The IHE AI Workflow for Imaging Profile addresses use cases to request, manage, perform, and monitor AI Inference on digital image data. AI Inference is the process of using trained AI Models or machine learning algorithms to make a prediction.
AI Inference tasks are managed using transactions based on RESTful protocols defined in 340 DICOM UPS-RS. This profile provides flexibility in how the AI Inference task is managed. AI Inference tasks may be managed using either a pull, triggered-pull or push workflow. AIW-I defines DICOM RESTful transactions, also referred to as DICOMweb, and DICOM DIMSE transactions, sometimes referred to as Conventional DICOM.
DicomObjects .NET Core 5 version (and any future versions) fully supports the conventional DICOM DIMSE transactions and all DICOM RESTful transactions as defined in part 18 of the Standard.
The table below is a summary of all transactions as described in the IHE AI Workflow for Imaging Profile.
|Transaction||Protocal/Service||Supported by DO.CORE|
|Procedure Update (RAD-13)||HL7 v2.5.1||No|
|Retrieve Image (RAD-16)||DICOM C-MOVE||Yes|
|Creator Images Stored (RAD-18)||DICOM C-STORE||Yes|
|Instance Availability Notification (RAD-49)||DICOM N-CREATE||Yes|
|Create UPS Workitem (RAD-80)||
|Query UPS Workitems (RAD-81)||
|Claim UPS Workitem (RAD-82)||
|Get UPS Workitem Contents (RAD-83)||
|Update UPS Workitem (RAD-84)||
|Complete UPS Workitem (RAD-85)||
|Manage UPS Subscription (RAD-86)||
|Send UPS Notification (RAD-87)||
|Request UPS Workitem Cancelation (RAD-88)||DICOM N-ACTION||Yes|
|Open Event Channel (RAD-109)||Open Notification Connection Transaction of the DICOM UPS-RS Worklist Service||No|