TCIA_Notebooks

This is a home for notebooks which demonstrate how to access and work with TCIA datasets. Most of them heavily leverage functionality from tcia_utils.

General Notebooks

• TCIA_Linux_Data_Retriever_App.ipynb - A tutorial on how to install the NBIA Data Retriever command-line Data Retriever utility on Linux and use it to download TCIA datasets
• TCIA_REST_API_Queries.ipynb - A Python tutorial on how to use the NBIA REST API to query radiology (DICOM) datasets
• TCIA_REST_API_Downloads.ipynb - A Python tutorial on how to use the NBIA REST API to download radiology (DICOM) datasets
• TCIA_Segmentations - A Python tutorial focused on using the TCIA APIs to identify segmentation data, find the corresponding reference series and visualize them together.
• TCIA_Series_UID_Report.ipynb - Ingests a file containing TCIA Series Instance UIDs (e.g. TCIA manifest file or CSV of UIDs) and creates reports that summarize those scans
• TCIA_Aspera_CLI_Downloads.ipynb - A short tutorial on how to download TCIA datasets that are made available through Aspera via the command line (rather than via the Aspera browser plugin). TCIA typically uses Aspera for downloading histopathology collections or radiology collections that were provided in a format other than DICOM.
• TCIA_DataCite_Queries.ipynb - TCIA issues a Digital Object Identifier (DOI) for each of its datasets through DataCite. This notebook demonstrates how the DataCite API can be used to programmatically access Collection metadata such as their DOI URL, title, publication date, licensing information and abstract.

AI and Visualization Notebooks

• TCIA_Image_Visualization_with_itkWidgets.ipynb - Example of downloading TCIA DICOM images and visualizing them as interactive cinematic volume renderings or as 2D slices, using itkWidgets.
• TCIA_MONAI_Model_Zoo.ipynb - Demonstrates downloading data from TCIA, downloading a pre-trained model from MONAI's Model Zoo, applying the model to the data to segment anatomic structures in that data, and then visually comparing model results with expert segmentations.
• TCIA_RTStruct_SEG_Visualization_with_itkWidgets.ipynb - Tutorial of downloading expert annotations as DICOM SEG and RTSTRUCT objects, converting them to labelmaps for use in training and evaluating AI models, and visualizing them with their source images in 3D or as overlays on 2D slices.
• TCIA_STL_Visualization_with_itkWidgets.ipynb - Shows how to download, convert, and visualize expert annotations and 3D printer models stored in STL format on TCIA for use in training and evaluating AI models.

Collection-specific Notebooks

• CPTAC.ipynb - A tutorial on accessing DICOM images and tumor annotations (3d segmentations & seed points) related to the CPTAC-UCEC (Corpus Endometrial Carcinoma), CPTAC-PDA (Pancreatic Ductal Adenocarcinoma), CPTAC-CCRCC (Clear Cell Renal Carcinoma), and CPTAC-HNSCC (Head and Neck Squamous Cell Carcinoma) datasets hosted on TCIA.
• TCIA_NCTN_Annotations.ipynb - A tutorial on accessing DICOM images, clinical data, and tumor annotations (3d segmentations & seed points) related to NCI Clinical Trial Network datasets hosted on TCIA.
• TCGA_Clinical.ipynb - A tutorial on accessing DICOM images and clinical data for patients from The Cancer Genome Atlas Program.