Hi,
Yes, I'm afraid this is a bug in avwstats++ - sorry about that.
It only occurs sometimes because it depends on where the maximum
value is in the order that the volume is traversed internally.
I've already fixed it in the code and the next patch will include this
fix, but in the meantime the best workaround is:
avwmaths++ 3209525_FA -mas dif-3209525LeftInf 3209525_FA_masked
avwstats++ 3209525_FA_masked -R
and only take the *maximum* value from this. The minimum and
means are already correct with the previous call to avwstats++ and
so you can use them still. It will just be necessary to do the lines above
to get the maximum (and make sure you do not use the -k option in
the second line as that is where the problem lies).
All the best,
Mark
Wayne Su wrote:
>Hi,
>
>I tried to use avwstats++ to get min/max FA values from seeds. The seeds
>only have 5 pixels. If it is seen from fslview, the FA values of these
>pixels are 0.232322 0.319473 0.486368, 0.627936 0.706525. When I use
>following command,
>
>Avwstats++ 3209525_FA -k dif-3209525LeftInf -R
>
>Here 3209525_FA is FA image, dif-3209525LeftInf is the seed image.
>
>I got (mean, min, max) FA values: 0.474525 0.232322 0.000000.
>
> The max value is wrong.
>
>This happens some time. There are about 8 cases wrong out of total 80 cases.
>
>I tried on Linux and Mac OS X. it came same result.
>
>Any solution?
>
>Regard,
>Wayne
>
>
|