYAoCgKPn36ZFPn1q1b8PHxwapVq9Rz366zCQwMRHFxMRYtWqSeKy4uVn97eXlBr9fj+fPniIqK+r+Pg4i+YCJDRACAjIwMhIeHIzQ0FBs2bMDgwYPR1NSECxcuIDMzExUVFejfvz8aGxuxZ88eREZG4ubNm8jKympzX+vWrUNCQgJ69OiBadOm4d27d7h58yZMJhMmT56MsLAwzJ49G9u2bUNAQADMZjMKCgowZ84cjBgx4qf7SUxMhF6vx6RJk9CrVy/U1tZi06ZN8PT0RFhYGADA19cX58+fx7Nnz+Du7g6tVgt/f3/U1NTg2LFjGDlyJAoKCnDmzBmba5tMJsTHx2PEiBEYM2YM8vPzUVZWhn79+ql11q9fj+TkZGi1WkydOhUfP37EvXv38Pr1ayxdurTN942IWtDei3SI6N/DbDaL0WgUHx8fURRFevbsKTNnzlS3JIuI7Ny5U3Q6nTg5OYnBYJDDhw8LAHWL9j/br7925swZ+fbPTVZWlgQGBoqjo6PodDoxmUxq2du3b8VkMolerxdHR0fp3bu3REVFSU1NjYj8d/v1j5w8eVKmT58uOp1OFEURvV4v8+bNk7KyMrWOxWKRKVOmiEajsdl+vXz5cnF3dxeNRiMLFy6UXbt2NRvXhg0bxMPDQzQajcTGxkpycrKMHj3apk5eXp4MGTJEFEWR7t27y7hx435pUTURtcxBpA2r/IiIqFVTpkyBt7c3cnNz2zsUor8GXy0REf2Curo6ZGVlwWAwoGPHjjh69CguXryICxcutHdoRH8VzsgQEf2CDx8+IDIyEg8ePEB9fT0CAwOxevVqzJ07t71DI/qrMJEhIiIiu8VPFBAREZHdYiJDREREdouJDBEREdktJjJERERkt5jIEBERkd1iIkNERER2i4kMERER2S0mMkRERGS3mMgQERGR3foP60M7umvWWCAAAAAASUVORK5CYII=",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "# Plot distribution of cancer stage for integer values vs tumor size\n",
        "df_subject.plot.scatter(x='cancer_stage', y='max_tumor_size')\n",
        "\n",
        "# Title\n",
        "plt.title('Distribution of cancer stage vs tumor size for Liver')\n",
        "\n",
        "# X label\n",
        "plt.xlabel('Cancer Stage')\n",
        "\n",
        "# Y label\n",
        "plt.ylabel('Tumor Size (cm)')\n",
        "\n",
        "# show file with maximum tumor size\n",
        "df_subject.loc[df_subject['max_tumor_size'].idxmax()]\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "gather": {
          "logged": 1697512591510
        },
        "jupyter": {
          "outputs_hidden": false,
          "source_hidden": false
        },
        "nteract": {
          "transient": {
            "deleting": false
          }
        }
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "Text(0.5, 0, 'Tumor length (less than 20cm) per subject')"
            ]
          },
          "execution_count": 15,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "df3d.columns\n",
        "#Plot largest_tumor_length_cm histogram, where tumor_length_cm is less than 20cm and gretar than zero, using 100 bins\n",
        "df3d[ (df3d.major_axis_length_cm > 0) &  (df3d.major_axis_length_cm < 20)].major_axis_length_cm.hist(bins=50)\n",
        "# Title\n",
        "plt.title(\"Histogram of largest 3D tumor length less than 20cm per subject\")\n",
        "# Y label\n",
        "plt.ylabel(\"Frequency\")\n",
        "# X label\n",
        "plt.xlabel(\"Tumor length (less than 20cm) per subject\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "gather": {
          "logged": 1697512604414
        },
        "jupyter": {
          "outputs_hidden": false,
          "source_hidden": false
        },
        "nteract": {
          "transient": {
            "deleting": false
          }
        }
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Tumor area vs number of tumors')"
            ]
          },
          "execution_count": 16,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 600x600 with 3 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "# cluster the tumor_area and number_of_tumors\n",
        "from sklearn.cluster import KMeans\n",
        "kmeans = KMeans(n_clusters=3)\n",
        "kmeans.fit(df2d[['tumor_area_cm2', 'number_of_tumors']])\n",
        "df2d['tumor_cluster'] = kmeans.predict(df2d[['tumor_area_cm2', 'number_of_tumors']])\n",
        "df2d.plot.scatter(x='tumor_area_cm2', y='number_of_tumors', c='tumor_cluster', colormap='viridis')\n",
        "\n",
        "# Axis title\n",
        "plt.xlabel('Tumor area (cm^2)')\n",
        "plt.ylabel('Number of tumors')\n",
        "plt.title('Tumor area vs number of tumors')\n",
        "\n",
        "# # show density plot of tumor_area and number_of_tumors\n",
        "import seaborn as sns\n",
        "sns.jointplot(x='tumor_area_cm2', y='number_of_tumors', data=df2d, kind='kde')\n",
        "\n",
        "# Axis title\n",
        "plt.xlabel('Tumor area (cm^2)')\n",
        "plt.ylabel('Number of tumors')\n",
        "plt.title('Tumor area vs number of tumors')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "gather": {
          "logged": 1697512609641
        },
        "jupyter": {
          "outputs_hidden": false,
          "source_hidden": false
        },
        "nteract": {
          "transient": {
            "deleting": false
          }
        }
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>tumor_area_cm2</th>\n",
              "      <th>number_of_tumors</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>tumor_cluster</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>3.407788</td>\n",
              "      <td>1.002576</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>22.713283</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>84.539155</td>\n",
              "      <td>1.076923</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "               tumor_area_cm2  number_of_tumors\n",
              "tumor_cluster                                  \n",
              "0                    3.407788          1.002576\n",
              "1                   22.713283          1.000000\n",
              "2                   84.539155          1.076923"
            ]
          },
          "execution_count": 17,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df2d['tumor_cluster'].value_counts()\n",
        "\n",
        "# Get mean value of tumor_area and number_of_tumors for each cluster\n",
        "df2d.groupby('tumor_cluster')[['tumor_area_cm2', 'number_of_tumors']].mean()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "gather": {
          "logged": 1697512612326
        },
        "jupyter": {
          "outputs_hidden": false,
          "source_hidden": false
        },
        "nteract": {
          "transient": {
            "deleting": false
          }
        }
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Index(['name', 'max_tumor_size', 'number_tumors', 'cancer_stage'], dtype='object')\n"
          ]
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "df_subject = pd.read_csv('/home/azureuser/cloudfiles/code/Users/data/msd_quantification/msd_pancreas_metrics_3D_subject.csv')\n",
        "print(df_subject.columns)\n",
        "# Plot scatter plot of tumor_area vs number_of_tumors where tumor_flag is 1\n",
        "df_subject.plot.scatter(x='max_tumor_size', y='number_tumors')\n",
        "# Axis title\n",
        "plt.xlabel('Max tumor size (cm)')\n",
        "plt.ylabel('Number of tumors')\n",
        "plt.title('Tumor size (cm) vs Number of Tumors for Liver')\n",
        "plt.ticklabel_format(style='plain')"
      ]
    }
  ],
  "metadata": {
    "kernel_info": {
      "name": "python38-azureml"
    },
    "kernelspec": {
      "display_name": "Python 3.8 - AzureML",
      "language": "python",
      "name": "python38-azureml"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.8.5"
    },
    "microsoft": {
      "host": {
        "AzureML": {
          "notebookHasBeenCompleted": true
        }
      },
      "ms_spell_check": {
        "ms_spell_check_language": "en"
      }
    },
    "nteract": {
      "version": "nteract-front-end@1.0.0"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 2
}