Bibliographic record
Abstract
Release Notes Tedana's 0.0.11 release includes a number of bug fixes and enhancements, and it's associated with publication of our Journal of Open Source Software (JOSS) paper! Beyond the JOSS paper, two major changes in this release are (1) outputs from the tedana and t2smap workflows are now BIDS compatible, and (2) we have overhauled how masking is performed in the tedana workflow, so that improved brain coverage is retained in the denoised data, while the necessary requirements for component classification are met. 🔧 Breaking changes The tedana and t2smap workflows now generate BIDS-compatible outputs, both in terms of file formats and file names. Within the tedana workflow, T2* estimation, optimal combination, and denoising are performed on a more liberal brain mask, while TE-dependence and component classification are performed on a reduced version of the mask, in order to retain the increased coverage made possible with multi-echo EPI. When running tedana on a user-provided mixing matrix, the order and signs of the components are no longer modified. This will not affect classification or the interactive reports, but the mixing matrix will be different. ✨ Enhancements tedana interactive reports now include carpet plots. The organization of the documentation site has been overhauled to be easier to navigate. We have added documentation about how to use tedana with fMRIPrep, along with a gist that should work on current versions of fMRIPrep. Metric calculation is now more modular, which will make it easier to debug and apply in other classification decision trees. 🐛 Bug fixes One component was not rendering in interactive reports, but this is fixed now. Inputs are now validated to ensure that multi-file inputs are not interpreted as single z-concatenated files. Changes since last stable release [JOSS] Add accepted JOSS manuscript to main (#813) @tsalo [FIX] Check data type in io.load_data (#802) @tsalo [DOC] Fix link to developer guidelines in README (#797) @tsalo [FIX] Figures of components with index 0 get rendered now (#793) @eurunuela [DOC] Adds NIMH CMN video (#792) @jbteves [STY] Use black and isort to manage library code style (#758) @tsalo [DOC] Generalize preprocessing recommendations (#769) @tsalo [DOC] Add fMRIPrep collection information to FAQ (#773) @tsalo [DOC] Add link to EuskalIBUR dataset in documentation (#780) @tsalo [FIX] Add resources folder to package data (#772) @tsalo [ENH] Use different masking thresholds for denoising and classification (#736) @tsalo [DOC, MAINT] Updated dependency version numbers (#763) @handwerkerd [REF] Move logger management to new functions (#750) @tsalo [FIX] Ignore non-significant kappa elbow when no non-significant kappa values exist (#760) @tsalo [ENH] Coerce images to 32-bit (#759) @jbteves [ENH] Add carpet plot to outputs (#696) @tsalo [FIX] Correct manacc documentation and check for associated inputs (#754) @tsalo [DOC] Reorganize documentation (#740) @tsalo [REF] Do not modify mixing matrix with sign-flipping (#749) @tsalo [REF] Eliminate component sorting from metric calculation (#741) @tsalo [FIX] Update apt in CircleCI (#746) @notZaki [DOC] Update resource page with dataset and link to Dash app visualizations (#745) @jsheunis [DOC] Clarify communication pathways (#742) @tsalo [FIX] Disable report logging during ICA restart loop (#743) @tsalo [REF] Replace metric dependency dictionaries with json file (#739) @tsalo [FIX] Add references back into the HTML report (#737) @tsalo [ENH] Allows iterative clustering (#732) @jbteves [REF] Modularize metric calculation (#591) @tsalo Rename sphinx functions to fix building error for docs (#727) @eurunuela [ENH] Generate BIDS Derivatives-compatible outputs (#691) @tsalo
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.004 | 0.014 |
| Meta-epidemiology (narrow) | 0.003 | 0.003 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.005 | 0.004 |
| Open science | 0.006 | 0.004 |
| Research integrity | 0.002 | 0.005 |
| Insufficient payload (model declined to judge) | 0.402 | 0.373 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".