Closed jpcoutu closed 6 years ago
By the way, I am happy to provide any information necessary regarding this added feature here or by email. We have been using this at Biospective since the pull request without any problem.
Looks like @rdvincent is ignoring this....
I realized that I was not freeing the memory allocated for sort_array, so I pushed a fix for that. Also, now that I have more experience looking into dcm2mnc I am thinking this could probably have been coded differently using only the general info pointer (similarly to how indices and coordinates are stored for sort_dimensions()), rather than using the file pointers. Not sure if that makes a big difference however.
I also noted in the add_philips_info sub that there is a TODO to examine whether the trigger time is ever useful in MRI. Just wanted to point out that it is useful to determine the GEMS slice order for fMRI as demonstrated in this code.
Oops, sorry it fell off my radar
This changes dcm2mnc to set the mincheader field acquisition:slice_order for GEMS fMRI scans. It calculates the slice order from the TemporalPositionIdentifier, TriggerTime and SliceLocation dicom tags when available. We tested this on our end and it works, and does not seem to alter normal behavior in other cases (e.g. Siemens and Philips scans).