Releases: DCAN-Labs/infant-abcd-bids-pipeline
Update JLF templates + fix Docker build issue
ENH: bump dcan-infant-pipeline v0.0.14 -> v0.0.15, JLF templates updated with skullstripped anats and updated segmentations (DCAN-Labs/dcan-infant-pipeline#8)
FIX: pin bids-validator version to 1.14.0 (addresses Docker build issue related to DCAN-Labs/abcd-hcp-pipeline#117)
Add compiled code from hotfix for v0.0.21
Update to internal tools with DBP v.4.0.11
This version contains updated time-series after re-applying the bandpass filter to the BOLD data. The new implementation includes zero padding to the BOLD data to minimize distortions in the edges of the time series. It’s important to note that this is not a bug, but rather an improvement. This release does not invalidate previous results, it reduces variance towards the beginning and end of the time-series data. In the previous release, those frames are discharged since are labeled as "outliers" in the provided mask. Using these updated data those frames at the end will not be discharged if the increased variance was related to the "edge effect", so users should be able to include more data in their analyses now.
Merge from develop
Back to layered build (uses: ubuntu --> external-software --> internal-tools --> infant-abcd-bids-pipeline).
This allows all of the pipelines to synchronize on tools by using the same internal-tools release. (Helps with debugging.)
But mainly, when multiple pipelines using the same internal-tools are loaded onto a node, uses much less Docker space.
Update Dockerfile w/ latest dcan_bold_processing
Update Dockerfile w/ latest dcan_bold_processing (includes merge of dcan_bold_processing dev branch intended but missing from v0.0.17)
Latest dcan_bold_processing
Fixes issues with run numbers.
Latest Executive Summary
Pulled in fix that caused the combined grey plot files to fail.
Get latest Executive Summary
The main difference in this release is that a bug was fixed in Executive Summary, and this build takes the bug fix. The bug was that the atlas was not always being passed from ExecutiveSummary.py to its preprocessing script. This caused anat-only processing to use the wrong atlas for Atlas-in-T1w and T1w-in-Atlas figures.
Note: while this was being testing, DockerHub began having problems with starting from our large internal-tools image. To work around, had to also flatten the build into a single Dockerfile. To see how layered builds were done, see releases prior to this one.
Added FileMapper as a stage in the pipeline.
Added a FileMapper stage. If user supplies path to a file mapper json, the stage will be run. To use the default file mapper json for infants, specify the path as '/opt/dcan-tools/filemapper/current_infant.json'
Fix the ANTs threading problem
There were 10 instances of ANTs and each was allocating up to 10 threads, but we were only asking for 10 cores. Now, each can ask for up to 2 threads. Must ask for 20 cores.