processing-scripts > Quality assurance at your site
Showing 1-2 of 2 posts
Feb 11, 2014 10:02 PM | Matthias Heil
Quality assurance at your site
Dear forum participants,
Our lab is developing a quality assurance pipeline to automatically check and generate reports of the acquired MR data. My task is to find out, what is already done at different sites and labs.
I would be glad if you like to share your QA steps (and maybe scripts). If you're interested in my inquiry results, I also would share them here.
I'm especially interested, if you're using fsl tools for that.
We are currently working on motion QA using python for website-embedded video reports (and already using mcflirt parameters), normalization results with reports of different slices and their overlap with MNI template.
In automated artefact detection, we're at the moment focussing on ghosting, spikes and zipper artefacts.
For quantified checks, we're propably will be using tSNR for EPI sequences, SNR for structural images. Since I'm just a beginner in the neuroscience field and the project has been started, thats all I can tell you right now :)
What I've already found:
http://cbs.fas.harvard.edu/science/core-...
http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/FI...
https://xwiki.nbirn.org:8443/bin/view/Fu...
Thanks in advance,
Matthias
Our lab is developing a quality assurance pipeline to automatically check and generate reports of the acquired MR data. My task is to find out, what is already done at different sites and labs.
I would be glad if you like to share your QA steps (and maybe scripts). If you're interested in my inquiry results, I also would share them here.
I'm especially interested, if you're using fsl tools for that.
We are currently working on motion QA using python for website-embedded video reports (and already using mcflirt parameters), normalization results with reports of different slices and their overlap with MNI template.
In automated artefact detection, we're at the moment focussing on ghosting, spikes and zipper artefacts.
For quantified checks, we're propably will be using tSNR for EPI sequences, SNR for structural images. Since I'm just a beginner in the neuroscience field and the project has been started, thats all I can tell you right now :)
What I've already found:
http://cbs.fas.harvard.edu/science/core-...
http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/FI...
https://xwiki.nbirn.org:8443/bin/view/Fu...
Thanks in advance,
Matthias
Feb 12, 2014 08:02 PM | Christian Haselgrove
RE: Quality assurance at your site
Matthias,
The INCF XNAT (its one-click server: http://xnat.incf.org/) uses the BIRN QA (http://www.nitrc.org/projects/bxh_xcede_...) for time series data, DTIPrep (https://www.nitrc.org/projects/dtiprep/) for diffusion data, and a custom approach for structural data that uses FSL FAST's tissue classification to estimate signal (mean intensity in the brain) and noise (standard deviation of intensity outside the head) to calculate SNR. Code for the latter can be found from line 155 at https://github.com/INCF/one_click/blob/m...
You might also want to check out Artifact Detection Tools (http://www.nitrc.org/projects/artifact_d...).
c
The INCF XNAT (its one-click server: http://xnat.incf.org/) uses the BIRN QA (http://www.nitrc.org/projects/bxh_xcede_...) for time series data, DTIPrep (https://www.nitrc.org/projects/dtiprep/) for diffusion data, and a custom approach for structural data that uses FSL FAST's tissue classification to estimate signal (mean intensity in the brain) and noise (standard deviation of intensity outside the head) to calculate SNR. Code for the latter can be found from line 155 at https://github.com/INCF/one_click/blob/m...
You might also want to check out Artifact Detection Tools (http://www.nitrc.org/projects/artifact_d...).
c