Hi everyone, 

When I use the MRIcroGL to convert files into nii, I got a cropped image (image attached). What could it be?

Thanks

What version of dcm2niix was used for the conversion of your DICOM images to NIfTI? I think if you convert your images with a recent release of dcm2niix it will resolve your problem. My guess is these images were acquired using a new proprietary encoding method (e.g. zip2) that requires a custom kludge in the image conversion tool. If my guess is correct, the latest (1.0.20241208) version of dcm2niix will resolve your issue.

https://github.com/rordenlab/dcm2niix/releases

Hi Chris,

Thanks for your reply. 

Per your suggestion, I used the most recent release but still got the same cropped image.  

I checked the JSON file, and I did not find any mention of zip2. 

1- Here is what appears in the MRIcroGL window when I convert just this acquisition:

C:\Users\username\AppData\Local\Temp\Rar$EXa4768.24269\MRIcroGL\Resources\dcm2niix.exe -f "%p_%i" -p y -z y -o "C:\Users\username\Desktop\mricrongl" "C:\Users\username\Desktop\mricrongl\raw\DICOM_MRI_STUDY-019\20240311\15520000"

Chris Rorden's dcm2niiX version v1.0.20241208  (JP2:OpenJPEG) (JP-LS:CharLS) MSC1942  (64-bit Windows)

Found 482 DICOM file(s)

Convert 482 DICOM as C:\Users\username\Desktop\mricrongl\ses-01_task_run-02_bold_MRI_STUDY_019a (80x80x26x482)

Conversion required 13.343000 seconds..

2- Below is what I get in the window; if I try to convert several acquisitions simultaneously, I get error messages, including one about slices not being stacked.  But I think this happens because the localizer is included in the folder. 

C:\Users\username\AppData\Local\Temp\Rar$EXa4768.31863\MRIcroGL\Resources\dcm2niix.exe -f "%p_%i" -p y -z y -o "C:\Users\username\Desktop\mricrongl" "C:\Users\username\Desktop\mricrongl\raw\DICOM_MRI_STUDY-019\20240311"

Chris Rorden's dcm2niiX version v1.0.20241208  (JP2:OpenJPEG) (JP-LS:CharLS) MSC1942  (64-bit Windows)

Found 1670 DICOM file(s)

Convert 482 DICOM as C:\Users\username\Desktop\mricrongl\ses-01_task_run-02_bold_MRI_STUDY_019 (80x80x26x482)

Slices not stacked: orientation varies (vNav or localizer?) [0 1 0 0 0 -1] != [1 0 0 0 0 -1]

Warning: Issue797: Check slice timing range 0..10.2969, TA= 10.2969, TR=3.98 ms)

Warning: Siemens MoCo? Bogus slice timing (range -1..-1, TR=3.98 seconds)

Warning: Weird CSA 'ProtocolSliceNumber' (System/Miscellaneous/ImageNumbering reversed): VALIDATE SLICETIMING AND BVECS

Convert 3 DICOM as C:\Users\username\Desktop\mricrongl\trufi_localizer_3-plane_MRI_STUDY_019_i00001 (256x256x3x1)

Warning: Issue797: Check slice timing range 0..10.3125, TA= 10.3125, TR=3.98 ms)

Warning: Siemens MoCo? Bogus slice timing (range -1..-1, TR=3.98 seconds)

Warning: Weird CSA 'ProtocolSliceNumber' (System/Miscellaneous/ImageNumbering reversed): VALIDATE SLICETIMING AND BVECS

Convert 3 DICOM as C:\Users\username\Desktop\mricrongl\trufi_localizer_3-plane_MRI_STUDY_019_i00004 (256x256x3x1)

Warning: Issue797: Check slice timing range 0..10.3125, TA= 10.3125, TR=3.98 ms)

Warning: Siemens MoCo? Bogus slice timing (range -1..-1, TR=3.98 seconds)

Convert 3 DICOM as C:\Users\username\Desktop\mricrongl\trufi_localizer_3-plane_MRI_STUDY_019_i00007 (256x256x3x1)

Convert 192 DICOM as C:\Users\username\Desktop\mricrongl\t1mprage_sag_MRI_STUDY_019 (256x256x192x1)

Convert 471 DICOM as C:\Users\username\Desktop\mricrongl\ses-01_task_run-01_bold_MRI_STUDY_019 (80x80x26x471)

Convert 325 DICOM as C:\Users\username\Desktop\mricrongl\REST_MB4_MRI_STUDY_019 (112x112x72x325)

Slices not stacked: dimensions vary across slices

Convert 190 DICOM as C:\Users\username\Desktop\mricrongl\t1mprage_sag_MRI_STUDY_019a (256x256x190x1)

Convert 1 DICOM as C:\Users\username\Desktop\mricrongl\t1mprage_sag_MRI_STUDY_019b (484x484x1x1)

Conversion required 77.504997 seconds.

In both cases, the image of the run_01(02)_bold appears cropped.

Thanks for your help.

I am unable to understand this behavior. The note about slice sizes varying makes me think that this dataset has been manipulated by different tools. I would check the provenance and see if the DICOMs exported directly from the scanner convert correctly. Feel free to share this dataset with my institutional email if you want more feedback:

  https://sc.edu/study/colleges_schools/ar...