BrainSuite BIDS App v23a

• In conjunction with the release of BrainSuite23a, we have released an updated version of the BrainSuite BIDS App
• The BrainSuite BIDS App is a containerized environment that enables you to run the BrainSuite environment using Docker or Singularity. All system requirements are included in the Docker, facilitating easy installation.
• The BrainSuite BIDS App includes BrainSuite Dashboard, a web-based system for monitoring the status of your study data as they are processed. This enables you to see errors and assess the quality of the derived data as they are created.
• The BrainSuite BIDS App does not include the BrainSuite GUI, which we recommend installing on your host machine if you are using Docker or Singularity for processing your data.

BrainSuite Statistics Toolbox in R (bstr)

• bssr had been renamed to bstr
• Updates to support BIDS compatibility
• Support excluding subjects from csv using the parameter exclude_col

BrainSuite Anatomical Pipeline

(new since 21a) The most major change to the software is a modification to the BrainSuite T1-weighted MRI processing workflow, which we have now termed the BrainSuite Anatomical Pipeline. The BrainSuite Anatomical Pipeline performs cortical surface extraction, cortical thickness estimation, surface-constrained volumetric registration, mapping of subject data to atlas space, and computation of subject-level statistics. We have separated the previous svreg program into multiple modules to improve the workflow.

We have developed new scripts to simplify batch processing of data. In the BrainSuite23a/bin folder, there is a script named brainsuite_anatomical_pipeline.sh (Mac/Linux) or brainsuite_anatomical_pipeline.cmd (Windows) which performs the full anatomical processing sequence. We have also created svreg_runall.{cmd,sh} which performs four steps using programs in the svreg bin folder: (cortical thickness estimation, surface-constrained volumetric registration, mapping of subject data to atlas space, and computation of subject-level statistics). The brainsuite_anatomical_pipeline script calls the cortical_surface_extraction and svreg_runall scripts.

BrainSuite Anatomical Pipeline Dialog

These changes to the workflow are also reflected in the BrainSuite Anatomical Pipeline Dialog, which is an update of the previous Cortical Surface Extraction dialog. The BAP Dialog also includes a new rewind feature. If you use the Automatically Save Results feature, then BrainSuite can rewind to a previous step by pressing on the rewind button (backwards circle arrow) next to the step name. This will rewind to the start of that step. This feature can also be used after closing and reopening BrainSuite.

USCBrain Atlas

The USCBrain atlas is now included as part of the BrainSuite release. The USCBrain atlas is a new high-resolution single-subject atlas, constructed using both anatomical and functional information to guide the parcellation of the cortex. For details, please see the webpage for the atlas or our preprint describing the atlas, Joshi AA, Choi S, Chong M, Sonkar G, Gonzalez-Martinez J, Nair D, Wisnowski JL, Haldar JP, Shattuck DW, Damasio H, Leahy RM (2020) “A Hybrid High-Resolution Anatomical MRI Atlas with Sub-parcellation of Cortical Gyri using Resting fMRI” bioRxiv 2020.09.12.294322; doi: https://doi.org/10.1101/2020.09.12.294322

Surface-constrained Volumetric Registration (SVReg) Workflow

The SVReg workflow has changed. There are now four separate executables:

1. thicknessPVC – cortical thickness estimation
2. svreg – surface and volume registration and atlas-based labeling
3. mapthickness – mapping thickness to atlas space
4. svreg_thickness2atlas – computation of subject-level statistics (e.g., GM and WM volumes within anatomical ROIs)

These can be called from a single script (svreg_runall), which runs them in the above order.

Cortical Thickness Estimation

We have removed the use of linked distance as an estimate for cortical thickness. All BrainSuite processing for cortical thickness now uses thicknessPVC, an implementation of the Anisotropic Laplace Equation method described in Joshi et al., 2018. The ALE method provides much greater accuracy than linked distance, which was used previously to provide a fast visualization of thickness during cortical surface generation.