• To identify or develop the DICOM mechanisms to support AI workflows, concentrating on the clinical context.  This includes the interfaces between application software and the back end DICOM infrastructure.

Current Supplements, Work and Objectives:

• Sup 219 JSON Representation of DICOM Structured Reports out for trial implementation
•  Sup 224 Service Discovery and Control, including review of Supplement 19 (in progress – may be withdrawn or completely revised).
• Strong desire to use more modern, simpler technology in Application Hosting, leveraging work done in DICOMweb.
• Desire to better support AI workflows
• Work item for creating examples using existing mechanisms for Application Hosting

Challenges and Opportunities (Environment):

• Mechanisms that are appropriate for one regulatory body are inappropriate for another.
• Ensuring that Application Hosting is of sufficient interest to the potential contributors to maintain active participation by experts, champions, vendors and clinical users.
• Requests for future work items will be dependent on the active participation of vendors and clinicians that perceive a need to expand the DICOM Standard in a specific area.
• Some items may require cooperation from other working groups, including WG-11 (Display), WG-17 (3D), WG-18 (Clinical Trials and Education), WG-20/HL7 IIWG, WG-27 (DICOMweb), all of the modality-specific working groups, and the HL-7 FHIR working group.
• Work may require shifts in the scope of the working group.
• Staying focused on the task at hand.
• The mechanisms utilized could become obsolete or broken.
• The technological landscape is constantly shifting, with various tools falling in and out of favor
• The use cases are continually shifting
• Regulatory issues may limit the clinical utility of the Standard.
• Specifying mechanisms that are easily incorporated and do not conflict with work done by other bodies.

Future Roadmap and Objectives (Committee Direction):

• Continued monitoring of potential use cases. There is concern that current DICOM mechanisms might not be adequate to cover some use cases, particularly bulk analysis of large repository data, e.g. for training deep learning neural networks.
• The WG expects to leverage existing Standards, insofar as possible. The WG has closely cooperated with HL7 in the past, and expects to continue to monitor what is happening in that space (e.g. FHIR).
• Further flesh out use cases/scenarios where applications of Artificial Intelligence or other analysis methods are a primary focus.
• The working group is considering, based on use examples, how to represent analytic results in DICOM in a machine-consumable and interoperable format.

Past Work

• Supplement 118 FT 2009
• Supplement 219 FD 2019
• Assisted with two IHE Profiles
• Application Hosting was included in the IHE Post Processing Work Flow (PPWF) profile.