Data Extraction is the process of linking a relational dataset with a Project cohort and generating an anonymous data release. This may include several datasets (e.g. biochemistry, prescribing etc) which might be filtered (e.g. only extract prescriptions for painkillers).
Image tables are treated exactly like regular datasets by RDMP. You can build a cohort from the tables (based on PatientID), link with other datasets and produce final cohort lists.
You can extract images as part of a Project by including the imaging table in the list of datasets extracted for the Project. Since the imaging table is just a normal table from RDMP's point of view this will result in a CSV file being produced with anonymised tag data.
If you want the images to also be fetched and anonymised then you should add the FoDicomAnonymiser component to your extraction pipeline (it is recommended to first clone your normal extraction pipeline).
The FoDicomAnonymiser looks at the System.Data.DataTable passing through the extraction pipeline. When it spots a dataset containing imaging file paths it copies them to the output directory after stripping identifiable tags and substituting UIDs / PatientIDs for suitable anonymous mappings.
Mappings are persisted so that repeated extractions of the same images for the same Project result in the same images being produced.
Extraction Component FoDicomBasedExtraction