LIDC

LIDC converts project-specific LIDC-IDRI XML nodule annotations into standardized DICOM objects to enable interoperable access and SQL-like querying of the LIDC-IDRI Computed Tomography (CT) scans for medical imaging and oncology research.

Key Features:

• Annotation format conversion: Converts project-specific LIDC XML annotations into DICOM Segmentation objects for nodules and DICOM Structured Reporting objects for nodule characterizations.
• Standardization and interoperability: Transforms annotations to DICOM to harmonize LIDC-IDRI data with other medical imaging datasets and enable use of general-purpose DICOM visualization and querying tools.
• Dataset support: Targets the LIDC-IDRI collection of over 1000 Computed Tomography (CT) scans annotated by multiple expert readers and hosted in The Cancer Imaging Archive (TCIA).
• Object-relational mapping (ORM): Implements an ORM layer using SQLAlchemy to expose dataset entities for SQL-like queries and attribute-based operations on data objects.
• FAIR alignment: Facilitates Findable, Accessible, Interoperable, and Reusable (FAIR) principles by converting annotations to standard DICOM formats.

Scientific Applications:

• Visualization and querying: Enables researchers to visualize and query LIDC nodule annotations using standard DICOM tools.
• Data harmonization: Supports cross-dataset comparisons and harmonization by standardizing annotation representation in DICOM.
• Imaging and oncology research integration: Facilitates incorporation of LIDC-IDRI annotations into broader medical imaging and oncology studies.
• Programmatic data selection: Enables selection and manipulation of nodules and annotations via SQL-like queries through the SQLAlchemy ORM.

Methodology:

Convert project-specific XML annotations to DICOM Segmentation and DICOM Structured Reporting objects, and provide an Object-relational mapping implemented with SQLAlchemy to enable SQL-like queries and attribute-based operations.