{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"# RSNA2023 inference notebook\n\n- Score 0.56(63rd solution)\n- [An overview of my solution](https://www.kaggle.com/competitions/rsna-2023-abdominal-trauma-detection/discussion/452255) \n","metadata":{"papermill":{"duration":0.007477,"end_time":"2023-10-12T23:34:52.772856","exception":false,"start_time":"2023-10-12T23:34:52.765379","status":"completed"},"tags":[]},"attachments":{"451afa6c-cb67-49fc-918f-dc94265c29ad.PNG":{"image/png":"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"}}},{"cell_type":"markdown","source":"# Import Libraries","metadata":{"papermill":{"duration":0.006633,"end_time":"2023-10-12T23:34:52.786734","exception":false,"start_time":"2023-10-12T23:34:52.780101","status":"completed"},"tags":[]}},{"cell_type":"code","source":"import os\nimport pandas as pd, numpy as np, random\nimport matplotlib.pyplot as plt\nimport gc\nimport cv2\nimport pydicom\n\nfrom IPython import display as ipd\nfrom glob import glob\nfrom tqdm.notebook import tqdm\n\nimport torch\nimport torch.nn as nn\nimport torch.nn.functional as F\nfrom torchvision import models\nfrom torch.utils.data import Dataset, DataLoader, Subset\nfrom torchvision.transforms.v2 import Resize, Compose, RandomHorizontalFlip, ColorJitter, RandomAffine, RandomErasing, ToTensor","metadata":{"_kg_hide-output":true,"papermill":{"duration":3.930746,"end_time":"2023-10-12T23:34:56.724417","exception":false,"start_time":"2023-10-12T23:34:52.793671","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2023-11-01T23:51:45.46788Z","iopub.execute_input":"2023-11-01T23:51:45.468996Z","iopub.status.idle":"2023-11-01T23:51:45.475385Z","shell.execute_reply.started":"2023-11-01T23:51:45.468958Z","shell.execute_reply":"2023-11-01T23:51:45.4743Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Configuration","metadata":{"papermill":{"duration":0.007108,"end_time":"2023-10-12T23:34:56.739012","exception":false,"start_time":"2023-10-12T23:34:56.731904","status":"completed"},"tags":[]}},{"cell_type":"code","source":"class CFG:\n    # device\n    device = 'cuda'\n    # seed for data-split, layer init, augs\n    seed = 42\n    # which folds of ckpt to use\n    num_folds = 1\n    # dicom to png size\n    resize_dim = 256\n    # size of training image\n    img_size = [256, 256]\n    # target column\n    target_col  = [\"bowel_healthy\", \"bowel_injury\", \"extravasation_healthy\",\n                   \"extravasation_injury\", \"kidney_healthy\", \"kidney_low\",\n                   \"kidney_high\", \"liver_healthy\", \"liver_low\", \"liver_high\",\n                   \"spleen_healthy\", \"spleen_low\", \"spleen_high\"]\n    emsemble = 'mean' # 'max' or 'mean'","metadata":{"papermill":{"duration":0.013756,"end_time":"2023-10-12T23:34:56.758685","exception":false,"start_time":"2023-10-12T23:34:56.744929","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2023-11-01T23:51:45.477576Z","iopub.execute_input":"2023-11-01T23:51:45.478213Z","iopub.status.idle":"2023-11-01T23:51:45.489249Z","shell.execute_reply.started":"2023-11-01T23:51:45.478161Z","shell.execute_reply":"2023-11-01T23:51:45.488329Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Reproducibility","metadata":{"papermill":{"duration":0.005827,"end_time":"2023-10-12T23:34:56.770614","exception":false,"start_time":"2023-10-12T23:34:56.764787","status":"completed"},"tags":[]}},{"cell_type":"code","source":"def seeding(SEED):\n    np.random.seed(SEED)\n    random.seed(SEED)\n    os.environ['PYTHONHASHSEED'] = str(SEED)\n    torch.manual_seed(SEED)\n    torch.cuda.manual_seed(SEED)\n    torch.cuda.manual_seed_all(SEED)\n    print('seeding done!!!')\nseeding(CFG.seed)","metadata":{"papermill":{"duration":0.017363,"end_time":"2023-10-12T23:34:56.794061","exception":false,"start_time":"2023-10-12T23:34:56.776698","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2023-11-01T23:51:45.490517Z","iopub.execute_input":"2023-11-01T23:51:45.490777Z","iopub.status.idle":"2023-11-01T23:51:45.501955Z","shell.execute_reply.started":"2023-11-01T23:51:45.490754Z","shell.execute_reply":"2023-11-01T23:51:45.501081Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Meta Data","metadata":{"papermill":{"duration":0.005835,"end_time":"2023-10-12T23:34:56.806104","exception":false,"start_time":"2023-10-12T23:34:56.800269","status":"completed"},"tags":[]}},{"cell_type":"code","source":"BASE_PATH = '/kaggle/input/rsna-2023-abdominal-trauma-detection'\n","metadata":{"papermill":{"duration":0.012968,"end_time":"2023-10-12T23:34:56.825241","exception":false,"start_time":"2023-10-12T23:34:56.812273","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2023-11-01T23:51:45.503095Z","iopub.execute_input":"2023-11-01T23:51:45.50339Z","iopub.status.idle":"2023-11-01T23:51:45.513426Z","shell.execute_reply.started":"2023-11-01T23:51:45.503367Z","shell.execute_reply":"2023-11-01T23:51:45.512527Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Test Paths","metadata":{"papermill":{"duration":0.00611,"end_time":"2023-10-12T23:34:56.837399","exception":false,"start_time":"2023-10-12T23:34:56.831289","status":"completed"},"tags":[]}},{"cell_type":"code","source":"test_df = pd.read_csv(f'{BASE_PATH}/test_series_meta.csv')\ntest_df['dicom_dir'] = BASE_PATH + '/' + 'test_images'\\\n                                    + '/' + test_df.patient_id.astype(str)\\\n                                    + '/' + test_df.series_id.astype(str)\ntest_dirs = test_df.dicom_dir.tolist()\n\ntest_paths = []\nfor test_dir in tqdm(test_dirs):\n    paths = sorted(glob(os.path.join(test_dir, '*dcm')),\n                   key=lambda x: int(x.split('/')[-1].split('.')[0]))\n    test_paths += [paths]\n\ntest_df['dicom_paths'] = test_paths\ntest_df = test_df[test_df.dicom_paths.map(len)>0] # in public test not all folder contains dicom file\n","metadata":{"papermill":{"duration":0.060416,"end_time":"2023-10-12T23:34:56.90408","exception":false,"start_time":"2023-10-12T23:34:56.843664","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2023-11-01T23:51:45.515353Z","iopub.execute_input":"2023-11-01T23:51:45.515629Z","iopub.status.idle":"2023-11-01T23:51:45.549989Z","shell.execute_reply.started":"2023-11-01T23:51:45.515602Z","shell.execute_reply":"2023-11-01T23:51:45.548953Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print('test_files:',test_df.shape[0])","metadata":{"papermill":{"duration":0.013124,"end_time":"2023-10-12T23:34:56.923528","exception":false,"start_time":"2023-10-12T23:34:56.910404","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2023-11-01T23:51:45.551465Z","iopub.execute_input":"2023-11-01T23:51:45.551805Z","iopub.status.idle":"2023-11-01T23:51:45.557337Z","shell.execute_reply.started":"2023-11-01T23:51:45.551772Z","shell.execute_reply":"2023-11-01T23:51:45.556457Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Dicom Utils","metadata":{"papermill":{"duration":0.007789,"end_time":"2023-10-12T23:34:58.039531","exception":false,"start_time":"2023-10-12T23:34:58.031742","status":"completed"},"tags":[]}},{"cell_type":"code","source":"def standardize_pixel_array(dcm: pydicom.dataset.FileDataset) -> np.ndarray:\n    \"\"\"\n    Source :\n    https://www.kaggle.com/code/theoviel/get-started-quicker-dicom-png-conversion\n    https://www.kaggle.com/competitions/rsna-2023-abdominal-trauma-detection/discussion/427217\n    \"\"\"\n    # Correct DICOM pixel_array if PixelRepresentation == 1.\n    pixel_array = dcm.pixel_array\n    if dcm.PixelRepresentation == 1:\n        bit_shift = dcm.BitsAllocated - dcm.BitsStored\n        dtype = pixel_array.dtype\n        pixel_array = (pixel_array << bit_shift).astype(dtype) >> bit_shift\n\n    intercept = float(dcm.RescaleIntercept)\n    slope = float(dcm.RescaleSlope)\n    center = int(dcm.WindowCenter)\n    width = int(dcm.WindowWidth)\n    low = center - width / 2\n    high = center + width / 2\n\n    pixel_array = (pixel_array * slope) + intercept\n    pixel_array = np.clip(pixel_array, low, high)\n\n    return pixel_array\n\n\ndef read_xray(path, fix_monochrome = True):\n    try:\n        dicom = pydicom.dcmread(path)\n        data = standardize_pixel_array(dicom)\n        data = data - np.min(data)\n        data = data / (np.max(data) + 1e-5)\n        if fix_monochrome and dicom.PhotometricInterpretation == \"MONOCHROME1\":\n            data = 1.0 - data\n        IMG_SIZE = [CFG.resize_dim, CFG.resize_dim]\n        data = cv2.resize(data, IMG_SIZE, cv2.INTER_LINEAR)\n    except:\n        data = np.zeros(shape=(CFG.resize_dim, CFG.resize_dim), dtype=np.float32)\n    \n    return data\n\ndef create_volume(dcm_dir, downsample_rate=1):\n\n    dcm_paths = sorted(\n            glob(os.path.join(dcm_dir, \"*dcm\")),\n            key=lambda x: int(x.split(\"/\")[-1].split(\".\")[0]),\n        )\n    downsample_rate=1\n\n    if len(dcm_paths) > 512:\n        downsample_rate = -(-len(dcm_paths) // 512)\n\n    volume = []\n    for dcm_path in dcm_paths[::downsample_rate]:\n        image = read_xray(dcm_path)\n        #image = preprocess_png(dcm_path)\n        volume.append(image)\n\n    return np.stack(volume, axis=0).astype(np.float32) # [flame, height, width]\n","metadata":{"_kg_hide-input":true,"papermill":{"duration":0.304781,"end_time":"2023-10-12T23:34:58.350835","exception":false,"start_time":"2023-10-12T23:34:58.046054","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2023-11-01T23:51:45.587383Z","iopub.execute_input":"2023-11-01T23:51:45.587676Z","iopub.status.idle":"2023-11-01T23:51:45.600376Z","shell.execute_reply.started":"2023-11-01T23:51:45.587652Z","shell.execute_reply":"2023-11-01T23:51:45.599438Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def do_scale_to_size(volume, scale_size):\n    n, l, s, s = volume.shape # scale to scale_size\n    if scale_size[0] != l:\n\n        volume = F.interpolate(\n            volume.unsqueeze(0),\n            size=scale_size,\n            mode='trilinear',\n            align_corners=False,\n        ).squeeze(0)\n\n    return volume","metadata":{"papermill":{"duration":0.014307,"end_time":"2023-10-12T23:34:58.391365","exception":false,"start_time":"2023-10-12T23:34:58.377058","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2023-11-01T23:51:45.60214Z","iopub.execute_input":"2023-11-01T23:51:45.602427Z","iopub.status.idle":"2023-11-01T23:51:45.615148Z","shell.execute_reply.started":"2023-11-01T23:51:45.602404Z","shell.execute_reply":"2023-11-01T23:51:45.614424Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def slice_volume(volume, organ_preds):\n\n    depth = volume.shape[1]\n    organ_preds_list = organ_preds.tolist()[0]\n\n    print(\"len: \" + str(depth))\n\n    if depth < 50:\n        return volume\n\n    if np.sum(organ_preds_list) < 1:\n        return volume\n    \n    ind_min = organ_preds_list.index(1)\n    ind_max = len(organ_preds_list) - list(reversed(organ_preds_list)).index(1)\n\n    # expand\n    ind_min = max(0, ind_min - 15)\n    ind_max = min(depth, ind_max + 15)\n\n    # print(\"len_organ: \" + str(len(organ_preds_list)) + \" ind_min: \" + str(ind_min) + \" ind_max: \" + str(ind_max) + \" target_frame_num:\" + str(ind_max-ind_min))\n    \n    return volume[:, ind_min:ind_max, :, :]\n\n\ndef preprocess_for_multi_stage(volume, organ_preds):\n    \"\"\"\n    preprocess the volume for 2stage model(injury predictions of 3 classes(liver kidney spleen)).\n    \"\"\"\n    volume_sliced = slice_volume(volume, organ_preds)                           # [batch_size, num_flame, height, width]\n\n    num_flame = 150\n    scale_size = (num_flame, CFG.img_size[0], CFG.img_size[0])                  # [num_flame, height, width]\n    volume_sliced = do_scale_to_size(volume_sliced, scale_size)\n\n    volume_sliced = volume_sliced.view(-1, 3, CFG.img_size[0], CFG.img_size[0]) # [num_flame/3, 3, height, width]\n    volume_sliced = volume_sliced.unsqueeze(dim = 0)                            # [1, num_flame/3, 3, height, width]\n\n    return volume_sliced\n\n\ndef preprocess_for_single(volume):\n\n    num_flame = 300\n    scale_size = (num_flame, CFG.img_size[0], CFG.img_size[0])    # [num_flame, height, width]\n    volume = do_scale_to_size(volume, scale_size)\n\n    volume = volume.view(-1, 3, CFG.img_size[0], CFG.img_size[0]) # [num_flame/3, 3, height, width]\n    volume = volume.unsqueeze(dim = 0)                            # [1, num_flame/3, 3, height, width]\n    \n    return volume","metadata":{"execution":{"iopub.status.busy":"2023-11-01T23:51:45.616283Z","iopub.execute_input":"2023-11-01T23:51:45.616544Z","iopub.status.idle":"2023-11-01T23:51:45.630421Z","shell.execute_reply.started":"2023-11-01T23:51:45.616523Z","shell.execute_reply":"2023-11-01T23:51:45.629565Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Data Check","metadata":{"papermill":{"duration":0.005932,"end_time":"2023-10-12T23:34:58.403676","exception":false,"start_time":"2023-10-12T23:34:58.397744","status":"completed"},"tags":[]}},{"cell_type":"code","source":"test_df.dicom_dir.iloc[0]","metadata":{"papermill":{"duration":0.015484,"end_time":"2023-10-12T23:34:58.425293","exception":false,"start_time":"2023-10-12T23:34:58.409809","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2023-11-01T23:51:45.631671Z","iopub.execute_input":"2023-11-01T23:51:45.632447Z","iopub.status.idle":"2023-11-01T23:51:45.646991Z","shell.execute_reply.started":"2023-11-01T23:51:45.632414Z","shell.execute_reply":"2023-11-01T23:51:45.646284Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Data Pipeline","metadata":{"papermill":{"duration":0.006247,"end_time":"2023-10-12T23:34:58.629008","exception":false,"start_time":"2023-10-12T23:34:58.622761","status":"completed"},"tags":[]}},{"cell_type":"code","source":"# dataset\nclass Abdominal_data(Dataset):\n    \"\"\"\n    Custom dataset class for handling abdominal trauma data classification.\n    \n    Args:\n        df (str): Path to the CSV file containing patient labels.\n    \"\"\"\n    \n    def __init__(self, df):\n        super().__init__()\n        \n        self.df = df.copy()\n\n        # collect all the image instance dirs\n        self.dicom_dirs = df.dicom_dir.tolist()\n        \n    def __len__(self):\n        \"\"\"\n        Returns the total number of samples in the dataset.\n        \"\"\"\n        \n        return len(self.dicom_dirs)\n    \n    def __getitem__(self, idx):\n        \"\"\"\n        Retrieves a sample from the dataset by index.\n        \n        Args:\n            idx (int): Index of the dataset to retrieve.\n        \n        Returns:\n            dict: A dictionary containing the image data and labels for different abdominal structures.\n        \"\"\"\n        dicom_dir = self.dicom_dirs[idx]\n        \n        patient_id = dicom_dir.split('/')[-2]\n\n        volume = create_volume(dicom_dir) # [flame, height, width]\n        volume = torch.tensor(volume, dtype = torch.float) \n\n        return {\n            'volume': volume,\n        }\n\n\n    def _get_subset(self, indices):\n        \"\"\"\n        Returns a subset of the dataset based on the provided indices.\n        \n        Args:\n            indices (list): List of indices to include in the subset.\n        \n        Returns:\n            Subset: A subset of the dataset.\n        \"\"\"\n        return Subset(self, indices)","metadata":{"papermill":{"duration":0.017094,"end_time":"2023-10-12T23:34:58.671824","exception":false,"start_time":"2023-10-12T23:34:58.65473","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2023-11-01T23:51:45.649299Z","iopub.execute_input":"2023-11-01T23:51:45.649575Z","iopub.status.idle":"2023-11-01T23:51:45.658768Z","shell.execute_reply.started":"2023-11-01T23:51:45.649551Z","shell.execute_reply":"2023-11-01T23:51:45.657819Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Model Configs","metadata":{"papermill":{"duration":0.006138,"end_time":"2023-10-12T23:34:58.73471","exception":false,"start_time":"2023-10-12T23:34:58.728572","status":"completed"},"tags":[]}},{"cell_type":"code","source":"# model config(single-stage model)\nCKPT_DIRS_single = [\n    ([256, 256], f'/kaggle/input/rsna-20231015-single-stage-model-1-ep2'),\n]\n\nMODEL_CONFIG_single = []\nfor img_size, ckpt_dir in  CKPT_DIRS_single:\n    paths = sorted(glob(os.path.join(ckpt_dir, '*pth')))[0:CFG.num_folds]\n    if len(paths)==0:\n        print('no model found for :',ckpt_dir)\n    MODEL_CONFIG_single.append([img_size, paths])\ndisplay(MODEL_CONFIG_single)","metadata":{"execution":{"iopub.status.busy":"2023-11-01T23:51:45.659795Z","iopub.execute_input":"2023-11-01T23:51:45.660115Z","iopub.status.idle":"2023-11-01T23:51:45.67584Z","shell.execute_reply.started":"2023-11-01T23:51:45.660092Z","shell.execute_reply":"2023-11-01T23:51:45.675015Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# model config(sub-volume detection )\nCKPT_DIRS_multi_1 = [\n    ([256, 256], f'/kaggle/input/rsna-20231015-multi-stage-model-organdet-1-ep4'),\n]\n\nMODEL_CONFIG_multi_1 = []\nfor img_size, ckpt_dir in  CKPT_DIRS_multi_1:\n    paths = sorted(glob(os.path.join(ckpt_dir, '*pth')))[0:CFG.num_folds]\n    if len(paths)==0:\n        print('no model found for :',ckpt_dir)\n    MODEL_CONFIG_multi_1.append([img_size, paths])\ndisplay(MODEL_CONFIG_multi_1)","metadata":{"papermill":{"duration":0.018293,"end_time":"2023-10-12T23:34:58.759197","exception":false,"start_time":"2023-10-12T23:34:58.740904","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2023-11-01T23:51:45.677035Z","iopub.execute_input":"2023-11-01T23:51:45.67738Z","iopub.status.idle":"2023-11-01T23:51:45.68637Z","shell.execute_reply.started":"2023-11-01T23:51:45.677351Z","shell.execute_reply":"2023-11-01T23:51:45.685435Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# model config(3 classes prediction )\nCKPT_DIRS_multi_2 = [\n    ([256, 256], f'/kaggle/input/rsna-20231015-multi-stage-model-pred-3-ep3'),\n]\n\nMODEL_CONFIG_multi_2 = []\nfor img_size, ckpt_dir in  CKPT_DIRS_multi_2:\n    paths = sorted(glob(os.path.join(ckpt_dir, '*pth')))[0:CFG.num_folds]\n    if len(paths)==0:\n        print('no model found for :',ckpt_dir)\n    MODEL_CONFIG_multi_2.append([img_size, paths])\ndisplay(MODEL_CONFIG_multi_2)","metadata":{"execution":{"iopub.status.busy":"2023-11-01T23:51:45.687526Z","iopub.execute_input":"2023-11-01T23:51:45.688019Z","iopub.status.idle":"2023-11-01T23:51:45.700972Z","shell.execute_reply.started":"2023-11-01T23:51:45.687989Z","shell.execute_reply":"2023-11-01T23:51:45.700099Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# single stage model architecture(CT-Net)\n\nclass CTNet_model(nn.Module):\n    \"\"\"Model for big data. ResNet18 then 3D conv then FC.\"\"\"\n    def __init__(self):\n        super().__init__()        \n        resnet = models.resnet18()\n        self.features = nn.Sequential(*(list(resnet.children())[:-2]))\n\n        #conv input torch.Size([1,100,512,8,8])\n        self.reducingconvs = nn.Sequential(\n            nn.Conv3d(100, 64, kernel_size = (3,3,3), stride=(3,1,1), padding=0),\n            nn.ReLU(),\n            nn.Conv3d(64, 32, kernel_size = (3,3,3), stride=(3,1,1), padding=0),\n            nn.ReLU(),\n            nn.Conv3d(32, 16, kernel_size = (3,2,2), stride=(3,2,2), padding=0),\n            nn.ReLU())\n\n        self.linear_bowel = nn.Linear(1152, 32)\n        self.silu_bowel = nn.SiLU()\n        self.drop_bowl = nn.Dropout(0.2)\n        self.bowel = nn.Linear(32, 1)\n\n        self.linear_extra = nn.Linear(1152, 32)\n        self.silu_extra = nn.SiLU()\n        self.drop_extra = nn.Dropout(0.2)\n        self.extravasation = nn.Linear(32, 1)\n\n        self.linear_kidney = nn.Linear(1152, 32)\n        self.silu_kidney = nn.SiLU()\n        self.drop_kidney = nn.Dropout(0.2)\n        self.kidney = nn.Linear(32, 3)\n\n        self.linear_liver = nn.Linear(1152, 32)\n        self.silu_liver = nn.SiLU()\n        self.drop_liver = nn.Dropout(0.2)\n        self.liver = nn.Linear(32,3)\n\n        self.linear_spleen = nn.Linear(1152, 32)\n        self.silu_spleen = nn.SiLU()\n        self.drop_spleen = nn.Dropout(0.2)\n        self.spleen = nn.Linear(32, 3)\n        \n    def forward(self, x):\n        #example shape: [1,100,3,256,256]\n        #example shape: [2,100,3,256,256]\n        shape = list(x.size())\n        batch_size = int(shape[0])\n\n        x = x.view(batch_size*100,3,256,256)\n        x = self.features(x) \n        x = x.view(batch_size,100,512,8,8)\n        x = self.reducingconvs(x)\n\n        #output is shape [batch_size, 16, 18, 2, 2]\n        x = x.view(batch_size, 16*18*2*2)\n\n\n        # output logits\n        bowel_fc1 = self.silu_bowel(self.linear_bowel(x))\n        bowel = self.bowel(self.drop_bowl(bowel_fc1))\n\n        extra_fc1 = self.silu_extra(self.linear_extra(x))\n        extravsation = self.extravasation(self.drop_extra(extra_fc1))\n\n        kidney_fc1 = self.silu_kidney(self.linear_kidney(x))\n        kidney = self.kidney(self.drop_kidney(kidney_fc1))\n\n        liver_fc1 = self.silu_liver(self.linear_liver(x))\n        liver = self.liver(self.drop_liver(liver_fc1))\n\n        spleen_fc1 = self.silu_spleen(self.linear_spleen(x))\n        spleen = self.spleen(self.drop_spleen(spleen_fc1))\n        \n        return bowel, extravsation, kidney, liver, spleen\n","metadata":{"papermill":{"duration":0.020773,"end_time":"2023-10-12T23:34:58.786685","exception":false,"start_time":"2023-10-12T23:34:58.765912","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2023-11-01T23:51:45.702144Z","iopub.execute_input":"2023-11-01T23:51:45.702435Z","iopub.status.idle":"2023-11-01T23:51:45.719096Z","shell.execute_reply.started":"2023-11-01T23:51:45.702413Z","shell.execute_reply":"2023-11-01T23:51:45.718136Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# multi-stage model architecture(sub-volume)\nclass Organ_det_model(nn.Module):\n    def __init__(self):\n        super().__init__()\n        \n        self.input = nn.Conv2d(1, 9, kernel_size = 3)\n        self.conv1 = nn.Conv2d(9, 3, kernel_size = 3)\n        self.gelu1 = nn.GELU()\n        self.gelu2 = nn.GELU()\n\n        # model = models.efficientnet_b0(weights = 'IMAGENET1K_V1')\n        model = models.efficientnet_b0()\n        \n        self.features = model.features\n        self.avgpool = model.avgpool\n        \n        self.linear_organ = nn.Linear(1280, 16)\n        self.silu_organ = nn.SiLU()\n        self.drop_organ = nn.Dropout(0.5)\n        self.organ = nn.Linear(16, 1)\n    \n    def forward(self, x):\n        \n        # extract features\n        x = self.gelu1(self.input(x))\n        x = self.gelu2(self.conv1(x))\n        x = self.features(x)\n        x = self.avgpool(x)\n        x = torch.flatten(x, 1)\n        \n        # output logits\n        organ_fc1 = self.silu_organ(self.linear_organ(x))\n        organ = self.organ(self.drop_organ(organ_fc1))\n        \n        return organ\n","metadata":{"execution":{"iopub.status.busy":"2023-11-01T23:51:45.720287Z","iopub.execute_input":"2023-11-01T23:51:45.720593Z","iopub.status.idle":"2023-11-01T23:51:45.733483Z","shell.execute_reply.started":"2023-11-01T23:51:45.720559Z","shell.execute_reply":"2023-11-01T23:51:45.73268Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# multi-stage model architecture(3-classes prediction)\n\nclass CTNet_organ_model(nn.Module):\n    \"\"\"Model for big data. ResNet18 then 3D conv then FC.\"\"\"\n    def __init__(self):\n        super().__init__()\n        # resnet = models.resnet18(weights=models.ResNet18_Weights.IMAGENET1K_V1)\n        resnet = models.resnet18()\n        self.features = nn.Sequential(*(list(resnet.children())[:-2]))\n        \n        #conv input torch.Size([1,50,512,8,8])\n        self.reducingconvs = nn.Sequential(\n            nn.Conv3d(50, 64, kernel_size = (3,3,3), stride=(3,1,1), padding=0),\n            nn.ReLU(),\n            nn.Conv3d(64, 32, kernel_size = (3,3,3), stride=(3,1,1), padding=0),\n            nn.ReLU(),\n            nn.Conv3d(32, 16, kernel_size = (3,2,2), stride=(3,2,2), padding=0),\n            nn.ReLU())\n\n        self.linear_kidney = nn.Linear(1152, 32)\n        self.silu_kidney = nn.SiLU()\n        self.drop_kidney = nn.Dropout(0.2)\n        self.kidney = nn.Linear(32, 3)\n\n        self.linear_liver = nn.Linear(1152, 32)\n        self.silu_liver = nn.SiLU()\n        self.drop_liver = nn.Dropout(0.2)\n        self.liver = nn.Linear(32,3)\n\n        self.linear_spleen = nn.Linear(1152, 32)\n        self.silu_spleen = nn.SiLU()\n        self.drop_spleen = nn.Dropout(0.2)\n        self.spleen = nn.Linear(32, 3)\n        \n    def forward(self, x):\n        shape = list(x.size())\n        #example shape: [1,50,3,256,256]\n        #example shape: [2,50,3,256,256]\n        batch_size = int(shape[0])\n        x = x.view(batch_size*50,3,256,256)\n        x = self.features(x) \n        x = x.view(batch_size,50,512,8,8)\n        x = self.reducingconvs(x)\n        #output is shape [batch_size, 16, 18, 2s, 2]\n        x = x.view(batch_size, 16*18*2*2)\n\n        # output logits\n        kidney_fc1 = self.silu_kidney(self.linear_kidney(x))\n        kidney = self.kidney(self.drop_kidney(kidney_fc1))\n\n        liver_fc1 = self.silu_liver(self.linear_liver(x))\n        liver = self.liver(self.drop_liver(liver_fc1))\n\n        spleen_fc1 = self.silu_spleen(self.linear_spleen(x))\n        spleen = self.spleen(self.drop_spleen(spleen_fc1))\n        \n        return kidney, liver, spleen\n    ","metadata":{"execution":{"iopub.status.busy":"2023-11-01T23:51:45.734842Z","iopub.execute_input":"2023-11-01T23:51:45.735168Z","iopub.status.idle":"2023-11-01T23:51:45.750093Z","shell.execute_reply.started":"2023-11-01T23:51:45.735139Z","shell.execute_reply":"2023-11-01T23:51:45.749214Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Utility","metadata":{"papermill":{"duration":0.006399,"end_time":"2023-10-12T23:34:58.799337","exception":false,"start_time":"2023-10-12T23:34:58.792938","status":"completed"},"tags":[]}},{"cell_type":"code","source":"def norm_to_one(x):\n    s = sum(x)\n    x=[xx/s for xx in x]\n    return x\n\ndef post_proc_v2(pred):\n    proc_pred = np.empty((pred.shape[0], 2*2 + 3*3), dtype='float32')\n\n    # bowel, extravasation\n    proc_pred[:, 0] = 1 - pred[:, 0] # bowel-healthy\n    proc_pred[:, 1] = pred[:, 0] # bowel-injured\n    proc_pred[:, 2] = 1 - pred[:, 1] # extra-healthy\n    proc_pred[:, 3] = pred[:, 1] # extra-injured\n    \n    # liver, kidney, sneel\n    proc_pred[:, 4:7] = pred[:, 2:5]\n    proc_pred[:, 7:10] = pred[:, 5:8]\n    proc_pred[:, 10:13] = pred[:, 8:11]\n    \n    return proc_pred","metadata":{"papermill":{"duration":0.014677,"end_time":"2023-10-12T23:34:58.820339","exception":false,"start_time":"2023-10-12T23:34:58.805662","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2023-11-01T23:51:45.751289Z","iopub.execute_input":"2023-11-01T23:51:45.752148Z","iopub.status.idle":"2023-11-01T23:51:45.76466Z","shell.execute_reply.started":"2023-11-01T23:51:45.752116Z","shell.execute_reply":"2023-11-01T23:51:45.763714Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Inference","metadata":{"papermill":{"duration":0.007396,"end_time":"2023-10-12T23:34:58.833919","exception":false,"start_time":"2023-10-12T23:34:58.826523","status":"completed"},"tags":[]}},{"cell_type":"code","source":"# fixed\nBATCH_SIZE = 1\n\ndevice = torch.device(CFG.device)","metadata":{"papermill":{"duration":0.023827,"end_time":"2023-10-12T23:34:58.872892","exception":false,"start_time":"2023-10-12T23:34:58.849065","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2023-11-01T23:51:45.765777Z","iopub.execute_input":"2023-11-01T23:51:45.766093Z","iopub.status.idle":"2023-11-01T23:51:45.775132Z","shell.execute_reply.started":"2023-11-01T23:51:45.766063Z","shell.execute_reply":"2023-11-01T23:51:45.774216Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# loading models\n\n# single-stage model\n_, fold_path_single = MODEL_CONFIG_single[0]\nmodel_single = torch.load(fold_path_single[0], map_location=device)\nmodel_single.eval()\n\n# multi-stage model(1/2 sub-volume detection)\n_, fold_path_1st = MODEL_CONFIG_multi_1[0]\nmodel_1 = torch.load(fold_path_1st[0], map_location=device)\nmodel_1.eval()\n\n# multi-stage model(2/2 prediction)\n_, fold_path_2nd = MODEL_CONFIG_multi_2[0]\nmodel_2 = torch.load(fold_path_2nd[0], map_location=device)\nmodel_2.eval()\n","metadata":{"execution":{"iopub.status.busy":"2023-11-01T23:51:45.777884Z","iopub.execute_input":"2023-11-01T23:51:45.77814Z","iopub.status.idle":"2023-11-01T23:51:45.990362Z","shell.execute_reply.started":"2023-11-01T23:51:45.778118Z","shell.execute_reply":"2023-11-01T23:51:45.989394Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Getting unique patient IDs from test dataset\npatient_ids = test_df['patient_id'].unique()\n\n# Initializing array to store predictions\npatient_preds = np.zeros(shape=(len(patient_ids), 2*2 + 3*3), dtype='float32')\n\n# Iterating over each patient\nfor pidx, patient_id in tqdm(enumerate(patient_ids), total=len(patient_ids), desc=\"Patients \"):\n    # Query the dataframe for a particular patient\n    patient_df = test_df[test_df[\"patient_id\"] == patient_id]\n    \n    # Initializing model predictions array\n    model_preds = np.zeros(shape=(1, 11), dtype=np.float32)\n    \n    print(\"=\"*25)\n    print(f\"   Patient ID: {patient_id}\")\n    print(\"=\"*25)\n    \n    # Getting volume paths for a patient\n    patient_paths = patient_df.dicom_dir.tolist()\n\n    test_data = Abdominal_data(patient_df)\n    test_dataloader = DataLoader(test_data, batch_size = BATCH_SIZE, shuffle = False)\n\n    sm = nn.Softmax()\n    sg = nn.Sigmoid()\n    b=0.; e=0.;k=0.; l=0.; s=0.\n    pred = []\n\n    for batch_data in test_dataloader:\n\n        # ------------------------------\n        # (1)single-stage prediction\n        # ------------------------------\n        # preprocess data for single stage model\n        volume_p = preprocess_for_single(batch_data['volume'])\n        inputs_single = volume_p.to(CFG.device)\n        b, e, k, l, s = model_single(inputs_single)\n\n        b = sg(b).data.cpu().numpy().copy()\n        e = sg(e).data.cpu().numpy().copy()\n        k = sm(k).data.cpu().numpy().copy()\n        l = sm(l).data.cpu().numpy().copy()\n        s = sm(s).data.cpu().numpy().copy()\n        pred += [b, e, k, l, s]\n\n        # ------------------------------\n        # (2)multi-stage prediction\n        # ------------------------------\n        # 1st stage (sub-volume)\n        o=0.\n        organ_preds = []\n        depth = batch_data['volume'].shape[1]\n        FRAME_BAT = 50\n        for d in range(0, depth, FRAME_BAT):\n            range1 = d\n            range2 = min(d + FRAME_BAT, depth)\n            input = batch_data['volume'][:, range1:range2, :, :].to(CFG.device)\n            input = input.view(-1, 1, CFG.img_size[0], CFG.img_size[0])\n            o = model_1(input)\n            o = o.view(1,-1)\n\n            o = sg(o).data.cpu().numpy().copy()\n            organ_preds += [o]\n        organ_preds = np.concatenate(organ_preds, axis=-1).astype('float32')\n        ORGAN_TH = 0.5\n        organ_preds = np.where(organ_preds > ORGAN_TH, 1,0)\n        print('organ_frame_num:' + str(np.sum(organ_preds)))\n        \n        # preprocess data for 2nd stage\n        volume_p = preprocess_for_multi_stage(batch_data['volume'], organ_preds)\n\n        # 2nd stage (3-classes prediction)\n        inputs = volume_p.to(CFG.device)\n        k, l, s = model_2(inputs)\n\n        #b=np.array([[0.]]) # fixed(no prediction)\n        #e=np.array([[0.]]) # fixed(no prediction)\n        k = sm(k).data.cpu().numpy().copy()\n        l = sm(l).data.cpu().numpy().copy()\n        s = sm(s).data.cpu().numpy().copy()\n        pred += [b, e, k, l, s]\n\n    # ------------------------------\n    # ensemble\n    # ------------------------------\n    pred = np.concatenate(pred, axis=-1).astype('float32') # reducing memory footprint\n    pred = pred[:len(patient_paths), :]\n    pred = pred.reshape(len(patient_paths)*2, 11)\n\n    if CFG.emsemble == 'max':\n        pred = np.max(pred, axis=0) # taking max prediction of all ct scans for a patient\n    else:\n        pred = np.mean(pred, axis=0) # taking average prediction of all ct scans for a patient\n\n    # Store model's prediction\n    model_preds += pred\n    \n    # Deleting variables to free up memory\n    del pred, volume_p, input, inputs_single; gc.collect()\n    \n    print('\\n')\n    \n    del test_data, test_dataloader, patient_paths; gc.collect()\n            \n    # Adding processed predictions to patient_preds\n    patient_preds[pidx, :] += post_proc_v2(model_preds)[0]\n    \n    del model_preds; gc.collect()\n\nprint(\"Prediction Done!\")","metadata":{"papermill":{"duration":10.57982,"end_time":"2023-10-12T23:35:09.459527","exception":false,"start_time":"2023-10-12T23:34:58.879707","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2023-11-01T23:51:45.991678Z","iopub.execute_input":"2023-11-01T23:51:45.991953Z","iopub.status.idle":"2023-11-01T23:51:47.9857Z","shell.execute_reply.started":"2023-11-01T23:51:45.99193Z","shell.execute_reply":"2023-11-01T23:51:47.984851Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# Submission","metadata":{"papermill":{"duration":0.006617,"end_time":"2023-10-12T23:35:09.473523","exception":false,"start_time":"2023-10-12T23:35:09.466906","status":"completed"},"tags":[]}},{"cell_type":"code","source":"# Create Submission\npred_df = pd.DataFrame({'patient_id':patient_ids,})\npred_df[CFG.target_col] = patient_preds.astype('float32')\n\n# Align with sample submission\nsub_df = pd.read_csv(f'{BASE_PATH}/sample_submission.csv')\nsub_df = sub_df[['patient_id']]\nsub_df = sub_df.merge(pred_df, on='patient_id', how='left')\n\n# Store submission\nsub_df.to_csv('submission.csv',index=False)\nsub_df.head()","metadata":{"papermill":{"duration":0.046267,"end_time":"2023-10-12T23:35:09.526574","exception":false,"start_time":"2023-10-12T23:35:09.480307","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2023-11-01T23:51:47.986675Z","iopub.execute_input":"2023-11-01T23:51:47.986923Z","iopub.status.idle":"2023-11-01T23:51:48.017565Z","shell.execute_reply.started":"2023-11-01T23:51:47.986901Z","shell.execute_reply":"2023-11-01T23:51:48.016654Z"},"trusted":true},"execution_count":null,"outputs":[]}]}