Page 1 of 1
AimsVoiStat Linux version
Posted: Fri May 20, 2005 4:32 pm
by bottlaender
Warning to users :
The content of the output file has changed in the last version.
The post-processing when using this file may be now wrong and have to be adapted.
To developpers : can you please inform the users for these modifications that alter the output files.
Posted: Fri May 20, 2005 5:44 pm
by Yann Cointepas
A lot of people are doing post-processing on the result of this command. I agree that modification of the output should be done more carefully.
StdDev for AimsVoiStat
Posted: Thu Jun 16, 2005 3:37 pm
by jdubois
The content of the output file has changed for "StdDev". What is calculated now? Is it the standard deviation over the VOI? What was calculated in the previous versions? The difference is huge between the results... Thank you for your reply!
Jessica
Posted: Tue Jun 21, 2005 10:10 pm
by riviere
Well, apparently nobody wants to answer you
The last significant modif to the VoiStat stuff appears to be:
Change 13797 by maroy@maroy-shfj on 2005/01/19 15:49:54
Major bug fixed in Roi Stat, concerning standard deviation of signal in each region.
Affected files ...
... //depot/aimsalgo/main/src/aimsalgo/roi/roistat.h#11 edit
Differences ...
==== //depot/aimsalgo/main/src/aimsalgo/roi/roistat.h#11 (text) ====
205c205
< std=(float)sqrt( (double)(std-square(mean)*pn2) )/(float)(pn2-1);
---
> std=(float)sqrt( (double)(std-square(mean)*pn2) / (float)(pn2-1) );
This seems to suggest the stdDev was wrong before (in all your earlier studies!)... but I don't know more about it.
Denis
Posted: Mon Aug 14, 2006 7:49 pm
by Tupo
It is really??