March 31, 2024 tedana newsletter
tedana-newsletter
tedana March 2024 Update
General updates
Tedana will have a poster at OHBM 2024 on Wednesday and Thursday, June 26-27. We will also make a list of multi-echo fMRI related content at OHBM, whether or not tedana is used. You can view and request to add content here: https://github.com/ME-ICA/ohbm-2024-multiecho
We’ve only recently started to use this new list. If you know others who might want these infrequent emails, invite them to join: https://groups.google.com/g/tedana-newsletter
We’ve had a productive few months of development with both user-facing and internal changes and we just released version 24.0.0!
tedana version 24.0.0
By default, tedana has been saving 4D volumes of the high-kappa components Accepted_bold.nii.gz and the low-kappa components Rejected_bold.nii.gz even though very few people use them and they require a lot of space. These will now only be saved if the program is run with --verbose. Additionally our final denoised time series was called desc-optcomDenoised_bold.nii.gz and this created confusion. It is now called desc-denoised_bold.nii.gz. This will break pipelines that looked for a file with the previous name. More details are in #1033
We noticed a small difference between the decision tree implemented in MEICA v2.5 and the tree we were calling kundu. We have renamed our existing kundu tree to tedana_orig, and there is now a new meica tree that should match the MEICAv2.5 method. In practice, tedana_orig and meica will either give identical results or meica will accept additional components. The additionally accepted components can have substantial variance and, upon visual inspection, usually look like they should have been rejected. Therefore, we've kept the same default (tedana_orig), but give both options to users. More details are in #952 and in a new section in our FAQ.
Different metrics, like kappa and rho, are calculated for each ICA component. While the code allowed for a range of different metrics, the list of what was calculated when tedana was run was impossible to change without editing the code. The metrics that were already specified in the decision tree json files will now be the ones calculated. The actual metric calculations still need to be defined within the code, but this change makes it practical to add a range of additional metrics that can vary by decision tree. More details are in #969
The tedana_report.html file now includes the mean T2* and S0 maps used in calculations #1040, consistent orientations for all images of brain slices #1045, version numbers for key python packages used during execution #1014, and the reference list is now properly rendered #1001.
Full release notes: https://github.com/ME-ICA/tedana/releases/tag/24.0.0
Active work
We are actively working on several additions where more contributors are welcome. This includes both direct contributions to the code, educational material, and documentation, as well as people who run new versions of the code on their data to make sure tedana runs as expected across a wider range of data sets. Active areas of work, which we expect to release in the next version of tedana include:
Adding in a more stable version of ICA which might include a more stable estimate of the total number of components. More details: #1013
Fitting user-provided time series to each component so that measures like head motion can be included in the decision tree process for classifying and labeling accepted or rejected components. #1064
Getting help with tedana or multi-echo fMRI
Questions about multi-echo fMRI and tedana usage or development can be posted at https://neurostars.org with tedana or multi-echo tags or as an issue or discussion at https://github.com/ME-ICA/tedana or https://mattermost.brainhack.org/brainhack/channels/tedana. We actively monitor all three message boards and try to efficiently respond.
Contributors
We are excited to welcome new contributors: @goodalse2019, @martinezeguiluz, and @martaarbizu. We continue to thank all of our contributors for their continued input and help with tedana. We always look forward to seeing more new faces! If you’re not sure where to start, please feel free to open an issue on github, ask a question on neurostars with the ‘tedana’ or ‘multi-echo’ tag, join our next monthly developer’s call, or join the tedana mattermost channel.