NITRC BXH/XCEDE Tools Forum: help

RE: bxh python missing analyzetools

Tristan Yates — Fri, 31 Jan 2025 16:29:20 GMT

Thank you so much for your rapid response! This is helpful and I think our needs are covered by the existing tools.  One quick note that bxhabsorb seems to work in creating a bxh file from the nii.gz file, but results in this warning: Error importing bxh.converters.dicom: No module named 'pydicom._storage_sopclass_uids'   warnings.warn("Error importing bxh.converters.dicom: %s" % (e,))

bxh python missing analyzetools

Syam Gadde — Thu, 30 Jan 2025 14:21:14 GMT

analyze2bxh can be replaced with bxhabsorb. And most tools will now output .nii.gz with a .bxh file.
If you're trying to create nifti files from a bxh file, we usually now use bxhselect INPUT.bxh OUTPUTPREFIX which will create OUTPREFIX.bxh and OUTPUTPREFIX.nii.gz

If that doesn't work for you, let me know!

bxh python missing analyze tools

Tristan Yates — Thu, 30 Jan 2025 3:16:44 GMT

Hello! I noticed the new bxh release for python, which is fantastic! However, both analyze2bxh and bxh2analyze (which to my knowledge are necessary tools for interacting with nifti images) seem to be missing from the package. Is there any reason for this? Would it be possible to include these tools in the updated release?

RE: volumemeans or volmeasures?

Syam Gadde — Tue, 31 May 2022 12:28:39 GMT

fmriqa_generate.pl calls fmriqa_volmeasures to generate that data so they should give you similar numbers, however fmriqa_generate.pl may remove some datapoints from the beginning before sending data to fmriqa_volmeasures. The center-of-mass is a crude measure that can be helpful. For more accurate position information you may consider tools like FSL's mcflirt.

volumemeans or volmeasures?

Mollie Neason — Mon, 30 May 2022 11:31:51 GMT

Hello,

I was wondering if I could have some clarification on the difference between the volumemeans output (cmassx/y/z from fmriqa_generate.pl) and the volmeasures output (from fmriqa_volmeasures)?

I am trying to use BXH tools to identify absolute motion artefacts in my fmri data - which would you advise would be best for this?

Thank you very much!
Mollie

Catalina / 64-bit Catalina issues

Hannah Swearingen — Mon, 14 Dec 2020 17:00:28 GMT

Hi,

I am trying to run the firm_phantomqa.pl program, but am running into some issues with the binary programs that come with the most recent package: bxh_xcede_tools-1.11.14-MacOSX.i686. This package is not compatible with MacOS Catalina or greater, as Catalina has dropped support for 32-bit executables.

For example, the error I get when attempting to run < ./bxhabsorb > is "Bad CPU type in executable"

Does anybody know a way to run these programs in macOS Catalina? Or is there an updated software package available that I can download to use?

Thanks,
Hannah

Can't find frequency direction in NIFTI header

Aras K — Thu, 15 Nov 2018 16:24:51 GMT

Dear all,

I am running the fBIRN QA pipeline version 1.10.7-lsb30.x86_64 with AFNI version AFNI_2011_12_21_1014. I have NIFTI fmri images and I convert them to BXH/XCEDE using analyze2bxh --xcede. This is followed by running the fBIRN QA pipeline on the BXH/XCEDE data files using fmriqa_phantomqa.pl.

The fBIRN QA pipeline fails to generate ghost measures with the following error:

"volnum ghostpercentage brightghostpercentageCan't find frequency direction in header! Not calculating ghost metrics (not EPI?)"

It appears that the pipeline cannot find the information pertaining to frequency/phase-encode directions in the NIFTI header. I tried to manually modify the NIFTI header by adding the field 'dim_info' but it did not resolve the issue. Any suggestion pertaining to a solution?

Thanks,

Aras

RE: dyld errors on Mac OS X

Matthew Johnson — Fri, 25 Mar 2016 21:16:54 GMT

Sounds good! Yes, it seems reasonable to me to just have people install ImageMagick for themselves... MacPorts and Homebrew make that pretty easy even if you're not running Linux.

dyld errors on Mac OS X

Syam Gadde — Fri, 25 Mar 2016 14:05:13 GMT

Thanks for doing the grunt work, and providing detailed workarounds to these issues. I will likely get rid of the ImageMagick executables packaged with bxh_xcede_tools since the standard ImageMagick tools are stable enough now. In the past, ImageMagick behavior would change drastically from minor version to minor version, but now things are less volatile.

dyld errors on Mac OS X

Matthew Johnson — Fri, 25 Mar 2016 1:32:38 GMT

Not really a help request per se as I just figured out my own problem, but I'm leaving the solution here for posterity. I have already submitted a bug report with similar information to that below.

The ImageMagick programs contained within the BXH/XCEDE tools don't appear to work on current versions of Mac OS X (I'm on El Capitan here, but from my searches it appears that this bug would happen for any version >= 10.7 Lion). I ran into this using fmriqa_generate.pl but any other functions that use ImageMagick likely have the same problem.

Errors like this appear when invoking ImageMagick's 'convert' or 'montage' tools:
dyld: Library not loaded: /System/Library/Frameworks/OpenCL.framework/Versions/A/Libraries/libclparser.dylib
Referenced from: /Users/mjohnson113/Downloads/bxh_xcede_tools-1.11.1-MacOSX.i686/bin/convert
Reason: image not found

From web searches, I deduced that the .dylib in question doesn't seem to exist anymore in Lion and up.

I fixed the problem for myself by installing ImageMagick via MacPorts and replacing the 'montage' and 'convert' binaries in the BXH/XCEDE package with the freshly built copies from MacPorts. Not sure what the best long-term solution would be for the project.

Anyway, if you are having the same problem, installing your own copy of ImageMagick and replacing the broken BXH/XCEDE copies with custom-built ones should, in theory, work for you too!