Closed davidmummy closed 2 months ago
Add Branch to fix the issue
Update -- we will also implement the vendor-specific image orientation (#57 ) as part of this issue, since the complex conjugate is tied up in the orientation for GE scanners. Here is the spec for that one:
Different vendors write their data differently to the MRD file, resulting in the orientation of images having different coordinate systems after processing.
Desired functionality:
Read the scanner_type from the MRD file (philips, ge, siemens)
Implement the correct sequence of flips and rotations for each vendor so that the results are in the same coordinate system across sites (i.e. the standard Duke system). We can just base this on our current test data from Cincinnati and Iowa.
This is ready to work at your convenience.
Assigning to myself to review.
Tested on MRDs from Duke subject 009-024A and Cincy ILD-HC-068_V2. Everything looks good and properly oriented.
Have requested updated MRD from Iowa to complete the collection. Will update as soon as I get it and then I think this will be ready to close.
When reading in the MRD, please check the scanner type. If it is GE, apply a complex conjugate to the data. Otherwise, do nothing.
This mirrors the functionality currently in our spectroscopy pipeline as well.