[MINC-development] accuracy in mritotal. better tool?
Yasunari Tosa
minc-development@bic.mni.mcgill.ca
05 Aug 2003 17:20:29 -0400
--=-MTMNSz2vzKXae07QG71B
Content-Type: text/plain
Content-Transfer-Encoding: 7bit
Hi, MINC people:
I'm not getting the accuracy in mritotal results.
I would appreciate it very much for your help in resolving the accuracy.
mritotal -verbose -debug -clobber -protocol icbm orig.mnc talairach.xfm
My test was to modify the world translation by a fixed amount and see
how the result changes
(therefore the image never changes).
What I was surprized was the following:
Even though the rotation and the scaling parts are quite stable, the
translation part error
is quite large (up to 1.48 mm).
The original volume is 256^3 with 1 mm voxel.
The last two column is the difference and the error from the original in
translation
The fourth column is the translation part.
Original : world_translation is (0,0,0):
=============================
1.10129165649414 0.0375710912048817 -0.0232741367071867 0.884597778320312
-0.0363568961620331 1.05287253856659 -0.020708842203021 -6.91514205932617
0.0234725382179022 0.0233993344008923 1.14845287799835 8.22117519378662
world_translation is (10, 0, 0)
======================= diff error
1.10129165649414 0.0375710912048817 -0.0232741367071867 -10.1283187866211 11.0129165649414 10 1.01291656494141
-0.0363568961620331 1.05287253856659 -0.020708842203021 -6.55157375335693 -0.36356830596924 0 -0.36356830596924
0.0234725382179022 0.0233993344008923 1.14845287799835 7.98645114898682 0.234724044799799 0 0.234724044799799
world_translation is (0,10,0)
====================== diff error
1.10129177570343 0.0375712215900421 -0.0232741497457027 0.508886873722076 0.375710904598236 0 0.375710904598236
-0.0363569855690002 1.05287170410156 -0.0207088124006987 -17.4438610076904 10.5287189483642 10 0.528718948364231
0.0234725456684828 0.0233993232250214 1.14845287799835 7.98718357086182 0.233991622924799 0 0.233991622924799
world_translation is (0, 0, 10)
======================= diff error
1.10129165649414 0.0375710912048817 -0.0232741367071867 1.11733937263489 -0.232741594314578 0 -0.232741594314578
-0.0363568961620331 1.05287253856659 -0.020708842203021 -6.70805358886719 -0.20708847045898 0 -0.20708847045898
0.0234725382179022 0.0233993344008923 1.14845287799835 -3.26335191726685 11.4845271110535 10 1.48452711105347
--
Yasunari Tosa, Ph.D.
R&D Enginner email: tosa@nmr.mgh.harvard.edu
Mass. General Hospital NMR Center phone:
Building 149-2301
13th Street fax :
Charlestown, MA 02129
--=-MTMNSz2vzKXae07QG71B
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: 7bit
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 TRANSITIONAL//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; CHARSET=UTF-8">
<META NAME="GENERATOR" CONTENT="GtkHTML/3.0.7">
</HEAD>
<BODY>
Hi, MINC people:<BR>
<BR>
I'm not getting the accuracy in mritotal results. <BR>
I would appreciate it very much for your help in resolving the accuracy.<BR>
<BR>
mritotal -verbose -debug -clobber -protocol icbm orig.mnc talairach.xfm<BR>
<BR>
My test was to modify the world translation by a fixed amount and see how the result changes<BR>
(therefore the image never changes).<BR>
<BR>
What I was surprized was the following:<BR>
<BR>
Even though the rotation and the scaling parts are quite stable, the translation part error<BR>
is quite large (up to 1.48 mm). <BR>
<BR>
The original volume is 256^3 with 1 mm voxel.<BR>
<BR>
The last two column is the difference and the error from the original in translation<BR>
The fourth column is the translation part.<BR>
<BR>
Original : world_translation is (0,0,0):<BR>
=============================<BR>
1.10129165649414 0.0375710912048817 -0.0232741367071867 0.884597778320312 <BR>
-0.0363568961620331 1.05287253856659 -0.020708842203021 -6.91514205932617 <BR>
0.0234725382179022 0.0233993344008923 1.14845287799835 8.22117519378662 <BR>
<BR>
world_translation is (10, 0, 0)<BR>
======================= diff error<BR>
1.10129165649414 0.0375710912048817 -0.0232741367071867 -10.1283187866211 11.0129165649414 10 1.01291656494141<BR>
-0.0363568961620331 1.05287253856659 -0.020708842203021 -6.55157375335693 -0.36356830596924 0 -0.36356830596924<BR>
0.0234725382179022 0.0233993344008923 1.14845287799835 7.98645114898682 0.234724044799799 0 0.234724044799799<BR>
<BR>
world_translation is (0,10,0)<BR>
====================== diff error <BR>
1.10129177570343 0.0375712215900421 -0.0232741497457027 0.508886873722076 0.375710904598236 0 0.375710904598236<BR>
-0.0363569855690002 1.05287170410156 -0.0207088124006987 -17.4438610076904 10.5287189483642 10 0.528718948364231<BR>
0.0234725456684828 0.0233993232250214 1.14845287799835 7.98718357086182 0.233991622924799 0 0.233991622924799<BR>
<BR>
world_translation is (0, 0, 10)<BR>
======================= diff error<BR>
1.10129165649414 0.0375710912048817 -0.0232741367071867 1.11733937263489 -0.232741594314578 0 -0.232741594314578<BR>
-0.0363568961620331 1.05287253856659 -0.020708842203021 -6.70805358886719 -0.20708847045898 0 -0.20708847045898<BR>
0.0234725382179022 0.0233993344008923 1.14845287799835 -3.26335191726685 11.4845271110535 10 1.48452711105347<BR>
<BR>
<TABLE CELLSPACING="0" CELLPADDING="0" WIDTH="100%">
<TR>
<TD>
<PRE>--
Yasunari Tosa, Ph.D.
R&D Enginner email: tosa@nmr.mgh.harvard.edu
Mass. General Hospital NMR Center phone:
Building 149-2301
13th Street fax :
Charlestown, MA 02129
</PRE>
</TD>
</TR>
</TABLE>
</BODY>
</HTML>
--=-MTMNSz2vzKXae07QG71B--