Sorry again for a “newbie” question, but I couldn’t find an answer, yet.

I have several dicom sequences for gre field mapping.
In order to perform field map correction, I assume I should create nii-s from them- one for magnitude, and other for phase.
How do I know from the dicom headers (or from the niftis, after a conversion) which is for the magnitude nii and which is for the phasediff?
(Any tutorial about field mapping calculation for beginners would be appreciated too)

I assume that the nii with field:
“EchoNumber”: 2,
is the nii for phase.

And I also assume that the following line tells which is which:
“ImageType”: [“ORIGINAL”, “PRIMARY”, “P”, “ND”],
Vs.
“ImageType”: [“ORIGINAL”, “PRIMARY”, “M”, “ND”],

dcm2niix issue 194
Depending on how the scanner was set up, you will probably have two echos, with either

One magnitude and one phase map per echo

One real and one imaginary image per echo

As Brai noted, if you have a Philips scanner the BIDS ImageType may tell you whether you have "P"hase, "M"agnitude, “R"eal or “I"maginary. If you have the latest version of dcm2niix converting your data it should recognize echo and image type by adding the extension “_e1”/”_e2” to denote echo and “” for magnitude, “_ph” for phase, “_real” for real and “_imaginary” for imaginary. The appearance of the image will also help you reveal these features. If you have real and imaginary images, you can use fslcomplex to combine them as a complex image and then run fslcomplex again to convert the complex image to phase and magnitude. I also wrote a Matlab/SPM script and I am sure if you look around you can find a numPy solution.