Experts,
I have been working on Freesurfer last 5 months and initially the recon processes were within normal time limits.
However, I am working on a different set of subjects now and its consuming days together. For my current data,autorecon1 is fine, it stops at a particular point in autorecon2 for infinity. Any suggestions would be helpful.
I am pasting the autorecon2 process here:
recon-all -autorecon2 -subjid ja Subject Stamp: freesurfer-Linux-centos4-stable-pub-v4.1.0
Current Stamp: freesurfer-Linux-centos4-stable-pub-v4.1.0
INFO: SUBJECTS_DIR is /usr/local/freesurfer/subjects
Actual FREESURFER_HOME /usr/local/freesurfer
-rw-rw-r-- 1 barnali barnali 257784 Jul 9 10:32 /usr/local/freesurfer/subjects/ja/scripts/recon-all.log
Linux barnali-work 2.6.24-16-server #1 SMP Thu Apr 10 13:58:00 UTC 2008 i686 GNU/Linux
#-------------------------------------
#@# EM Registration Thu Jul 9 10:35:21 PDT 2009
/usr/local/freesurfer/subjects/ja/mri
mri_em_register -mask brainmask.mgz nu.mgz /usr/local/freesurfer/average/RB_all_2008-03-26.gca transforms/talairach.lta
using MR volume brainmask.mgz to mask input volume...
reading 1 input volumes...
logging results to talairach.log
reading '/usr/local/freesurfer/average/RB_all_2008-03-26.gca'...
average std = 6.9 using min determinant for regularization = 4.7
0 singular and 1812 ill-conditioned covariance matrices regularized
reading 'nu.mgz'...
freeing gibbs priors...done.
bounding unknown intensity as < 14.9 or > 790.2
total sample mean = 84.0 (478 zeros)
************************************************
spacing=8, using 2185 sample points, tol=1.00e-05...
************************************************
register_mri: find_optimal_transform
find_optimal_transform: nsamples 2185, passno 0, spacing 8
resetting wm mean[0]: 102 --> 107
resetting gm mean[0]: 64 --> 64
input volume #1 is the most T1-like
using real data threshold=8.0
using (107, 91, 112) as brain centroid...
mean wm in atlas = 107, using box (90,77,91) --> (123, 104,132) to find MRI wm
before smoothing, mri peak at 137
after smoothing, mri peak at 151, scaling input intensities by 0.709
scaling channel 0 by 0.708609
initial log_p = -92132.8
************************************************
First Search limited to translation only.
************************************************
Found translation: (-2.8, -1.7, 1.7): log p = -39503.6
****************************************
Nine parameter search. iteration 0 nscales = 0 ...
****************************************
Result so far: scale 1.000: max_log_p=-37619.9, old_max_log_p =-39503.6 (thresh=-39464.1)
1.062 0.000 0.000 -10.944;
0.000 1.053 0.139 -22.515;
0.000 -0.131 0.991 18.328;
0.000 0.000 0.000 1.000;
****************************************
Nine parameter search. iteration 1 nscales = 0 ...
****************************************
Result so far: scale 1.000: max_log_p=-35219.5, old_max_log_p =-37619.9 (thresh=-37582.3)
1.062 0.000 0.000 -10.944;
0.000 1.266 0.027 -29.117;
0.000 0.025 1.003 -1.688;
0.000 0.000 0.000 1.000;
****************************************
Nine parameter search. iteration 2 nscales = 0 ...
****************************************
Result so far: scale 1.000: max_log_p=-35219.5, old_max_log_p =-35219.5 (thresh=-35184.3)
1.062 0.000 0.000 -10.944;
0.000 1.266 0.027 -29.117;
0.000 0.025 1.003 -1.688;
0.000 0.000 0.000 1.000;
reducing scale to 0.2500
****************************************
Nine parameter search. iteration 3 nscales = 1 ...
****************************************
Result so far: scale 0.250: max_log_p=-32338.2, old_max_log_p =-35219.5 (thresh=-35184.3)
1.096 0.000 0.000 -15.249;
0.000 1.208 0.026 -25.944;
0.000 0.026 1.035 -8.782;
0.000 0.000 0.000 1.000;
****************************************
Nine parameter search. iteration 4 nscales = 1 ...
****************************************
Result so far: scale 0.250: max_log_p=-31970.6, old_max_log_p =-32338.2 (thresh=-32305.8)
1.096 0.000 0.000 -17.124;
0.000 1.189 0.025 -21.806;
0.000 0.026 1.035 -6.907;
0.000 0.000 0.000 1.000;
****************************************
Nine parameter search. iteration 5 nscales = 1 ...
****************************************
Result so far: scale 0.250: max_log_p=-31970.6, old_max_log_p =-31970.6 (thresh=-31938.6)
1.096 0.000 0.000 -17.124;
0.000 1.189 0.025 -21.806;
0.000 0.026 1.035 -6.907;
0.000 0.000 0.000 1.000;
reducing scale to 0.0625
****************************************
Nine parameter search. iteration 6 nscales = 2 ...
****************************************
Result so far: scale 0.062: max_log_p=-31028.4, old_max_log_p =-31970.6 (thresh=-31938.6)
1.083 0.000 0.000 -14.039;
0.000 1.194 0.025 -22.828;
0.000 0.026 1.039 -7.345;
0.000 0.000 0.000 1.000;
****************************************
Nine parameter search. iteration 7 nscales = 2 ...
****************************************
Result so far: scale 0.062: max_log_p=-30899.3, old_max_log_p =-31028.4 (thresh=-30997.4)
1.079 0.000 0.000 -13.957;
0.000 1.194 0.025 -22.828;
0.000 0.026 1.043 -7.785;
0.000 0.000 0.000 1.000;
****************************************
Nine parameter search. iteration 8 nscales = 2 ...
****************************************
Result so far: scale 0.062: max_log_p=-30683.0, old_max_log_p =-30899.3 (thresh=-30868.4)
1.070 0.000 0.000 -12.858;
0.000 1.194 0.025 -22.828;
0.000 0.026 1.039 -7.344;
0.000 0.000 0.000 1.000;
****************************************
Nine parameter search. iteration 9 nscales = 2 ...
****************************************
Result so far: scale 0.062: max_log_p=-30108.0, old_max_log_p =-30683.0 (thresh=-30652.3)
1.058 0.000 0.000 -11.225;
0.000 1.194 0.025 -22.828;
0.000 0.026 1.051 -9.140;
0.000 0.000 0.000 1.000;
****************************************
Nine parameter search. iteration 10 nscales = 2 ...
****************************************
Result so far: scale 0.062: max_log_p=-29914.1, old_max_log_p =-30108.0 (thresh=-30077.9)
1.058 0.000 0.000 -11.225;
0.000 1.180 0.025 -21.172;
0.000 0.027 1.059 -10.034;
0.000 0.000 0.000 1.000;
****************************************
Nine parameter search. iteration 11 nscales = 2 ...
****************************************
Result so far: scale 0.062: max_log_p=-29914.1, old_max_log_p =-29914.1 (thresh=-29884.1)
1.058 0.000 0.000 -11.225;
0.000 1.180 0.025 -21.172;
0.000 0.027 1.059 -10.034;
0.000 0.000 0.000 1.000;
min search scale 0.025000 reached
***********************************************
Computing MAP estimate using 2185 samples...
***********************************************
dt = 5.00e-06, momentum=0.80, tol=1.00e-05
l_intensity = 1.0000
Aligning input volume to GCA...
Transform matrix
1.05773 0.00000 0.00000 -11.22528;
0.00000 1.17956 0.02501 -21.17203;
0.00000 0.02668 1.05912 -10.03400;
0.00000 0.00000 0.00000 1.00000;
nsamples 2185
Quasinewton: input matrix
1.05773 0.00000 0.00000 -11.22528;
0.00000 1.17956 0.02501 -21.17203;
0.00000 0.02668 1.05912 -10.03400;
0.00000 0.00000 0.00000 1.00000;
v3p/netlib/opt/lbfgs.c: lb3_1.lp > 0
outof QuasiNewtonEMA: 012: -log(p) = 29914.1 tol 0.000010
Resulting transform:
1.058 0.000 0.000 -11.225;
0.000 1.180 0.025 -21.172;
0.000 0.027 1.059 -10.034;
0.000 0.000 0.000 1.000;
pass 1, spacing 8: log(p) = -29914.1 (old=-92132.8)
transform before final EM align:
1.058 0.000 0.000 -11.225;
0.000 1.180 0.025 -21.172;
0.000 0.027 1.059 -10.034;
0.000 0.000 0.000 1.000;
**************************************************
EM alignment process ...
Computing final MAP estimate using 244171 samples.
**************************************************
dt = 5.00e-06, momentum=0.80, tol=1.00e-07
l_intensity = 1.0000
Aligning input volume to GCA...
Transform matrix
1.05773 0.00000 0.00000 -11.22528;
0.00000 1.17956 0.02501 -21.17203;
0.00000 0.02668 1.05912 -10.03400;
0.00000 0.00000 0.00000 1.00000;
nsamples 244171
Quasinewton: input matrix
1.05773 0.00000 0.00000 -11.22528;
0.00000 1.17956 0.02501 -21.17203;
0.00000 0.02668 1.05912 -10.03400;
0.00000 0.00000 0.00000 1.00000;
dfp_em_step_func: 011: -log(p) = 3772889.8
after pass:transform: ( 1.07, 0.01, 0.01, -11.23)
( 0.00, 1.18, 0.03, -21.17)
( 0.00, 0.03, 1.06, -10.03)
v3p/netlib/opt/lbfgs.c: lb3_1.lp > 0
pass 2 through quasi-newton minimization...
v3p/netlib/opt/lbfgs.c: lb3_1.lp > 0
outof QuasiNewtonEMA: 013: -log(p) = 3772889.8 tol 0.000000
final transform:
1.067 0.008 0.009 -11.225;
0.001 1.180 0.027 -21.172;
0.001 0.027 1.059 -10.034;
0.000 0.000 0.000 1.000;
writing output transformation to transforms/talairach.lta...
registration took 33 minutes and 19 seconds.
#--------------------------------------
#@# CA Normalize Thu Jul 9 11:08:40 PDT 2009
/usr/local/freesurfer/subjects/ja/mri
mri_ca_normalize -mask brainmask.mgz nu.mgz /usr/local/freesurfer/average/RB_all_2008-03-26.gca transforms/talairach.lta norm.mgz
using MR volume brainmask.mgz to mask input volume...
reading 1 input volumes
reading atlas from '/usr/local/freesurfer/average/RB_all_2008-03-26.gca'...
reading transform from 'transforms/talairach.lta'...
reading input volume from nu.mgz...
resetting wm mean[0]: 102 --> 107
resetting gm mean[0]: 64 --> 64
input volume #1 is the most T1-like
using real data threshold=8.0
using (107, 91, 112) as brain centroid...
mean wm in atlas = 107, using box (90,77,91) --> (123, 104,132) to find MRI wm
before smoothing, mri peak at 137
after smoothing, mri peak at 151, scaling input intensities by 0.709
scaling channel 0 by 0.708609
using 244171 sample points...
INFO: compute sample coordinates transform
1.067 0.008 0.009 -11.225;
0.001 1.180 0.027 -21.172;
0.001 0.027 1.059 -10.034;
0.000 0.000 0.000 1.000;
INFO: transform used
finding control points in Left_Cerebral_White_Matter....
found 41584 control points for structure...
bounding box (125, 59, 30) --> (189, 159, 190)
It stops here and doesnt process any further. I tried twice, with same result.
I work on a dual core Ubuntu 5GB machine
Thanks
Barnali