Introduction to DICOM Imaging and FHIR
summary
DICOM handles large clinical images (often 12–16 bit, pathology files ~300 MB) and guarantees unique UIDs; FHIR mirrors that via the ImagingStudy resource which can publish endpoints pointing to DICOMweb (retrieve with QIDO/WADO or request JPEG rendering via an HTTP header). The talk cites ~two trillion images annually (50% in the U.S.), explains study→series→instance mapping, thumbnails are hard to pick, and shows a hands‑on exercise using a SIIM test server and API key to fetch real studies.
AI-written summary · about
chapters (16)
- 0:00Intro
- 1:09Slide credits go to Brad Genereaux
- 3:27Beyond Radiology: Enterprise Imaging
- 7:23DICOM is Imaging
- 8:43Image Hierarchy
- 10:18Image Metadata
- 18:02ImagingStudy (R4)
- 19:05ImagingStudy Example
- 21:02The "Series" Part
- 21:33The Instance Part
- 22:23Endpoint Resource
- 25:52Example: Get Study Thumbnail
- 31:33Thumbnail Use Case Caveats
- 33:32Enrich Modality-side Acquisition
- 34:27IHE Radiology Profiles on DICOMweb
- 37:07Society for Imaging Informatics in Medicine (SIIM)
related talks
46mDICOM and FHIR
40mDICOM web concepts: an introduction
33mSet DICOM on FHIR: metadata and images in FHIR workflows
14mConnecting the dots: FHIR, DICOM and IHE AIR
43mSmart Imaging Access
33mSponsored by DMI: This could have been a Questionnaire
40mClinical Studies: How Can FHIR Help?
19mThe Medical Informatics Initiative Germany