LNDb v4

LNDb v4 provides pulmonary nodule annotations derived from medical reports linked to manually annotated lung scans to facilitate research into lung cancer diagnosis.

Key Features:

• Report-to-Annotation Conversion: Converts medical reports into nodule annotations by matching textual information with manually annotated data from the Lung Nodule Database (LNDb).
• Tabular Data Release: Provides a tabular data file containing information from 292 medical reports in the LNDb.
• Nodule Characterization Files: Includes files detailing nodule characteristics and corresponding matches to the manually annotated data.
• Cross-Modal Data Integration: Bridges textual medical reports with image-based manual annotations to enable comparative research between these two data types.

Scientific Applications:

• Lung Cancer Diagnosis Research: Enables further research studies in lung cancer by connecting existing medical reports with manual annotations.
• Computer-Aided Diagnosis Development: Supports the development of systems that streamline the identification of pulmonary nodules by leveraging medical reports as sources for automatic image annotation retrieval.
• Comparative Data Type Analysis: Fosters research on the advantages and disadvantages of report-derived annotations versus manual image annotations.

Methodology:

Medical reports were converted into nodule annotations by matching textual information with manually annotated data from the Lung Nodule Database (LNDb).