Hi Rob,
hi Ragini-
> I wonder if scanning a simple water phantom would show reduced signal in the direction of the vibration if the bed were suitably loaded
A gel phantom is probably better to provide some diffusion signal and to visualize the vibration-induced signal loss..
See also:
Ogura A, Maeda F, Miyai A, Hayashi K, Hongoh T (2006): Effect of vibration caused by time-varying magnetic fields on diffu- sion-weighted MRI. Nippon Hoshasen Gijutsu Gakkai zasshi 62:565–569.
> Siemens has since rectified the problem
The fix does not completely eliminate the vibrations but it attenuates them and shifts the fundamental frequency peak to lower values so that the artifact does not become apparent.
Note that, according to my experience, eddy cannot "clean" the vibration artifact from the data.
This is because it consistently affects and increases with the diffusion weighting in x.
Eddy cannot "create" clean data when you never measure it.
It can when there are outliers but in this situation the data is not randomly affected.
What you could do is to build a confound regressor representing the artefact (see https://pdfs.semanticscholar.org/8e89/03f8092c27d159879a3a2429a771d21b7be8.pdf), run dtifit using the --cni flag and --save_tensor and dtigen to "build" "clean" diffusion data from that.
It may clearly not be an optimal solution for you but potentially worth a trial depending on what you are aiming to do.
Cheers,
Andreas
Am 16.01.19, 22:56 schrieb "FSL - FMRIB's Software Library im Auftrag von Reid, Robert I. (Rob)" <[log in to unmask] im Auftrag von [log in to unmask]>:
You might catch head motion caused by subject irritation due to vibration that way, but eddy's motion estimates are necessarily based on volume-to-volume measurements, so most vibrations will be too fast to detect that way.
See https://onlinelibrary.wiley.com/doi/full/10.1002/hbm.22846 and https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3906573/ for a different approach. There's also another paper about doing "auto-actuated" (i.e. bed + diffusion gradient actuated) MRE, but the MRE experts across the hall prefer to use purpose-built actuators.
I wonder if scanning a simple water phantom would show reduced signal in the direction of the vibration if the bed were suitably loaded.
--
Robert I. Reid, Ph.D. | Sr. Analyst/Programmer, Information Technology
Aging and Dementia Imaging Research | Opus Center for Advanced Imaging Research
Mayo Clinic | 200 First Street SW | Rochester, MN 55905 | mayoclinic.org
-----Original Message-----
From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf Of Jacob Antony Alappatt
Sent: Tuesday, January 15, 2019 3:21 PM
To: [log in to unmask]
Subject: [EXTERNAL] [FSL] Eddy motion parameters - Usage
Our diffusion data was acquired with a bed that had been reportedly vibrating. We want to make sure none of our results are correlated with motion.
After running eddy (from FSL/5.0) we end up with 6 motion (3 rotation & 3 translation) parameters, for each volume of the acquisition. If I want a single number from this acquisition, to be used as a regressor, or for correlations, how would I post-process this file?
It can't be as simple as averaging across all volumes, plus I don't know how to transform the 6 motion parameters into a single representative value, per volume. Any help on this would be great.
########################################################################
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
########################################################################
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
########################################################################
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
|