Robust-tedana: Advancing Multi-Echo fMRI Denoising
Poster No:
1039
Submission Type:
Abstract Submission
Authors:
Bahman Tahayori1,2, Robert Smith1,2, David Vaughan1,2, Eric Pierre1,2, Daniel Handwerker3, Chris Tailby1,2, Graeme Jackson1,2, David Abbott1,2, for the Australian Epilepsy Project Investigators1
Institutions:
1The Florey Institute of Neuroscience and Mental Health, Heidelberg, VIC, Australia, 2Florey Department of Neuroscience and Mental Health, The University of Melbourne, Parkville, VIC, Australia, 3National Institute of Mental Health, Bethesda, MD
First Author:
Bahman Tahayori
The Florey Institute of Neuroscience and Mental Health|Florey Department of Neuroscience and Mental Health, The University of Melbourne
Heidelberg, VIC, Australia|Parkville, VIC, Australia
The Florey Institute of Neuroscience and Mental Health|Florey Department of Neuroscience and Mental Health, The University of Melbourne
Heidelberg, VIC, Australia|Parkville, VIC, Australia
Co-Author(s):
Robert Smith
The Florey Institute of Neuroscience and Mental Health|Florey Department of Neuroscience and Mental Health, The University of Melbourne
Heidelberg, VIC, Australia|Parkville, VIC, Australia
The Florey Institute of Neuroscience and Mental Health|Florey Department of Neuroscience and Mental Health, The University of Melbourne
Heidelberg, VIC, Australia|Parkville, VIC, Australia
David Vaughan
The Florey Institute of Neuroscience and Mental Health|Florey Department of Neuroscience and Mental Health, The University of Melbourne
Heidelberg, VIC, Australia|Parkville, VIC, Australia
The Florey Institute of Neuroscience and Mental Health|Florey Department of Neuroscience and Mental Health, The University of Melbourne
Heidelberg, VIC, Australia|Parkville, VIC, Australia
Eric Pierre
The Florey Institute of Neuroscience and Mental Health|Florey Department of Neuroscience and Mental Health, The University of Melbourne
Heidelberg, VIC, Australia|Parkville, VIC, Australia
The Florey Institute of Neuroscience and Mental Health|Florey Department of Neuroscience and Mental Health, The University of Melbourne
Heidelberg, VIC, Australia|Parkville, VIC, Australia
Chris Tailby
The Florey Institute of Neuroscience and Mental Health|Florey Department of Neuroscience and Mental Health, The University of Melbourne
Heidelberg, VIC, Australia|Parkville, VIC, Australia
The Florey Institute of Neuroscience and Mental Health|Florey Department of Neuroscience and Mental Health, The University of Melbourne
Heidelberg, VIC, Australia|Parkville, VIC, Australia
Graeme Jackson
The Florey Institute of Neuroscience and Mental Health|Florey Department of Neuroscience and Mental Health, The University of Melbourne
Heidelberg, VIC, Australia|Parkville, VIC, Australia
The Florey Institute of Neuroscience and Mental Health|Florey Department of Neuroscience and Mental Health, The University of Melbourne
Heidelberg, VIC, Australia|Parkville, VIC, Australia
David Abbott, PhD
The Florey Institute of Neuroscience and Mental Health|Florey Department of Neuroscience and Mental Health, The University of Melbourne
Heidelberg, VIC, Australia|Parkville, VIC, Australia
The Florey Institute of Neuroscience and Mental Health|Florey Department of Neuroscience and Mental Health, The University of Melbourne
Heidelberg, VIC, Australia|Parkville, VIC, Australia
for the Australian Epilepsy Project Investigators
The Florey Institute of Neuroscience and Mental Health
Heidelberg, VIC, Australia
The Florey Institute of Neuroscience and Mental Health
Heidelberg, VIC, Australia
Introduction:
Multi-echo (ME) acquisition has been shown to improve differentiation of neural signals from noise in functional MRI (fMRI) data (Kundu, 2012). TE-Dependent ANAlysis (tedana) is an open-source tool for removing noise in ME-fMRI data that can support Multi-Band (MB) acquisition as well (DuPre, 2021). However, tedana can, in some instances, remove signal arising from neuronal activity to the extent in some single-subject analyses that most activation is removed. While manual curation of denoising can mitigate such, this made tedana unsuitable for studies such as the Australian Epilepsy Project (AEP), where we are collecting fMRI data from 4000 participants and, therefore, require automated analyses that yield robust results at the single-subject level. We therefore set about developing an improved pipeline that yields robust results.
Methods:
We developed Robust-tedana (Fig. 1), a pipeline incorporating three main steps: 1) Applying Marchenko-Pastur Principal Component Analysis (MP-PCA) to unprocessed fMRI data for early thermal noise suppression (Ades-Aron, 2021), 2) Preprocessing individual thermal-denoised echoes with fMRIPrep (Esteban, 2019) and 3) Multi-echo-based denoising using a modified version of tedana. These modifications include: preserving all components at the PCA step to disable thermal noise suppression; replacing the Independent Component Analysis (ICA) with robust ICA (Anglada-Girotto, 2022), which clusters many ICA executions, with different initial seed values, to yield a consensus result robust to potentially large stochastic variation; and modifying tedana's component classification algorithm.
We collected T1-weighted and multi-band multi-echo (MBME) fMRI data for 250 AEP participants performing a language task in a 3T Siemens PrismaFit MRI scanner, with the following parameters: three echoes at TE = [15 33.25 51.5]ms, TR = 0.9s, MB factor = 4, FOV =216mm, voxel size = 3*3*3 mm3 and 202 volumes per subject with anterior-posterior phase encoding direction.
The standard tedana pipeline (version 23.0.2, Fig. 1 left) and Robust-tedana pipeline (Fig. 1 right) were compared by quantifying supra-threshold activation volumes and mean z-scores within a pre-defined language ROI.
We collected T1-weighted and multi-band multi-echo (MBME) fMRI data for 250 AEP participants performing a language task in a 3T Siemens PrismaFit MRI scanner, with the following parameters: three echoes at TE = [15 33.25 51.5]ms, TR = 0.9s, MB factor = 4, FOV =216mm, voxel size = 3*3*3 mm3 and 202 volumes per subject with anterior-posterior phase encoding direction.
The standard tedana pipeline (version 23.0.2, Fig. 1 left) and Robust-tedana pipeline (Fig. 1 right) were compared by quantifying supra-threshold activation volumes and mean z-scores within a pre-defined language ROI.
Results:
Robust-tedana demonstrated increased mean z-scores and activation volumes within the language area compared to tedana. Investigating individual subject activation maps, there were occasional cases where TEDANA yielded only minimal test statistics commensurate with engagement with the language task; for these problematic subjects, Robust-tedana reported more extensive language network activation. A few examples for extreme cases are shown in Fig. 2. The radiological orientation is used in this figure (subject's left on image right). Robust-tedana yielded more extensive activation for 87% and 80% of subjects in terms of activation volume and mean z-score, respectively. The group average of the mean z-scores within the language ROI was significantly higher when using the Robust-tedana pipeline compared to the tedana pipeline. The mean z-score increased from 1.41 to 1.73 (p-value <10^(-11)). Most importantly, activation in expected areas was observed in all subjects when using Robust-tedana.
Conclusions:
Robust-tedana improved the denoising of ME-fMRI data, increasing the capacity to detect neural activity at both the individual and the group level. Thus, this pipeline can be utilised for denoising large cohorts, where automated pipelines are desirable. The alterations to tedana described herein have recently been integrated into tedana as part of update 24.0.2.
Modeling and Analysis Methods:
Activation (eg. BOLD task-fMRI) 1
Neuroinformatics and Data Sharing:
Workflows 2
Keywords:
Data analysis
fMRI CONTRAST MECHANISMS
FUNCTIONAL MRI
Language
Open-Source Code
Workflows
Other - Multi-echo fMRI
1|2Indicates the priority used for review
Abstract Information
By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute your work in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels, the OHBM website, or other electronic publications and media.
The Open Science Special Interest Group (OSSIG) is introducing a reproducibility challenge for OHBM 2025. This new initiative aims to enhance the reproducibility of scientific results and foster collaborations between labs. Teams will consist of a “source” party and a “reproducing” party, and will be evaluated on the success of their replication, the openness of the source work, and additional deliverables. Click here for more information. Propose your OHBM abstract(s) as source work for future OHBM meetings by selecting one of the following options:
Please indicate below if your study was a "resting state" or "task-activation” study.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Was this research conducted in the United States?
Were any human subjects research approved by the relevant Institutional Review Board or ethics panel? NOTE: Any human subjects studies without IRB approval will be automatically rejected.
Were any animal research approved by the relevant IACUC or other animal research panel? NOTE: Any animal studies without IACUC approval will be automatically rejected.
Please indicate which methods were used in your research:
For human MRI, what field strength scanner do you use?
Which processing packages did you use for your study?
Provide references using APA citation style.
Anglada-Girotto, M., et al (2022). robustica: customizable robust independent component analysis. BMC bioinformatics, 23(1), 519.
DuPre, E.M., et al (2021). TE-dependent analysis of multi-echo fMRI with Tedana. Journal of Open Source Software, 6(66), 3669.
Esteban O., et al (2019). fMRIPrep: a robust preprocessing pipeline for functional MRI. Nature Methods, 16(1), 111–116.
Kundu, P., et al (2012). Differentiating BOLD and non-BOLD signals in fMRI time series using multi-echo EPI. NeuroImage, 60(3), 1759-1770.