Anand Joshi

Forum Replies Created

Viewing 15 posts - 31 through 45 (of 172 total)
  • Author
    Posts
  • in reply to: measuring hippocampus volume children #1884
    Anand Joshi
    Moderator

    Currently, we don’t have a children-specific atlas, however, the results generated by the default atlases should be reasonable for 4yr+ children.

    in reply to: Unable to access SVReg labeled volume #1883
    Anand Joshi
    Moderator

    Hi, The labeled volume should be automatically stored in the same directory as the MRI scan. They have filenames <subjectid>.svreg.label.nii.gz
    Are you able to find that file?
    Here is the list of output files
    http://brainsuite.org/processing/svreg/output/

    Please let me know if you find these files generated in your mri directory.

    in reply to: Hardware recommendations #1879
    Anand Joshi
    Moderator

    Hi Tod,
    SVReg definitely makes use of available cores. So for surface registration, it processes 2 cortical hemispheres on 2 separate threads. During volume registration, whenever possible, it processes 3 coordinates on 3 threads.
    Moreover, each of the separate threads use multiple CPUs during various operations such as conjugate gradient minimization and interpolations.

    So yes, the more cores, the better it is for SVReg processing.

    Additionally, you can also process multiple subjects simultaneously if you have computing power.

    The system you mention is definitely great (both).

    in reply to: svreg command line "cores" #1844
    Anand Joshi
    Moderator

    Can you just type path in the matlab command window?

    How are you running the svreg execuable from witin matlab? Are you using !svreg.exe …. ?

    Can you send me steps to reproduce the issue?

    in reply to: svreg command line "cores" #1842
    Anand Joshi
    Moderator

    It might be that it is trying to use a different runtime version when called from matlab.
    Note that the environment variables and the path is different in dos and in matlab.
    check environment variables in matlab by typing env. Make sure that the correct MCR is first on the path.

    in reply to: svreg command line "cores" #1840
    Anand Joshi
    Moderator

    Did you install 32 bit windows MCR by any chance? A 64 bit version is required.

    in reply to: svreg command line "cores" #1837
    Anand Joshi
    Moderator

    Thank you for using BrainSuite and reporting this issue.
    You can run the svreg on the command line in a single threaded mode using -U flag.

    http://brainsuite.org/processing/svreg/usage/

    in reply to: Brainsuite vs Freesurfer #1832
    Anand Joshi
    Moderator

    One of the unique features of BrainSuite is that the volumetric registration used is ‘surface constrained volumetric registration’. This means that during the registration process, cortical surface and subcortical structures are aligned simultaneously leading to a consistent registration.
    Another unique feature is that the cortical thickness analysis used in BrainSuite uses computation of cortical thickness using partial tissue fraction information.
    For the subcortex, the total volumes of various structures, as well as gray matter and white matters, are reported in detail by BrainSuite. I am not aware of any other functionality that other software has and is missing in BrainSuite for structural analysis.

    • This reply was modified 4 years, 10 months ago by Anand Joshi.
    in reply to: About Tractography options #1830
    Anand Joshi
    Moderator

    Hi Fatima,
    Please take a look at BrainSuite diffusion pipeline:
    http://brainsuite.org/processing/diffusion/

    Here is a step by step tutorial on how to use it.
    http://brainsuite.org/quickstart/dti/

    in reply to: Brainsuite vs Freesurfer #1827
    Anand Joshi
    Moderator

    Hi Leon, Thank you for using BrainSuite and I am glad that your study was positively received.
    I would answer by saying that BrainSuite release is relatively frequent (approximately every 6 months). The latest release has been Feb 19, this year. I think this is more frequent than other competing software. I doubt it can be called outdated due to that reason. Additionally, the interactive GUI and fast processing speed allow rapid experimentation and visualization.
    Unless there is some feature that is needed for the study, that is not in BrainSuite, it is unclear why a switch is needed, especially since the results of the analysis seem to be not as expected. But the reviewers seem to like them.

    Usually, most researchers and different labs have their favorite software and pipelines. The user base of BrainSuite is quite large and is growing.

    Additionally, we as developers and maintainers of the software are ready to support and enhance the software.

    • This reply was modified 4 years, 10 months ago by Anand Joshi.
    • This reply was modified 4 years, 10 months ago by Anand Joshi.
    in reply to: interruption of sequence #1809
    Anand Joshi
    Moderator

    SVReg is a bit memory intensive, so if you running multiple subjects simultaneously, then it might fail due to memory issue. Also please make sure that most all the other program windows are closed.
    Please make sure that you are using the latest version (19a). I can’t think of any other reason.

    in reply to: Brain Total Volume . #1807
    Anand Joshi
    Moderator

    You get subject.roiwise.stats.txt after running svreg.
    This has a detailed report of volumes of different tissues.

    in reply to: svreg warp #1805
    Anand Joshi
    Moderator

    Discussed this by email with you but posting here so that it is useful to others.
    Yes, the program expects .gz extension. So if you compress it back the program should work.

    in reply to: Cortical thickness in hippocampus #1804
    Anand Joshi
    Moderator

    Thank you for using BrainSuite and very happy to hear that you like it.
    The SVReg module computes thickness only for the cortical surface using inner and pial cortical surface representations.
    Since the hippocampus and amygdala are subcortical structures, their volumes are not computed, but gray and white matter volumes are reported. The ‘thickness’ computation for hippocampus will involve generating surfaces only for that structure and not the entire brain. We don’t have that functionality implemented.

    in reply to: Matlab MCR #1556
    Anand Joshi
    Moderator

    The easiest is perhaps reinstall it.

Viewing 15 posts - 31 through 45 (of 172 total)