1.5.0 RC2

Release Notes

Two hallmark changes conducive to a new minor release line have been included in version 1.5.0: the upgrade of PyBIDS to the 0.9 series and the split of SDCflows off from fMRIPrep codebase.
PyBIDS 0.9.x has a better handling of the indexed dataset that will permit some optimizations to fMRIPrep's memory fingerprint.
SDCflows now are found at poldracklab/sdcflows, and have been split to allow a more granular and thorough testing in collaboration with M. Cieslak, A. Adebimpe, and T. Satterthwaite.
Some other bugfixes, improvements to the documentation and minor features are also shipped with the new release.

CHANGES

• FIX: Postpone pandoc conversion of boilerplate after workflow has fully run (#1710) @oesteban
• FIX: Use MNI152NLin2009cAsym for SDC if no templates are specified (#1703) @effigies
• FIX: Correct BOLD-T1w registration description if DoF != 9 (#1701) @effigies
• ENH: Make fmriprep print defaults for arguments with help (#1735) @gdevenyi
• ENH: Add check for updates and check whether version has been flagged. (#1715) @oesteban
• ENH: Add TaskName metadata to BOLD outputs (#1714) @effigies
• REF: Separate SDCflows from fMRIPrep (#1670) @oesteban
• DOC: Update .zenodo.json (#1732) @utooley
• DOC: Add link to documentation in help of --output-spaces argument (#1722) @oesteban
• DOC: adding recon-all issue to faq (#1622) @franklin-feingold
• DOC: Add copyright waiver to boilerplate and reviewer note to docs (#1691) @effigies
• MAINT: Bump pybids and nipype dependencies (#1744) @effigies
• MAINT: Move regression tests of EPI masks over to Niworkflows (#1716) @oesteban
• MAINT: Remove old extensions entity selector for PyBIDS queries (#1707) @oesteban
• MAINT: Use PyBIDS 0.9.x via niworkflows/smriprep PRs (#1695) @effigies

1.5.0rc1

Release Notes

Two hallmark changes conducive to a new minor release line have been included in
version 1.5.0: the upgrade of PyBIDS to the 0.9 series and the split of SDCflows
off from fMRIPrep codebase.
PyBIDS 0.9.x has a better handling of the indexed dataset that will permit some
optimizations to fMRIPrep's memory fingerprint.
SDCflows now are found at poldracklab/sdcflows,
and have been split to allow a more granular and thorough testing in collaboration
with M. Cieslak, A. Adebimpe, and T. Satterthwaite.
Some other bugfixes, improvements to the documentation and minor features are also
shipped with the new release.

CHANGES

• ENH: Add TaskName metadata to BOLD outputs (#1714) @effigies
• REF: Separate SDCflows from fMRIPrep (#1670) @oesteban
• MAINT: Remove old extensions entity selector for PyBIDS queries (#1707) @oesteban
• FIX: Postpone pandoc conversion of boilerplate after workflow has fully run (#1710) @oesteban
• DOC: adding recon-all issue to faq (#1622) @franklin-feingold
• FIX: Use MNI152NLin2009cAsym for SDC if no templates are specified (#1703) @effigies
• FIX: Correct BOLD-T1w registration description if DoF != 9 (#1701) @effigies
• MAINT: Use PyBIDS 0.9.x via niworkflows/smriprep PRs (#1695) @effigies
• DOC: Add copyright waiver to boilerplate and reviewer note to docs (#1691) @effigies

1.4.1

Release Notes

As of 1.4.1, the new infant and pediatric templates added to TemplateFlow are available to
brain extraction and spatial normalization.
Containers do not set the TEMPLATEFLOW_HOME environment variable anymore, allowing
increased flexibility when running them (especially Singularity, for which the filesystem
is generally set read-only).
Additionally, BOLD files in native space can be generated and a minor bug related to the
handling of FreeSurfer outputs have been included.

CHANGES

• ENH: Show informative warning for phase1/2 type of fieldmaps (#1689) @oesteban
• DOC: Add WHT to Zenodo (#1683) @wiheto
• FIX: Finalizing support for new templates and their command line modifiers (#1671) @oesteban
• FIX: Do not set TEMPLATEFLOW_HOME (#1669) @oesteban
• ENH: Allow template modifiers (a la --output-spaces) in skull-stripping (#1666) @oesteban
• ENH: Write outputs in native-BOLD space (#1646) @oesteban
• FIX: FreeSurfer failed without adding some fs* to --output-spaces (#1643) @oesteban

1.4.1rc5

rel(1.4.1): Update CHANGES [skip ci]

1.4.1rc4

Merge pull request #1671 from oesteban/anat/improved-several-templates

FIX: Finalizing support for new templates and their command line modifiers

1.4.1rc3

Testing whether TEMPLATEFLOW_HOME can be left unset.

1.4.1rc2

Release Notes

As of 1.4.1, the new infant and pediatric templates added to TemplateFlow are available to
brain extraction and spatial normalization.
Additionally, BOLD files in native space can be generated and a minor bug related to the
handling of FreeSurfer outputs has been included.

CHANGES

• ENH: Allow template modifiers (a la --output-spaces) in skull-stripping (#1666) @oesteban
• ENH: Write outputs in native-BOLD space (#1646) @oesteban
• FIX: FreeSurfer failed without adding some fs* to --output-spaces (#1643) @oesteban

1.4.1rc1

Testing manylinux1 wheels generation

1.4.0

Release Notes

The new 1.4 series include several new features, several maintenance patches, and numerous bugfixes. The largest change to fMRIPrep's interface is the new --output-spaces argument that allows running spatial normalization to one or more standard templates, and also to indicate that data preprocessed and resampled to the individual's anatomical space should be generated. The implementation of this option will be completed in future releases to include new nonstandard spaces (e.g., this BOLD run's native space) and custom templates providing a path. For example, the following fMRIPrep options:

  --template MNI152NLin6Asym --output-space template T1w fsaverage5 ----template-resampling-grid 2mm

now would be accomplished with:

  --output-spaces MNI152NLin6Asym:res-2 anat fsaverage:den-10k

with the difference that more templates could be specified if needed, e.g.,

  --output-spaces MNI152NLin6Asym:res-2 anat fsaverage:den-10k MNI152NLin2009cAsym:native

Related anatomical preprocessing workflows from sMRIPrep have gone through thorough revisions. In particular, the brain extraction workflow now is implemented in pure Nipype.

Users will notice the addition of two new subsections in the reports generated by fMRIPrep. The first addition describes the cumulative variance explained by successive a/tCompCor components. A second addition shows the correlations between the confounding regressors that fMRIPrep writes to the corresponding file, and their correlation to the global signal.

Series 1.4 increasingly relies on PyBIDS to handle not only inputs, but also outputs and reporting. The reports generation system has been deeply refactored to improve its generalizability across BIDS-Apps and addressing some rendering problems (e.g., when resizing ICA-AROMA components decompositions). Finally, there were several updates to packaging, testing and documentation, which should hopefully improve the experience for new users and contributors.

With thanks to Yaroslav Halchenko, Dan Lurie, Adriana Rivera-Dompenciel, Franklin Feingold, Markus Sneve, Anibal Heinsfeld, and James Kent for contributions.

CHANGES

• [FIX] Incorrect transforms being applied to aparc/asegs in standard space (#1636) @oesteban
• [FIX] Writing functional derivatives in standard spaces (#1632) @oesteban
• [FIX] Resampling of BOLD into standard spaces (#1627) @oesteban
• [FIX] Re-enable correct versioning within container (#1623) @oesteban
• [FIX] Update spline fitting distance for BOLD bias-field correction (#1603) @markushs
• [FIX] Mount Nipype config file under new $HOME (#1602) @oesteban
• [FIX] Not having template as one --output-space crashes fMRIPrep (#1560) @oesteban
• [ENH] Add --dummy-scans option (#1559) @jdkent
• [ENH] Reduce BSplineFieldmap memory usage (#1609) @effigies
• [ENH] Confound model enhancement (#1487, #1586) @rciric
• [ENH] Allow multiple {non,}standard spaces (--output-spaces) (#1596) @oesteban
• [ENH] Switch to the refactored report generation from NiWorkflows (#1599) @oesteban
• [ENH] Force compression of derivative NIfTI volumes (#1600) @effigies
• [ENH] Do not allow writing derivatives directly into the BIDS root folder (#1589) @oesteban
• [ENH] Support 4D SBRefs when generating the bold reference (#1581) @oesteban
• [ENH] Remove pre-existing citation files before running (#1567) @yarikoptic
• [DOC] Improve appearance of parameter types in API docs (#1633) @anibalsolon
• [DOC] Confound enhancement documentation (#1625) @rciric
• [DOC] Add FAQ, Tips, Tricks section to RTD (cont. #1601) (#1610) @franklin-feingold
• [DOC] Companion of #1596 + punctual improvements of docs (#1605) @oesteban
• [DOC] Add examples/clarify CONTRIBUTORS.md (#1566) @jdkent
• [DOC] Add .bidsignore requirement to docs on lesion masks (#1574) @danlurie
• [DOC] Bump references of python3.6 to python3.7 (#1562) @jdkent
• [MAINT] Consolidate build configuration in setup.cfg (#1607) @effigies
• [MAINT] Progress bids-validator 1.2.3 (#1583) @yarikoptic
• [MAINT] Pin nilearn!=0.5.0,!=0.5.1 to avoid problems plotting mgz images (#1585) @oesteban
• [MAINT] Group all 'finished running with errors' sentry messages (#1552) @chrisgorgo
• [MAINT] Include hash of dataset_description.json in sentry metadata (#1553) @chrisgorgo
• [CI] Reduce pointless expenditures (#1614) @effigies
• [CI] Use caches to pass data between jobs (#1608) @effigies
• [CI] Build docs outside of container (#1606) @effigies
• [CI] Fix label on CI step "Skipping doc building job" (#1564) @ariveradompenciel

1.4.0a1

A second release candidate, addressing a critical issue found on 1.4.0a0 (#1626)