{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "0e9e5b6e",
   "metadata": {},
   "source": [
    "# OrcaImage segmentation baseline\n",
    "\n",
    "## Goal\n",
    "Train a small MONAI model, run inference while preserving NIfTI geometry, and create a submission archive for the `segmentation-v1` evaluator."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "81beecc6",
   "metadata": {},
   "source": [
    "## Setup\n",
    "Install the starter dependencies and unpack the synthetic Liver CT bundled inside this notebook. No separate dataset download is required for the demo. GPU is optional for this smoke baseline."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6405da3a",
   "metadata": {},
   "outputs": [],
   "source": [
    "from pathlib import Path\n",
    "import importlib.util, subprocess, sys\n",
    "required = ['torch', 'monai', 'nibabel', 'numpy', 'matplotlib']\n",
    "missing = [name for name in required if importlib.util.find_spec(name) is None]\n",
    "if missing:\n",
    "    subprocess.check_call([sys.executable, '-m', 'pip', 'install', '-r', 'requirements.txt'])\n",
    "import torch\n",
    "print('PyTorch', torch.__version__)\n",
    "print('device', 'cuda' if torch.cuda.is_available() else 'cpu')\n",
    "assert Path('train.py').exists(), 'Run this notebook from the extracted starter-kit directory.'"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3ff5d012",
   "metadata": {},
   "source": [
    "### Bundled demo data\n",
    "The following cell restores a small synthetic Liver CT from the notebook itself. It is for validating the pipeline and file contract, not for clinical model development."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7c9c26a1",
   "metadata": {},
   "outputs": [],
   "source": [
    "import base64, io, zipfile\n",
    "from pathlib import Path\n",
    "demo_payload = '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'\n",
    "data_dir = Path('data')\n",
    "data_dir.mkdir(exist_ok=True)\n",
    "with zipfile.ZipFile(io.BytesIO(base64.b64decode(demo_payload))) as archive:\n",
    "    archive.extractall(data_dir)\n",
    "print('bundled Liver demo:', sorted(str(path) for path in data_dir.rglob('*.nii.gz')))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a5266ccb",
   "metadata": {},
   "source": [
    "## Steps\n",
    "### 1. Train a deterministic smoke model\n",
    "The demo uses generated anatomy so the notebook remains runnable when test labels are hidden. Replace this with the published training split for a real competition."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "469d9b45",
   "metadata": {},
   "outputs": [],
   "source": [
    "%run train.py"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c3ae0924",
   "metadata": {},
   "source": [
    "### 2. Inspect the public cohort"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f278a5a4",
   "metadata": {},
   "outputs": [],
   "source": [
    "import nibabel as nib\n",
    "import matplotlib.pyplot as plt\n",
    "images = sorted(Path('data').rglob('*.nii.gz'))\n",
    "print(f'{len(images)} public image(s)')\n",
    "if images:\n",
    "    sample = nib.load(images[0])\n",
    "    volume = sample.get_fdata()\n",
    "    print(images[0], volume.shape, sample.header.get_zooms()[:3])\n",
    "    plt.imshow(volume[:, :, volume.shape[2] // 2].T, cmap='gray', origin='lower'); plt.axis('off');"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f66c6b40",
   "metadata": {},
   "source": [
    "### 3. Run inference and package predictions\n",
    "The output retains the source affine and uses the required `labels/<case-id>.nii.gz` paths."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b075f6e7",
   "metadata": {},
   "outputs": [],
   "source": [
    "if images:\n",
    "    %run inference.py --data data --checkpoint outputs/model.pt\n",
    "    %run make_submission.py --predictions outputs/predictions --output submission.zip\n",
    "else:\n",
    "    print('Download and extract the challenge dataset first; then rerun this cell.')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5f0936ee",
   "metadata": {},
   "source": [
    "## Checks\n",
    "Before uploading, confirm that every expected case exists, shapes and affines match the public images, and only declared integer labels are present."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b5f74d20",
   "metadata": {},
   "outputs": [],
   "source": [
    "import zipfile\n",
    "if Path('submission.zip').exists():\n",
    "    with zipfile.ZipFile('submission.zip') as archive:\n",
    "        print('submission members:', archive.namelist())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cf4fff02",
   "metadata": {},
   "source": [
    "## Next steps\n",
    "Replace synthetic training with the challenge training cohort, add MONAI transforms and cross-validation, then retain only inference and packaging in the final reproducible run."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}
