Mindboggle Versions Save

This release fixes v1.3.7, which had a unicode character in the docstrings of three files, which caused an error. It also includes a Singularity recipe in neurodocker.sh.

Version 1.3.7 includes all changes since v1.2.4_thickinthehead_revision:

September 24, 2019 (v1.3.7): Satrajit Ghosh added etelemetry capabilities.

June 14, 2019 (v1.3.6): Joachim Giard patched the curvature C++ code to support non-default curvature measures.

June 11, 2019 (v1.3.5):

• Conda installed a version of vtk that broke the C++ code (https://github.com/nipy/mindboggle/issues/171).
  • updated neurodocker.sh to install VTK 8.2
• Nipype is breaking with "UnicodeEncodeError" when it tries to generate reports
  • added "encoding='utf-8'" in nipype files in the docker image to resolve issue #175
• Zernike moments stopped working because scipy.misc was deprecated.
  • replaced with scipy.special
• Non-default (Gaussian, min, max) curvature file names were incomplete.
  • fixed the names so that they have the same filestem as the mean curvature file
• Image links broken in docs and in the Jupyter notebook tutorial.
  • moved images to a new GitHub repository "nipy/mindboggle-assets"

We have revised the thickinthehead algorithm from the original published version. We removed upsampling to reduce memory issues for large image volumes, and replaced the estimated volume of middle cortical layer with an estimate of its surface area. We made these revisions to be less susceptible to deviations in voxel size from isometric 1mm^3 voxels for which thickinthehead was originally built (see issue #149).

This minor release fixes issue #145: "mindboggle123 volume results change each time because of stochastic element in ANTs segmentation algorithm".

Every time the mindboggle123 script is run, FreeSurfer, ANTs, and Mindboggle generate output. FreeSurfer and Mindboggle always generate the same output given the same input, but we were alerted today that the Atropos segmentation algorithm in ANTs has a stochastic element that is on by default ("-u 1"), leading to slightly different volume shape measures each time it is run. Since mindboggle123 script calls this algorithm, every time it is run, we likewise get slightly different volume shape measures. Surface shape measures are unaffected. By adding the "-u 0" flag to the antsCorticalThickness.sh call in the mindboggle123 script, we now get the same results every time.

This minor release adds a couple of pull requests, including one that contains additional colormap scripts (#98 by @Shuo-Han) and another (#132 by @satra) that makes ANTs run much faster with much less memory requirements (in mindboggle123: "corticalthickness.inputs.use_floatingpoint_precision = True").

This minor release adds arguments to the mindboggle123 command to enable more flexible use of FreeSurfer.

"--fs_flags": Include additional recon-all flags like nuintensitycor-3T

"--fs_T2image": Include an optional T2 image to use with FreeSurfer

This release of Mindboggle enables a user to execute (1) recon-all in FreeSurfer, (2) antsCorticalThickness.sh in ANTs, and (3) mindboggle with a single docker command:

docker run --rm -ti -v $HOST:$DOCK nipy/mindboggle $IMAGE --id $ID

The container is run as an executable, which in turn runs a nipype script called "mindboggle123." The user doesn't have to enter the bash shell inside the container as before, unless finer control over the individual commands is required.

This release of Mindboggle was created after a nipype workshop at MIT:

http://nipy.org/workshops/2017-03-boston/index.html

This version includes atlas and transform files (less than 400kb) in a data/ directory rather than downloading and caching them so everything runs offline except for some docstring tests. A new Dockerfile and online instructions permit running FreeSurfer's recon-all and ANTs's antsCorticalThickness.sh and Mindboggle pipelines all within one Docker container. It also fixes a bug that could potentially allow for negative surface areas. This version will soon be followed by a revised Jupyter notebook tutorial.

This release was used to analyze the Mindboggle-101 data set (https://osf.io/nhtur/). The analysis was published later:

Klein A, Ghosh SS, Bao FS, Giard J, HameY, Stavsky E, Lee N, Rossa B, Reuter M, Neto EC, Keshavan A. (2017) Mindboggling morphometry of human brains. PLoS Computational Biology 13(3): e1005350. http://dx.doi.org/10.1371/journal.pcbi.1005350

Documentation of the Mindboggle-101 data is in the following publication:

101 labeled brain images and a consistent human cortical labeling protocol Arno Klein, Jason Tourville. Frontiers in Brain Imaging Methods. 6:171. http://dx.doi.org/10.3389/fnins.2012.00171

This official release of Mindboggle coincides with the publication of the Mindboggle article in PLoS Computational Biology:

Klein A, Ghosh SS, Bao FS, Giard J, HameY, Stavsky E, Lee N, Rossa B, Reuter M, Neto EC, Keshavan A. (2017) Mindboggling morphometry of human brains. PLoS Computational Biology 13(3): e1005350. doi:10.1371/journal.pcbi.1005350