Siemens DICOM B0map to .nii

Software Support
, ,
1

Hi,

I am trying to learn how to use fieldmap images to do distortion correction. I am using a SIEMENS scanner and the fieldmap sequence that came with it. It has stored two “B0map” files for me. The other has 2 sets of magnitude images and the other has one phase difference image. They are in DICOM format. I have usually used freesurfers mri_convert for the structural images and also for the EPI series and it has worked perfectly. Now that I try to use the same command for these it says:

Found 42 DICOM file(s)
error: No such file or directory
error: niiReadFromMriFsStruct(): bad number of dimensions (0)

is there a flag that I should be using for this? How can I convert the magnitude and phase images to nifti?

Also the two Magnitude images are stored in the way that they have almost the same name for example:
1.3.12.2.1107.5.2.19.45177.30000025072105262377000000025-4-1-enya21
1.3.12.2.1107.5.2.19.45177.30000025072105262377000000025-4-1-zav2tz
1.3.12.2.1107.5.2.19.45177.30000025072105262377000000025-4-2-1i1z3gn
1.3.12.2.1107.5.2.19.45177.30000025072105262377000000025-4-2-hlzzbx
and so on, first of all: am I supposed to use only one of these images for
fsl_prepare_fieldmap <phase_image> <magnitude_image> <out_image> <deltaTE (in ms)>
or should I know which to use, and how to do the separation or merging of these images?

Thank you in advance!
-Nella

2

I am not familiar with mri_convert, so will let other experts answer on that. You may want to try out dcm2niix and see what information it provides on errors. I provide notes on Siemens fieldmaps here, that note the two important caveats. First, your acquisition must be set to save both the magnitude and phase. Second, Siemens will use the same instance number for both echos, so you need to make sure that your PACS archive does not overwrite some images with others, in which case you can get errors. Perhaps this is why you are getting the “bad number of dimensions”.

Note that you will want to provide the NIfTI not the DICOM images to fsl_prepare_fieldmap. The filenames you provide sound like DICOM filenames. With dcm2niix, you can choose your filenaming, so -f %m_%s_%p will create a human-readable filename with the manufacturer, series number and protocol name, e.g. Si_2_gre_field_magph_e2.json.

If you use dcm2niix, it will store the echo times in the JSON sidecar, and you can use that to determine the difference, aka deltaTE.