model_performance.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model Performance\n",
    "We can assess the models performance by comparing the AI's predictions to the manual corrections and other datasets: [Medical Segmentation Decathlon](http://medicaldecathlon.com/) (MSD), [QIN-Prostate](https://wiki.cancerimagingarchive.net/display/Public/QIN-PROSTATE-Repeatability), and [PROMISE12](https://promise12.grand-challenge.org/)\n",
    "\n",
    "## Prerequisites\n",
    "Install python packages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%capture\n",
    "%pip install pandas pydicom pydicom-seg scikit-learn seaborn segmentationmetrics requests"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Import required python packages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "from pathlib import Path\n",
    "import requests\n",
    "import zipfile\n",
    "import io\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import scipy.stats as stats\n",
    "import pydicom\n",
    "import pydicom_seg\n",
    "import SimpleITK as sitk\n",
    "import segmentationmetrics as sm\n",
    "from segmentationmetrics.surface_distance import compute_surface_dice_at_tolerance\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "sns.set_theme(style=\"whitegrid\")\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt_colors = {\n",
    "    \"ne2\": \"#5eceb0\",\n",
    "    \"rad1\": \"#9e70e1\",\n",
    "    \"tp\": \"k\",\n",
    "    \"fp\": \"b\",\n",
    "    \"fn\": \"r\",\n",
    "}\n",
    "\n",
    "revewer_cmap = sns.color_palette([plt_colors[\"ne2\"], plt_colors[\"rad1\"]])\n",
    "vol_cmap = sns.color_palette([plt_colors[\"fp\"], plt_colors[\"fn\"]])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Manual correction\n",
    "To compare model performance against the manual corrections, we need to download the dicom-seg files of the ai predictions and the manual corrections from zenodo.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "qa_dir = Path(\"qa-results\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "def download_inference(\n",
    "    extract_dir=\"qa-results\",\n",
    "    url=\"https://zenodo.org/record/10009368/files/prostate-mr.zip\",\n",
    "):\n",
    "    # download the zip file and extract it\n",
    "    r = requests.get(url)\n",
    "    z = zipfile.ZipFile(io.BytesIO(r.content))\n",
    "    z.extractall(extract_dir)\n",
    "\n",
    "\n",
    "if not qa_dir.exists() or not (qa_dir / \"ai-segmentations-dcm\").exists():\n",
    "    download_inference()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Description\n",
    "The dataset was reviewed by two reviewers, a radiologist and a 'non-expert'. The non-expert was a medical student with some familiarity with radiology images. For this project, the non-expert scored every ai prediction while the radiologist only scored the validation set.\n",
    "\n",
    "Alone with the segmentation files, we also have a qa-resulst.csv file that gives more information on the quality assessment feedback from the manual reviewers.\n",
    "\n",
    "| column                      | description                                                                                      |\n",
    "| --------------------------- | ------------------------------------------------------------------------------------------------ |\n",
    "| Reviewer                    | Either `rad1` or `ne1` for Radiologist 1 and Non-expert 1 respectively                           |\n",
    "| Validataion                 | if this scan was in the validation set                                                           |\n",
    "| Collection                  | IDC collection name                                                                              |\n",
    "| PatientID                   | PatientID from dicom                                                                             |\n",
    "| StudyDate                   | StudyDate from dicom                                                                             |\n",
    "| StudyDate_suffix            | incrementing value to differentiate different scans from the same StudyDate                      |\n",
    "| StudyInstanceUID            | StudyInstanceUID from dicom                                                                      |\n",
    "| SeriesInstanceUID           | SeriesInstanceUID from dicom                                                                     |\n",
    "| AISegmentation              | filename of ai segmentation                                                                      |\n",
    "| LikertScore                 | Reviewers opinion of how accurate the AI prediction was, see table below fr                      |\n",
    "| CommentsAboutAISegmentation | comments from the reviewer about the ai segmentation, usually how the AI underperformed          |\n",
    "| CommentsAboutScan           | comments from the reviewer about the scan, could be about unusual anatomy or scan artifacts.     |\n",
    "| CorrectedSegmentation       | file name of the manually corrected segmentation, this file is in `qa-segmentations-dcm`` folder |\n",
    "| AimiProjectYear | The year the project was completed. This is used to differentiate between the 2023 and 2024 projects. |\n",
    "\n",
    "The Likert scale used for scoring was:\n",
    "\n",
    "| Score | Meaning |\n",
    "| --- | --- |\n",
    "| 5 | Strongly Agree - Use-as-is (i.e., clinically acceptable, and could be used for treatment without change) |\n",
    "| 4 | Agree - Minor edits that are not necessary. Stylistic differences, but not clinically important. The current segmentation is acceptable |\n",
    "| 3 | Neither agree nor disagree - Minor edits that are necessary. Minor edits are those that the review judges can be made in less time than starting from scratch or are expected to have minimal effect on treatment outcome |\n",
    "| 2 | Disagree - Major edits. This category indicates that the necessary edit is required to ensure correctness, and sufficiently significant that user would prefer to start from the scratch |\n",
    "| 1 | Strongly disagree - Unusable. This category indicates that the quality of the automatic annotations is so bad that they are unusable.|\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "9.80% of scans were validated (34/347)\n",
      "34 scans were reviewed by radiologist and non-expert\n"
     ]
    }
   ],
   "source": [
    "df = pd.read_csv(\n",
    "    qa_dir / \"qa-results.csv\",\n",
    "    dtype={\n",
    "        \"PatientID\": str,\n",
    "        \"StudyDate\": str,\n",
    "        \"StudyDate_suffix\": str,\n",
    "        \"LikertScore\": int,\n",
    "    },\n",
    ")\n",
    "df.replace({\"rad1\": \"Radiologist\", \"ne1\": \"Non-expert\"}, inplace=True)\n",
    "\n",
    "# get validation subset\n",
    "val_df = df[df[\"Validation\"]]\n",
    "validation_size = len(val_df[\"SeriesInstanceUID\"].unique())\n",
    "total_size = len(df[\"SeriesInstanceUID\"].unique())\n",
    "print(\n",
    "    f\"{validation_size/total_size*100:.2f}% of scans were validated ({validation_size}/{total_size})\"\n",
    ")\n",
    "\n",
    "# get validation done by different reviewers\n",
    "rad_df = val_df.loc[val_df[\"Reviewer\"] == \"Radiologist\"]\n",
    "ne_df = val_df.loc[\n",
    "    (val_df[\"Reviewer\"] == \"Non-expert\")\n",
    "    & (val_df[\"AISegmentation\"]).isin(rad_df[\"AISegmentation\"])\n",
    "]\n",
    "\n",
    "rad_df = rad_df.sort_values(by=[\"AISegmentation\"])\n",
    "ne_df = ne_df.sort_values(by=[\"AISegmentation\"])\n",
    "\n",
    "assert len(rad_df) == len(ne_df)\n",
    "assert len(rad_df) == len(rad_df[\"AISegmentation\"].unique())\n",
    "\n",
    "print(f\"{len(rad_df)} scans were reviewed by radiologist and non-expert\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Typically, we could caluclate the Kendall-𝜏 score between the two reviewers. Kendall-𝜏 is a measure of correlation between raters of ordinal data. This tells us how much we can trust the scores given by the non-expert reviewer for the portion of the dataset the Radiologist did not review.\n",
    "However, for this dataset both of the reviewers rated all the scans in the validation set as 5. so there is 100% agreement and Kendall-𝜏 is undefined. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Kendall's 𝜏: nan, p-value: nan\n",
      "Percentage Agreement: 100.00%\n"
     ]
    }
   ],
   "source": [
    "tau, p_value = stats.kendalltau(rad_df.LikertScore.values, ne_df.LikertScore.values)\n",
    "print(f\"Kendall's 𝜏: {tau:.2f}, p-value: {p_value:.2f}\")\n",
    "\n",
    "# Calculate percentage agreement\n",
    "\n",
    "total_cases = len(rad_df.LikertScore)\n",
    "matching_cases = np.sum(rad_df.LikertScore.values == ne_df.LikertScore.values)\n",
    "percentage_agreement = (matching_cases / total_cases) * 100\n",
    "\n",
    "print(f\"Percentage Agreement: {percentage_agreement:.2f}%\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x7f3d25edf6a0>"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640.75x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "g = sns.catplot(\n",
    "    data=val_df,\n",
    "    kind=\"count\",\n",
    "    x=\"LikertScore\",\n",
    "    order=[1, 2, 3, 4, 5],\n",
    "    hue=\"Reviewer\",\n",
    "    hue_order=[\"Radiologist\", \"Non-expert\"],\n",
    "    palette=revewer_cmap,\n",
    ")\n",
    "g.set_axis_labels(\"\", \"Counts\")\n",
    "g.set(title=f\"Likert Score by Reviewer\")\n",
    "g.set_xticklabels(\n",
    "    labels=[\n",
    "        \"Strongly\\nDisagree\",\n",
    "        \"Disagree\",\n",
    "        \"Neither Agree\\nnor Disagree\",\n",
    "        \"Agree\",\n",
    "        \"Strongly\\nAgree\",\n",
    "    ],\n",
    "    rotation=60,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "# pie_df = (\n",
    "#     rad_df[\"LikertScore\"].value_counts().reset_index().sort_values(by=\"LikertScore\")\n",
    "# )\n",
    "# plt.pie(\n",
    "#     pie_df[\"count\"].values,\n",
    "#     labels=pie_df[\"LikertScore\"].values,\n",
    "#     labeldistance=0.9,\n",
    "# )\n",
    "# plt.title(\"Likert Score Distribution for Radiologist\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Each of the reviewers rated all of the validation set as 5, meaning the AI segmentations were fine as is and did not even need stylistic changes. Thats a great result, but doesn't leave us with much to look at in the way of metrics between ai and manual corrections. Instead we can compare this models performance against other datasets with existing expert segmentations.\n",
    "\n",
    "We will compare the results across three metrics: Dice Similarity Coefficient (DSC), 95% Hausdorff distance, and Normalized Mean Surface Distance (NSD).\n",
    "* DSC measures volume similarity between two segmentations.\n",
    "* 95% Hausdorff distance measures the agreement between surfaces. This is the distance were 95% of points on the surface are within the specified distance of the other surface.\n",
    "* NSD measures the average distance between the surfaces. It includes an acceptable tolerance that is dependant on the task. For prostate segmentation, the Medical Segmentation Decathlon set this tolerance to 4 mm."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_dcm_seg(seg_path: Path, label=1):\n",
    "    dcm = pydicom.dcmread(str(seg_path))\n",
    "    reader = pydicom_seg.SegmentReader()\n",
    "    result = reader.read(dcm)\n",
    "    image = result.segment_image(label)\n",
    "    image = sitk.Cast(image, sitk.sitkUInt8)\n",
    "    return image\n",
    "\n",
    "\n",
    "def load_nii_seg(seg_path: Path, label=1):\n",
    "    img = sitk.ReadImage(str(seg_path))\n",
    "    img = sitk.Cast(img == label, sitk.sitkUInt8)\n",
    "    return img\n",
    "\n",
    "\n",
    "def load_seg(seg_path: Path, label=1):\n",
    "    if seg_path.suffix == \".dcm\":\n",
    "        return load_dcm_seg(seg_path, label)\n",
    "    else:\n",
    "        return load_nii_seg(seg_path, label)\n",
    "\n",
    "\n",
    "def resize_label(img: sitk.Image, ref_img: sitk.Image, interp=sitk.sitkNearestNeighbor):\n",
    "    resampler = sitk.ResampleImageFilter()\n",
    "    resampler.SetReferenceImage(ref_img)\n",
    "    resampler.SetInterpolator(interp)\n",
    "    resampler.SetDefaultPixelValue(0)\n",
    "    resampled_img = resampler.Execute(img)\n",
    "    return resampled_img\n",
    "\n",
    "\n",
    "def calc_metrics_for_label(ai_seg_file, qa_seg_file, label_value=1, label_suffix=\"\"):\n",
    "    ai_img = load_seg(ai_seg_file, label_value)\n",
    "    qa_img = load_seg(qa_seg_file, label_value)\n",
    "    qa_img = resize_label(qa_img, ai_img)  # match the size of the ai_img\n",
    "    ai_arr = sitk.GetArrayFromImage(ai_img)\n",
    "    qa_arr = sitk.GetArrayFromImage(qa_img)\n",
    "    spacing = ai_img.GetSpacing()[::-1]  # numpy is reversed dimensions from sitk\n",
    "\n",
    "    qa_img = resize_label(qa_img, ai_img)  # match the size of the ai_img\n",
    "    m = sm.SegmentationMetrics(ai_arr, qa_arr, spacing)\n",
    "\n",
    "    metrics = {\n",
    "        \"dice\": m.dice,\n",
    "        \"hausdorff_distance_95\": m.hausdorff_distance,\n",
    "        \"mean_surface_distance\": m.mean_surface_distance,\n",
    "        \"mean_surface_distance_tol_4\": compute_surface_dice_at_tolerance(\n",
    "            m._surface_dist, 4\n",
    "        ),  # from medical segmentation decathlon\n",
    "    }\n",
    "    if label_suffix:\n",
    "        metrics = {f\"{k}_{label_suffix}\": v for k, v in metrics.items()}\n",
    "    return metrics\n",
    "\n",
    "\n",
    "def add_metrics_to_df(df):\n",
    "    for i, row in df.iterrows():\n",
    "        ai_seg_file = qa_dir / \"ai-segmentations-dcm\" / row[\"AISegmentation\"]\n",
    "        if row[\"LikertScore\"] < 5:\n",
    "            qa_seg_file = qa_dir / \"qa-segmentations-dcm\" / row[\"CorrectedSegmentation\"]\n",
    "        else:\n",
    "            qa_seg_file = ai_seg_file\n",
    "\n",
    "        assert ai_seg_file.exists()\n",
    "        assert qa_seg_file.exists()\n",
    "\n",
    "        metrics = calc_metrics_for_label(ai_seg_file, qa_seg_file)\n",
    "        metrics[\"label\"] = \"prostate\"\n",
    "        for k, v in metrics.items():\n",
    "            df.loc[i, k] = v"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Performance on other datasets\n",
    "\n",
    "Lets check how the model performs on:\n",
    "* Task 5 of the [Medical Segmentation Decathlon](http://medicaldecathlon.com/)\n",
    "* [PROMISE12](https://promise12.grand-challenge.org/)\n",
    "* [QIN-Prostate-Repeatability](https://wiki.cancerimagingarchive.net/display/Public/QIN-PROSTATE-Repeatability)\n",
    "\n",
    "Use the `prostate_mr_qa_datasets.ipynb` notebook to process and download these datasets in nifti format.\n",
    "\n",
    "Extract the zip files and you should have the following folders:\n",
    "\n",
    "* MSD/Task05_Prostate/T2imagesTr/\n",
    "* MSD/Task05_Prostate/prostateLabelsTr/\n",
    "* promise12/imagesTr/\n",
    "* promise12/labelsTr/\n",
    "* qin-prostate-repeatability/imagesTr/\n",
    "* qin-prostate-repeatability/labelsTr/\n",
    "\n",
    "### Run model\n",
    "\n",
    "Run the following commands to run the model in the container and predict the prostate segmentation for the three datasets.\n",
    "\n",
    "* MSD\n",
    "\n",
    "```bash\n",
    "mkdir -p {REPO_DIR}/MSD/Task05_Prostate/predTr\n",
    "docker run --gpus all -v {REPO_DIR}/MSD/Task05_Prostate/T2imagesTr:/data/input -v {REPO_DIR}/MSD/Task05_Prostate/predTr:/data/output bamf_prostate_mr:latest --nifti\n",
    "```\n",
    "\n",
    "* PROMISE12\n",
    "\n",
    "```bash\n",
    "mkdir -p {REPO_DIR}/promise12/predTr\n",
    "docker run --gpus all -v {REPO_DIR}/promise12/imagesTr:/data/input -v {REPO_DIR}/promise12/predTr:/data/output bamf_prostate_mr:latest --nifti\n",
    "```\n",
    "\n",
    "* QIN-Prostate-Repeatability (QPR)\n",
    "\n",
    "```bash\n",
    "mkdir -p {REPO_DIR}/qin-prostate-repeatability/predTr\n",
    "docker run --gpus all -v {REPO_DIR}/qin-prostate-repeatability/imagesTr:/data/input -v {REPO_DIR}/qin-prostate-repeatability/predTr:/data/output bamf_prostate_mr:latest --nifti\n",
    "```\n",
    "\n",
    "Then we can calcuate metrics for these datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "msd_preds = Path(\"MSD/Task05_Prostate/predTr\")\n",
    "msd_gt = Path(\"MSD/Task05_Prostate/prostateLabelsTr\")\n",
    "\n",
    "msd_df = pd.DataFrame.from_dict(\n",
    "    {\n",
    "        \"ai\": sorted(list(msd_preds.glob(\"*.nii.gz\"))),\n",
    "        \"gt\": sorted(list(msd_gt.glob(\"*.nii.gz\"))),\n",
    "        \"dataset\": \"MSD\",\n",
    "    }\n",
    ")\n",
    "\n",
    "p12_preds = Path(\"promise12/predTr\")\n",
    "p12_gt = Path(\"promise12/labelsTr\")\n",
    "p12_df = pd.DataFrame.from_dict(\n",
    "    {\n",
    "        \"ai\": sorted(list(p12_preds.glob(\"*.nii.gz\"))),\n",
    "        \"gt\": sorted(list(p12_gt.glob(\"*.nii.gz\"))),\n",
    "        \"dataset\": \"PROMISE12\",\n",
    "    }\n",
    ")\n",
    "\n",
    "qin_preds = Path(\"qin-prostate-repeatability/predTr\")\n",
    "qin_gt = Path(\"qin-prostate-repeatability/labelsTr\")\n",
    "qin_df = pd.DataFrame.from_dict(\n",
    "    {\n",
    "        \"ai\": sorted(list(qin_preds.glob(\"*.nii.gz\"))),\n",
    "        \"gt\": sorted(list(qin_gt.glob(\"*.nii.gz\"))),\n",
    "        \"dataset\": \"QIN-Prostate\",\n",
    "    }\n",
    ")\n",
    "\n",
    "df = pd.concat([msd_df, p12_df, qin_df], ignore_index=True)\n",
    "\n",
    "\n",
    "for i, row in df.iterrows():\n",
    "    ai_seg_file = row[\"ai\"]\n",
    "    qa_seg_file = row[\"gt\"]\n",
    "\n",
    "    assert ai_seg_file.exists()\n",
    "    assert qa_seg_file.exists()\n",
    "    assert ai_seg_file.name == qa_seg_file.name, f\"{ai_seg_file} != {qa_seg_file.name}\"\n",
    "\n",
    "    metrics = calc_metrics_for_label(ai_seg_file, qa_seg_file)\n",
    "    metrics[\"label\"] = \"prostate\"\n",
    "    for k, v in metrics.items():\n",
    "        df.loc[i, k] = v"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">dice</th>\n",
       "      <th colspan=\"2\" halign=\"left\">mean_surface_distance_tol_4</th>\n",
       "      <th colspan=\"2\" halign=\"left\">hausdorff_distance_95</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dataset</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MSD</th>\n",
       "      <td>0.90</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.97</td>\n",
       "      <td>0.04</td>\n",
       "      <td>3.65</td>\n",
       "      <td>0.98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PROMISE12</th>\n",
       "      <td>0.91</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.97</td>\n",
       "      <td>0.04</td>\n",
       "      <td>4.10</td>\n",
       "      <td>1.93</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>QIN-Prostate</th>\n",
       "      <td>0.88</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.95</td>\n",
       "      <td>0.06</td>\n",
       "      <td>6.06</td>\n",
       "      <td>4.62</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              dice       mean_surface_distance_tol_4        \\\n",
       "              mean   std                        mean   std   \n",
       "dataset                                                      \n",
       "MSD           0.90  0.04                        0.97  0.04   \n",
       "PROMISE12     0.91  0.04                        0.97  0.04   \n",
       "QIN-Prostate  0.88  0.04                        0.95  0.06   \n",
       "\n",
       "             hausdorff_distance_95        \n",
       "                              mean   std  \n",
       "dataset                                   \n",
       "MSD                           3.65  0.98  \n",
       "PROMISE12                     4.10  1.93  \n",
       "QIN-Prostate                  6.06  4.62  "
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# mean metric by dataset\n",
    "df[[\"dataset\", \"dice\", \"mean_surface_distance_tol_4\", \"hausdorff_distance_95\"]].groupby(\n",
    "    \"dataset\"\n",
    ").agg(\n",
    "    {\n",
    "        \"dice\": [\"mean\", \"std\"],\n",
    "        \"mean_surface_distance_tol_4\": [\"mean\", \"std\"],\n",
    "        \"hausdorff_distance_95\": [\"mean\", \"std\"],\n",
    "    }\n",
    ").round(\n",
    "    2\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "nnUNet was the winning algorithm in the Medical Segmentation Decathlon. MSD scored entries for Dice Similarity Coefficient and Noramlized Surface Distance (tolerance=4mm) for the prostate task. In MSD, there were two prostate tasts for the peripheral zone and transition zone. Both scores are given in the table for the nnUNet MSD entry.\n",
    "\n",
    "| Algorithm     | Mean DSC | Mean NSD |\n",
    "| ------------- | -------- | -------- |\n",
    "| nnUNet MSD    | 0.76, 0.90     | 0.96, 0.99     |\n",
    "| nnUNet (ours) |  0.90   |   0.97  |\n",
    "\n",
    "Comparing the model performance across stats and datasets, our model performed well and comparable to nnUNets winning Medical Decathlon entry."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x7f3c8e3ffa60>"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "g = sns.catplot(data=df, x=\"dataset\", y=\"dice\", kind=\"violin\", color=\"0.9\", inner=None)\n",
    "g2 = sns.swarmplot(data=df, x=\"dataset\", y=\"dice\")\n",
    "g.set_axis_labels(\"\", \"DSC\")\n",
    "g.set(ylim=(0.5, 1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x7f3dcc7c3eb0>"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "g = sns.catplot(data=df, x=\"dataset\", y=\"mean_surface_distance_tol_4\", kind=\"box\")\n",
    "g.set_axis_labels(\"Dataset\", \"Normalized Surface Dice\\n(tolerance=4mm)\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x7f3dcc769c00>"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "g = sns.catplot(data=df, x=\"dataset\", y=\"hausdorff_distance_95\", kind=\"box\")\n",
    "g.set_axis_labels(\"Dataset\", \"95% Hausdorff Distance (mm)\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": ".venv",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.12"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}