Successes and challenges in extracting information from DICOM image databases for audit and research.

The article focuses on the successes and challenges of extracting metadata from Digital Imaging and Communications in Medicine (DICOM) image databases, particularly Picture Archiving and Communication Systems (PACS), for audit and research purposes in radiography, with an emphasis on breast screening. It highlights the clinical and research benefits of accessing rich metadata stored alongside medical images, such as workflow optimization, clinical audits (e.g., compression force variability and dose audits), evaluation of clinical outcomes, and support for artificial intelligence development. Key challenges include limited and inefficient PACS search tools, data security and governance concerns, difficulties in anonymization, inconsistent metadata standards, and the complexity of linking imaging data with clinical records. The article reviews existing solutions like retrospective and prospective data extraction, use of DICOM structured reports, and Trusted Research Environments (TREs), emphasizing that improved data access would be facilitated by specifying requirements during PACS procurement. Overall, enhanced metadata extraction and utilization can improve imaging quality, workflow, and patient care while supporting research and innovation.