Hello,
I looked at this a while ago and think that I found the respective piece of code that generates the overall rating. Something like
% == snip ==
temp = spm_select(Inf,'any', 'Select cat12 quality indicator files');
all = cell(size(temp,1),2);
for i = 1:size(temp,1)
curr = load(temp(i,:));
all{i,1} = deblank(temp(i,:));
all{i,2} = min(100,max(0,105 - curr.S.qualityratings.IQR*10));
end;
% == snap ==
may do the trick. If you use subdir, one of my all time favorite functions from the Mathworks file exchange, it will be even easier to assess all files, or all of one group.
[log in to unmask]" class="GroupWiseMessageBody">
Cheers
Marko
PS: Disclaimer: could not verify if this still works as I am not at my desk so you may want to double-check...
Dear all,
I am analysing data from a multicentric study. We have 5 different groups, each one acquired partially at centre 1 and partially at centre 2 (the number of included subjects is not balanced across centres). The first thing that I want to know is if there is
a systematic difference in data quality between the two centres and, more importantly, if there is an interaction between group and centre on data quality. To do this on the T1 images I have preprocessed everything with CAT, and now I am trying to write a
routine to extract the quality parameter from the mat file of the reports in order to run statistical analyses on them.
When I open the .mat file of the report I see that there are at least two field that could contain information of interest to my aim, namely the ones called quality measures and quality ratings. Within these two field, there are several other subfields, whose
naming is not (at lest for me) self-explanatory. In particular, within the quality measures field I see the following subfields
res.vx.vol
res.vx.voli
res.RMS
res.BB
tissue.mn
tissue.mnr
tissue.weighting
tissue.std
tissue.stdr
contrast
contrastr
NCR
ICR
Similarly, within the quality ratings field, I find the following subfields
res.vx.vol
res.RMS
res.BB
tissue.mn
tissue.std
contrast
contrastr
NCR
ICR
IQR
Does someone know what these fields stand for ? Can someone give a hint about which one(s) of these measures should I extract to check for systematic differences in data quality between the two centres ?
Thank in advance for any help
Best
Alain
