Dear Alfonso,

I have been working on preprocessing resting-state fMRI data with field maps to correct AP-PA distortion. I have 4 runs of rsfMRI images (two AP rsfMRI images and two PA rsfMRI images) and 2 runs of field map images (one AP field map and one PA field map) for one participant. My field map output from the Siemens scanner consists of just one NIfTI file without separate phase/magnitude images.

Following previous posts in this forum, I used FSL topup to create an estimated field map volume (in Hz units) from the AP and PA field map images. This process resulted in a single field map image.

I then entered that field map image into CONN as a “fmap” secondary dataset (same fmap for 4 sessions of resting-state fMRI: 2 AP and 2 PA) and used the preprocessing pipeline for volume-based analysis when field maps are available.

However, during preprocessing in CONN, I noticed a suspicious message in the history window:

“Preparing functional Creation of voxel-displacement maps (VDM) for Susceptibility Distortion Correction
units of FieldMap /Users/data/…/myfieldmap.nii not found, assuming Hz
Creating vdm file for subject 1 session 1...”

Did I make any mistakes? How can I get CONN to recognize the units of my field map? 
I would sincerely appreciate your help.

Thanks,
Hyeonmin

Dear Hyeonmin,

You may simply create a file named myfieldmap.json in the same location as your myfieldmap.nii file and with the following text content:

{
 "Units": "Hz"
}

That said, lacking this information CONN is assuming Hz units, so since that is the correct assumption in your case explicitly adding this information is not strictly necessary.

Best

Alfonso

Originally posted by Hyeonmin An:

Dear Alfonso,

I have been working on preprocessing resting-state fMRI data with field maps to correct AP-PA distortion. I have 4 runs of rsfMRI images (two AP rsfMRI images and two PA rsfMRI images) and 2 runs of field map images (one AP field map and one PA field map) for one participant. My field map output from the Siemens scanner consists of just one NIfTI file without separate phase/magnitude images.

Following previous posts in this forum, I used FSL topup to create an estimated field map volume (in Hz units) from the AP and PA field map images. This process resulted in a single field map image.

I then entered that field map image into CONN as a “fmap” secondary dataset (same fmap for 4 sessions of resting-state fMRI: 2 AP and 2 PA) and used the preprocessing pipeline for volume-based analysis when field maps are available.

However, during preprocessing in CONN, I noticed a suspicious message in the history window:

“Preparing functional Creation of voxel-displacement maps (VDM) for Susceptibility Distortion Correction
units of FieldMap /Users/data/…/myfieldmap.nii not found, assuming Hz
Creating vdm file for subject 1 session 1...”

Did I make any mistakes? How can I get CONN to recognize the units of my field map? 
I would sincerely appreciate your help.

Thanks,
Hyeonmin

Hello, Alfonso, 

Thank you so much for your response. I created the `.json` file with the required lines, but I’m still receiving a message stating that there is no unit of Hz.

I used `fslmerge` and `topup` to prepare the images, but it doesn’t seem to be correcting the distortions. Am I correct in understanding that I should use the `topup` command with the raw field map images? I merged the two AP and PA field map images using `fslmerge` and then applied `topup` to the result.

I’ve attached the images after running `topup`. Are these the correct images to use for fieldmap correction?

Thank you for your guidance.

Best,

Hyeonmin