{"metadata":{"kernelspec":{"name":"python3","display_name":"Python 3","language":"python"},"language_info":{"name":"python","version":"3.10.13","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"none","dataSources":[{"sourceId":71549,"databundleVersionId":8561470,"sourceType":"competition"}],"dockerImageVersionId":30746,"isInternetEnabled":true,"language":"python","sourceType":"notebook","isGpuEnabled":false}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"base_path = '../input/rsna-2024-lumbar-spine-degenerative-classification'","metadata":{"execution":{"iopub.status.busy":"2024-08-03T05:42:35.393906Z","iopub.execute_input":"2024-08-03T05:42:35.394275Z","iopub.status.idle":"2024-08-03T05:42:35.399353Z","shell.execute_reply.started":"2024-08-03T05:42:35.394245Z","shell.execute_reply":"2024-08-03T05:42:35.398151Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"# How to figure out sagittal image direction\n\nAs discussed in this [page](https://www.kaggle.com/competitions/rsna-2024-lumbar-spine-degenerative-classification/discussion/521997), the direction of MRI scans in this dataset is not ordered for the saggital images, meaning for some images lower instance number means left, whereas for some other lower instances number means right. \n\nThis can be inferred from the label coordinates, as their keypoints represent left or right of the patient (see fig1). As a result, model cannot learn the direction of the 3d image and cannot differentiate between left and right degeneration diagnoses. \n\nThis can be solved using the Image Position (Patient) information in the dicom metadata. As stated in [this webpage](https://blog.redbrickai.com/blog-posts/introduction-to-dicom-coordinate) about dicom images (see fig2).\n\n> The positive X direction is towards the left of the patient.\n\n> The positive Y direction is towards the back of the patient (anterior to posterior).\n\n> The positive Z direction is towards top of the patient (inferior to superior).\n\nWe know that the x coordinate in the Image Position (Patient) describes the relative direction of the scan to the patient. The higher it is, the more left it is to the actual patient. Vice versa. Therefore, we can take the first and last dicom from the folder, get the difference between the x Image Position coordinate of the two, and infer whether the series is from left to right, or right to left, in the patient's perspective. Once we get this information, we can simply flip the dicom series number for those that has unmatching orientation. See result in fig3.\n\n# How to use the code\n\nYou can take the output of this notebook. It contains an annotated df_series_description, with information of whether series is left to right.\n\n![fig1](data:image/png;base64,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)\n\n![fig2](https://cdn.prod.website-files.com/642ff723b80bac51fafabacb/6432e06f637d39358edb9b2c_1*s1N_h0tyT2DPGjH8R5G7wg.png)\n\n![fig3](data:image/png;base64,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)\n","metadata":{}},{"cell_type":"code","source":"import pandas as pd\nimport numpy as np\nimport pydicom\nfrom glob import glob\nfrom functools import partial","metadata":{"execution":{"iopub.status.busy":"2024-08-03T05:42:35.401616Z","iopub.execute_input":"2024-08-03T05:42:35.402042Z","iopub.status.idle":"2024-08-03T05:42:35.410895Z","shell.execute_reply.started":"2024-08-03T05:42:35.402004Z","shell.execute_reply":"2024-08-03T05:42:35.409816Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def get_dicom_metadata(filename):\n    dcm = pydicom.dcmread(filename)\n    result = {}\n    for element in dcm:\n        if element.name == 'Pixel Data': continue\n        result[element.name] = element.value\n    return result\ndef get_position(folder_path):\n    filename = glob(f'{folder_path}/*.dcm')[0]\n    metadata = get_dicom_metadata(filename)\n    return metadata['Patient Position']\ndef get_dicom_count(folder_path):\n    files = glob(f'{folder_path}/*.dcm')\n    return len(files)\ndef get_neural_position(series_id, dicom_count, df_co):\n    dicom_count = df_series[df_series['series_id'] == series_id].iloc[0]['dicom_count']\n    slices = df_co[df_co['series_id'] == series_id].reset_index(drop=True)\n    slices_left = slices[slices['condition'] == 'Left Neural Foraminal Narrowing'].reset_index(drop=True)\n    slices_right = slices[slices['condition'] == 'Right Neural Foraminal Narrowing'].reset_index(drop=True)\n    left_position, right_position = None, None\n    if len(slices_left) != 0:\n        left_position = sum(slices_left['instance_number'].values) / len(slices_left) / dicom_count\n    if len(slices_right) != 0:\n        right_position = sum(slices_right['instance_number'].values) / len(slices_right) / dicom_count\n    return left_position, right_position\ndef get_neural_direction(folder_path):\n    files = glob(f'{folder_path}/*.dcm')\n    files.sort(key=lambda x: int(x[:-4].split('/')[-1]))\n    x_position_first = get_dicom_metadata(files[0])['Image Position (Patient)'][0]\n    x_position_final = get_dicom_metadata(files[-1])['Image Position (Patient)'][0]\n    return x_position_first > x_position_final\n\ndf = pd.read_csv(f'{base_path}/train.csv')\ndf_series = pd.read_csv(f'{base_path}/train_series_descriptions.csv')\ndf_series = df_series[df_series['study_id'].isin(df['study_id'])].reset_index(drop=True)\ndf_series = df_series[df_series['series_description'].str.contains('Sagittal')]\ndf_series['folder_path'] = df_series.apply(lambda x: f\"{base_path}/train_images/{x['study_id']}/{x['series_id']}\", axis=1)\ndf_series['position'] = df_series['folder_path'].apply(get_position)\ndf_series['dicom_count'] = df_series['folder_path'].apply(get_dicom_count)\ndf_co = pd.read_csv(f'{base_path}/train_label_coordinates.csv')\ndf_co = df_co[df_co['series_id'].isin(df_series['series_id'])].reset_index(drop=True)\n\ndf_series['left'] = df_series.apply(lambda x: get_neural_position(x['series_id'], x['dicom_count'], df_co)[0], axis=1)\ndf_series['right'] = df_series.apply(lambda x: get_neural_position(x['series_id'], x['dicom_count'], df_co)[1], axis=1)\ndf_series['left_to_right'] = df_series['folder_path'].apply(get_neural_direction)\n\ndf_series.head(20)","metadata":{"execution":{"iopub.status.busy":"2024-08-03T05:42:35.412428Z","iopub.execute_input":"2024-08-03T05:42:35.413182Z","iopub.status.idle":"2024-08-03T05:44:51.151397Z","shell.execute_reply.started":"2024-08-03T05:42:35.413143Z","shell.execute_reply":"2024-08-03T05:44:51.150164Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_series = df_series.dropna().reset_index(drop=True)\ndf_series.head(20)","metadata":{"execution":{"iopub.status.busy":"2024-08-03T05:48:18.769826Z","iopub.execute_input":"2024-08-03T05:48:18.770188Z","iopub.status.idle":"2024-08-03T05:48:18.790518Z","shell.execute_reply.started":"2024-08-03T05:48:18.770162Z","shell.execute_reply":"2024-08-03T05:48:18.789434Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"true_case_count, false_case_count = 0, 0\nfor left, right, left_to_right in df_series[['left', 'right', 'left_to_right']].values:\n    if (left < right) == left_to_right:\n        true_case_count += 1\n    else:\n        false_case_count += 1\nprint(true_case_count, false_case_count)","metadata":{"execution":{"iopub.status.busy":"2024-08-03T05:49:50.667164Z","iopub.execute_input":"2024-08-03T05:49:50.667553Z","iopub.status.idle":"2024-08-03T05:49:50.678190Z","shell.execute_reply.started":"2024-08-03T05:49:50.667522Z","shell.execute_reply":"2024-08-03T05:49:50.676961Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import matplotlib.pyplot as plt\n\n_, axes = plt.subplots(2, figsize=(10, 10))\naxes[0].hist(df_series['left'], bins=100)\naxes[0].set_title('Left neural foraminal narrowing relative position in the entire volume')\naxes[1].hist(df_series['right'], bins=100)\naxes[1].set_title('Right neural foraminal narrowing relative position in the entire volume')","metadata":{"execution":{"iopub.status.busy":"2024-08-03T06:02:50.668552Z","iopub.execute_input":"2024-08-03T06:02:50.668947Z","iopub.status.idle":"2024-08-03T06:02:51.465745Z","shell.execute_reply.started":"2024-08-03T06:02:50.668915Z","shell.execute_reply":"2024-08-03T06:02:51.464697Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_series['left_ordered'] = df_series.apply(lambda x: x['left'] if x['left_to_right'] else 1 - x['left'], axis=1)\ndf_series['right_ordered'] = df_series.apply(lambda x: x['right'] if x['left_to_right'] else 1 - x['right'], axis=1)\ndf_series.head()","metadata":{"execution":{"iopub.status.busy":"2024-08-03T05:55:28.646055Z","iopub.execute_input":"2024-08-03T05:55:28.646853Z","iopub.status.idle":"2024-08-03T05:55:28.710346Z","shell.execute_reply.started":"2024-08-03T05:55:28.646815Z","shell.execute_reply":"2024-08-03T05:55:28.709275Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"_, axes = plt.subplots(2, figsize=(10, 10))\naxes[0].hist(df_series['left_ordered'], bins=100)\naxes[0].set_title('Left neural foraminal narrowing relative position in the entire volume (result)')\naxes[1].hist(df_series['right_ordered'], bins=100)\naxes[1].set_title('Right neural foraminal narrowing relative position in the entire volume (result)')","metadata":{"execution":{"iopub.status.busy":"2024-08-03T06:09:25.512664Z","iopub.execute_input":"2024-08-03T06:09:25.513118Z","iopub.status.idle":"2024-08-03T06:09:26.337211Z","shell.execute_reply.started":"2024-08-03T06:09:25.513089Z","shell.execute_reply":"2024-08-03T06:09:26.336002Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_series.to_csv('df_series_annotated.csv', index=False)","metadata":{"execution":{"iopub.status.busy":"2024-08-03T06:12:26.463868Z","iopub.execute_input":"2024-08-03T06:12:26.464368Z","iopub.status.idle":"2024-08-03T06:12:26.511308Z","shell.execute_reply.started":"2024-08-03T06:12:26.464327Z","shell.execute_reply":"2024-08-03T06:12:26.510171Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}