Stabilitycalc is part of a suite of python tools we use to monitor the
stability of our fMRI system. We use a version of the fBIRN procedure
to run daily phantom quality assurance tests on our research MR system.
These phantom scans are then converted to NIFTI format and analyzed
with stabilitycalc to look for system instabilities that could degrade
our fMRI data.
The stabilitycalc package generates reports for individual scans, and
tabulates performance over time, generating summary reports that put
current performance into the context of normal operation. We use this
to quickly identify scanner problems, and the individual reports give
a good indication of what sort of problem we are having, which in turn
helps us troubleshoot much more effectively.
The current release includes the stabilitycalc program itself, which
generates reports for individual scans. The goal for the final release
is to include all of the tools that generate summary data. Since this
software sort of accreted over time (and started out partially as a way
for me to learn python), there is a fair amount of redesign that needs
to be done to make the package more reliable and generalizable to other
scanner manufacturers and workflows. Another important part of this
development effort is to remove some rather ghastly kludges in the
code. I released the code in large measure to shame myself into
improving it (and, if possible, provide a service for anybody else
trying to set up a QA procedure).
One of the design goals of the set of tools is to use pure python to
minimize machine architecture and compilation headaches, and to avoid
costly proprietary environments such as matlab.
Inputs:
Output:
stabilitycalc produces an HTML report:
STABILITYDIR/procresults/output.html