Dear Tim, Thank you for your efforts and kind response. Yes, the problem was with subject directory name which is resolved now and the scripts are running without error however I'm still trying to find and tweak parameters that help me to get better results. Most probably I will ask you and others more about it in near future.
Best, Mahmoud On Tue, Dec 29, 2015 at 1:49 PM, Timothy B. Brown <tbbr...@wustl.edu> wrote: > Hi Mahmoud, > > Sorry for the delay in answering your question below. I was away for the > holidays for a while. > > Have you already resolved the issue you describe (the "no output filename > specified" error in the FreeSurferPipeline)? I noticed a subsequent > message from you to the hcp-users list in which you state: > > I've applied HCP pipeline scripts to our structural data and it looks like > everything goes well to the end of PostFreeSurferPipelline. > > That being the case, does that mean that you've already resolved the > problem with the "Error: no output filename specified!" in the > FreeSurferPipeline? > > If you have, you can likely ignore the remainder of this message. > > ------------------------- > > I have taken a look at the log files that you sent, and it appears to me > that the FreeSurferHiresPial.sh script (which is invoked as part of the > FreeSurferPipeline) is successfully running through line 52 in the current > latest version of the FreeSurferHiresPial.sh script. That line looks like: > > 52: mri_surf2surf --s $SubjectID --sval-xyz pial.T2 --reg $regII > $mridir/orig.mgz --tval-xyz --tval pial --hemi rh > > After that line, there are statements that extract some information from > the brain.finalsurfs.mgz file and create the > <studydir>/<subjectid>/T1w/<subjectid>/mri/c_ras.mat matrix file. > > 54: MatrixX=`mri_info $mridir/brain.finalsurfs.mgz | grep "c_r" | cut -d > "=" -f 5 | sed s/" "/""/g` > 55: MatrixY=`mri_info $mridir/brain.finalsurfs.mgz | grep "c_a" | cut -d > "=" -f 5 | sed s/" "/""/g` > 56: MatrixZ=`mri_info $mridir/brain.finalsurfs.mgz | grep "c_s" | cut -d > "=" -f 5 | sed s/" "/""/g` > 57: echo "1 0 0 ""$MatrixX" > $mridir/c_ras.mat > 58: echo "0 1 0 ""$MatrixY" >> $mridir/c_ras.mat > 59: echo "0 0 1 ""$MatrixZ" >> $mridir/c_ras.mat > 60: echo "0 0 0 1" >> $mridir/c_ras.mat > > Next is a call to mris_convert on line 62 of the FreeSurferHiresPial.sh > script. > > 62: mris_convert "$surfdir"/lh.white "$surfdir"/lh.white.surf.gii > > I don't seem to find any evidence in your log files that the call to > mris_convert is succeeding. So your run seems to be ending with the "Error: > no output filename specified!" message somewhere between lines 54 and 62. > > It seems a bit unlikely that the problem is in lines 54 through 60, but > just to check, can you confirm for me that the file > <studydir>/<subjectid>/T1w/<subjectid>/mri/c_ras.mat exists? In the case > shown in your logs, I believe this file should actually be named: > /home/mydata/HCPReady/111_Sigma7/T1w/111_Sigma7/mir/c_ras.mat. If it > exists, is it a 4 line text file that looks something like the following? > > 1 0 0 -1.0000 > 0 1 0 -17.5000 > 0 0 1 19.0000 > 0 0 0 1 > > The last numeric values on each of the first three lines will probably > differ from the above. > > Assuming that the c_ras.mat file is successfully created, then I would > note that line 62 is the first use of the $surfdir variable in the > FreeSurferHiresPial.sh script (after setting its value). In the case > shown in your log files, the $surfdir variable should have a value of > /home/mydata/HCPReady/111_Signma7/T1w/111_Sigma7/surf. > > Could you please verify that such a directory exists and is writable? > > Thanks, > > Tim > > On Tue, Dec 22, 2015, at 07:06, Mahmoud wrote: > > Hello Tim and others, > > Following the previous posts, I renamed my dataset directory and ran the > PreFreeSurferPipeline on it and it looks like finished successfully. > After running the FreeSurferPipeline I got the same error "Error: no > output filename specified!" . I've attached the FreeSurferPipeline output > and error log here for you reference. > > I appreciate your help and time. > > Thank you, > Mahmoud > > On Fri, Dec 18, 2015 at 11:39 AM, Mahmoud <zeydab...@gmail.com> wrote: > > Dear Tim, > > I got it. Thank you for detailed explanation. > > Best, > Mahmoud > > On Fri, Dec 18, 2015 at 11:29 AM, Timothy B. Brown <tbbr...@wustl.edu> > wrote: > > Hi Mahmoud, > > Short answer: > > Because of directory structure conventions and file naming conventions > used in the pipeline scripts, you cannot rename the subject directory and > expect that the pipelines will run without (at least) also renaming a > number of other files in the subject directory. So renaming the subject > directory will break quite a number of things. > > Long answer: > > The FreeSurferPipeline in particular and the other pipeline scripts in > general depend upon a directory structure for the data that has a root > *study* folder in which each subdirectory of that *study* folder is named > for the subject id. So if the data you are processing are in a directory > named my_study, and you had only two subjects with subject ids xyz and abc, > then you would need to have xyz and abc subdirectories in your my_study > directory. > > Seems like you've got that all set up correctly initially. > > But...the pipeline scripts also depend upon file naming conventions that > use the subject id in the file names. For example, if the subject > subdirectory is named xyz, then a number of files in that directory tree > will be expected to be named starting with xyz. For example, there is an > expectation that there will be a file named my_study/xyz > /unprocessed/3T/T1_MPR1/xyz_3T_T1w_MPR1.nii.gz. Notice that the two > places the subject id is used have to match. Similarly, many of the > pipelines will generate output files that are used by subsequent pipelines > using a naming convention in which the names of the generated files will > start with the subject id. > > So, I suspect that this is the problem you are encountering. When you run > FreeSurferPipeLineBatch and tell it to process subject xyz_1, the > FreeSurferPipeline script is expecting to find a number of input files with > names that start with xyz_1. But they're not there because your xyz_1 > directory is just a copy of the original xyz directory. So many of the > files in it start with simply xyz. > > Hope that helps, > > Tim > > On Fri, Dec 18, 2015, at 07:55, Mahmoud wrote: > > Dear experts, > > After running the PreFreeSurferPipelineBatch for subjectID = xyz I changed > the ID (i.e. renamed the subject directory name) to xyz_1 and ran the > FreeSurferPipeLineBatch which ended up with this error in the error log > file (I just copied the few last lines): > > measuring cortical thickness... > writing cortical thickness estimate to 'thickness' file. > positioning took 5.6 minutes > Error: no output filename specified! > > Is this because of that name change? if so, is there any solution to it ? > > Thank you! > Mahmoud > > _______________________________________________ > HCP-Users mailing list > HCP-Users@humanconnectome.org > http://lists.humanconnectome.org/mailman/listinfo/hcp-users > > -- > Timothy B. Brown > Business & Technology Application Analyst III > Pipeline Developer (Human Connectome Project) > tbbrown(at)wustl.edu > ________________________________________ > The material in this message is private and may contain Protected > Healthcare Information (PHI). > If you are not the intended recipient, be advised that any unauthorized > use, disclosure, copying > or the taking of any action in reliance on the contents of this > information is strictly prohibited. > If you have received this email in error, please immediately notify the > sender via telephone or > return mail. > > Email had 2 attachments: > > - FreeSurferPipeline.sh.e1326051 > 25k (application/octet-stream) > - FreeSurferPipeline.sh.o1326051 > 465k (application/octet-stream) > > -- > Timothy B. Brown > Business & Technology Application Analyst III > Pipeline Developer (Human Connectome Project) > tbbrown(at)wustl.edu > ________________________________________ > The material in this message is private and may contain Protected > Healthcare Information (PHI). > If you are not the intended recipient, be advised that any unauthorized > use, disclosure, copying > or the taking of any action in reliance on the contents of this > information is strictly prohibited. > If you have received this email in error, please immediately notify the > sender via telephone or > return mail. > _______________________________________________ HCP-Users mailing list HCP-Users@humanconnectome.org http://lists.humanconnectome.org/mailman/listinfo/hcp-users
