open-discussion > "Automatic" distortion correction - Siemens Magnetom Skyra
Showing 1-3 of 3 posts
Display:
Results per page:
Mar 9, 2022  02:03 PM | Zac Green
"Automatic" distortion correction - Siemens Magnetom Skyra
Hello, 

For my dissertation work, I am using a large dataset of resting-state fMRI data gathered by a collaborator. They performed these scans on a Siemens Magnetom Skyra. I process the data via FSL 

In the protocol for the scans, their is a Resolution - Filter Image setting within the rsfMRI series that is listed as "Distortion Correction - On" 

My center and advisory team is not as familiar with SIEMENS scanners, and I haven't been able to get a straight answer as to if this setting means that I should skip the offline distortion correction steps that can be performed in FSL? I assume this is the case, as my attempts to apply distortion correction using the acquired field maps has only made the fMRI images look worse 

Thanks for any help that you can provide
Mar 9, 2022  04:03 PM | Paul Camacho - University of Illinois at Urbana-Champaign
RE: "Automatic" distortion correction - Siemens Magnetom Skyra
Hi Zac,

To better understand the use case, can you provide more details on the sequence that they used for the resting-state fMRI acquisition?

From our site experience with other Siemens machines, I know we use the CMRR multiband sequence for fMRI - for which the Distortion Correction setting is listed as off in the protocol. The field maps (phase encode PA and AP) used for offline TOPUP distortion correction are collected right before the resting-state bold series in this protocol.

Cheers,
Paul
Mar 10, 2022  03:03 PM | burke
RE: "Automatic" distortion correction - Siemens Magnetom Skyra
Hi Zac,

to the best of my knowledge, the "Distortion Correction" option on the Siemens Scanner refers to gradient non-linearity correction (e.g. here for more details https://mriquestions.com/gradient-linear...). This option therefore helps to correct for distortions in images with large Field-of-View, e.g. in abdominal imaging, that are caused by non-linear effects at larger distance from the isocenter. Our Siemens field applications specialist recommended to not use the distortion correction option in head imaging.

The distortions that are caused by local variations of susceptibility, e.g. in frontal parts of the brain will not be adresed by this distortion correction filter. Those susceptibility caused signal voids can be compensated by fieldmapping approaches like as the one you mentioned using two images with reversed phase encoding directions (PA and AP)  to calculate a fieldmap and then applying the fieldmap to the distorted images to recover signal.

For myself, I would not skip the distortion correction step in FSL as this step adresses a different type of distortion than the distortion correction filter on the Siemens scanner.

Best
Michael



Originally posted by Zac Green:
Hello, 

For my dissertation work, I am using a large dataset of resting-state fMRI data gathered by a collaborator. They performed these scans on a Siemens Magnetom Skyra. I process the data via FSL 

In the protocol for the scans, their is a Resolution - Filter Image setting within the rsfMRI series that is listed as "Distortion Correction - On" 

My center and advisory team is not as familiar with SIEMENS scanners, and I haven't been able to get a straight answer as to if this setting means that I should skip the offline distortion correction steps that can be performed in FSL? I assume this is the case, as my attempts to apply distortion correction using the acquired field maps has only made the fMRI images look worse 

Thanks for any help that you can provide