Project Week 25/Conversion of DICOM Single Frame MR to Enhanced Multiframe

Back to Projects List

Key Investigators

• Michael Onken (OpenConnections, Germany)
• Steve Pieper (Isomics Inc., USA)
• Andras Lasso (Queen's University, Canada)
• Csaba Pinter (Queen's University, Canada)

Project Description

Objective	Approach and Plan	Progress and Next Steps
Create tools and best practices for creating DICOM standard compliant multiframe MR instances from DICOM MR single frames. Define target use cases and input datasets What acquisition protocols will we address What sample data should we use to drive development Try to be 'document driven' in the sense that conversion should be parameterized by this document from David Clunie automatically parse xml apply to specific datasets	Approach: Define XML or JSON profile that specifies the DICOM attribute mapping required from single frame instances to the created multi-frame instance Such an editable profile makes sense to make the conversion configurable by the user. Besides the mapping rules, "commands" to the conversion engine can be added later, that filter the input, perform generalizations on attribute values, or something similar For the start, focus on a specific example to start, i.e. make sure the profile is sufficient to cover a selected use case Create such a profile from the DICOM Standard (XML) text Implement the engine (based on DCMTK) that utilizes such a profile in order to perform the conversion Probably using XSLT The plan for the project week is to have most of the XML profile done at the end of the week. The implementation of the conversion engine will happen afterwards.

Background and References

• QIICR wiki page on the topic
• Compare and contrast with this javascript prototype
• Example multiframe MR and SEG of the SPL brain atlas