Impacts of analytic workflows and modeling decisions on the estimated task fMRI activity.
Presented During:
Friday, June 27, 2025: 11:30 AM - 12:45 PM
Brisbane Convention & Exhibition Centre
Room:
P2 (Plaza Level)
Poster No:
1044
Submission Type:
Abstract Submission
Authors:
Michael Demidenko1, Jeanette Mumford1, Russell Poldrack1
Institutions:
1Stanford University, Stanford, CA
First Author:
Co-Author(s):
Introduction:
Functional magnetic resonance imaging (fMRI) tasks often produce signals that are difficult to detect, especially when studying individual differences (Poldrack et al., 2017, Elliott et al., 2020). Researchers often prioritize power in their analytic workflows by reducing collinearity (Liu et al., 2001). The Monetary Incentive Delay (MID; Knutson et al., [2001]) task has a multi-component trial structure that can exacerbate biases in the estimated BOLD activity when behavioral and BOLD timeseries are misaligned and subject-level models omit task-relevant regressors. Here, we highlight a GE timing issue and important model misspecification in MID subject-level models, and its impact on the estimated BOLD activity released as part of the Adolescent Brain Cognitive Development (ABCD) study® fMRI data.
Methods:
The ABCD sample used here is from the ABCD-BIDS Community Collection (ABCC:
Feczko et al., [2021]). The BIDS standard input data were preprocessed using fMRIPrep v23.1.4 (Esteban et al., 2022). To evaluate the timing of activity between scanners in the MID task, N = 50 subjects were subsampled from each of the GE, Siemens and Philips scanners. Peristimulus plots were created to visualize TR-by-TR in the mean percent signal change locked to the Probe onset in the MID task for a Left Motor region of interest across GE, Siemens and Philips scanner Baseline BOLD data. To illustrate effects of modeling decisions at the subject-level, N = 500 subjects were subsampled from the Siemens data Year 2 data (Demidenko et al., 2024). Subject-level general linear models (GLMs) were fit to the subject's run-level timeseries data using Nilearn. Two separate GLMs were estimated, the Data Analysis, Informatics & Resource Center's (DAIRC) model and, based on recent recommendations (Mumford et al., 2024), a Saturated model. The DAIRC instantaneous model (Hagler et al., 2019) fits an impulse function with a temporal derivative at the onset of the Cue and Feedback stimuli (a subset of the trial-by-trial structure). Our Saturated model fits a boxcar function using the trial onset and true durations for the Cue, Fixation, Probe and Feedback stimuli (all components of the trial-by-trial structure). For each of the 11 contrasts, one-sample t-test of the within-subject contrast maps was estimated using FSL's randomise using threshold-free cluster enhancement (TFCE). The DAIRC and Saturated models were compared using a two-sided t-test of the mean.
Feczko et al., [2021]). The BIDS standard input data were preprocessed using fMRIPrep v23.1.4 (Esteban et al., 2022). To evaluate the timing of activity between scanners in the MID task, N = 50 subjects were subsampled from each of the GE, Siemens and Philips scanners. Peristimulus plots were created to visualize TR-by-TR in the mean percent signal change locked to the Probe onset in the MID task for a Left Motor region of interest across GE, Siemens and Philips scanner Baseline BOLD data. To illustrate effects of modeling decisions at the subject-level, N = 500 subjects were subsampled from the Siemens data Year 2 data (Demidenko et al., 2024). Subject-level general linear models (GLMs) were fit to the subject's run-level timeseries data using Nilearn. Two separate GLMs were estimated, the Data Analysis, Informatics & Resource Center's (DAIRC) model and, based on recent recommendations (Mumford et al., 2024), a Saturated model. The DAIRC instantaneous model (Hagler et al., 2019) fits an impulse function with a temporal derivative at the onset of the Cue and Feedback stimuli (a subset of the trial-by-trial structure). Our Saturated model fits a boxcar function using the trial onset and true durations for the Cue, Fixation, Probe and Feedback stimuli (all components of the trial-by-trial structure). For each of the 11 contrasts, one-sample t-test of the within-subject contrast maps was estimated using FSL's randomise using threshold-free cluster enhancement (TFCE). The DAIRC and Saturated models were compared using a two-sided t-test of the mean.
Results:
Perstimulus plots illustrate that the BOLD timeseries locked to the Probe differed across GE and the Siemens/Philips data for Win, Lose and Neutral stimuli (Fig. 1A). As a result, data releases through release 5.0 for the Cue and Feedback locked impulse response models misestimate activity in GE data. Trimming one fewer dummy volumes in the GE data improved alignment between the BOLD timeseries and task events (Fig. 1B). The DAIRC and Saturated models elicited activity in similar regions (Fig. 2A). However, the models differed in key Cue and Feedback contrasts, particularly in condition versus baseline contrasts. This, in part, is related to the model misspecification in the DAIRC model, whereby the unmodeled true duration of the Cue and Feedback, and the unmodeled Fixation and Probe stimuli biased estimates of adjacent components in the GLM. Given that the Cue and Fixation components meaningfully differ (Fig. 2B) and do not have high collinearity, these task-relevant components should be included in subject-level models of the BOLD timeseries.
·Figure 1. Peristimulus Plots
·Figure 2. Activation Maps.
Conclusions:
Deviations in the scanner protocols and model misspecifications impacted BOLD estimates in the ABCD MID task fMRI data. Correcting the volumes improved the alignment between the timeseries and behavioral data. Modeling all trial-by-trial components recovered task-relevant activity without increasing collinearity in the MID task. Omitting task-relevant regressors results in BOLD estimates that impact validity and interpretation.
Emotion, Motivation and Social Neuroscience:
Reward and Punishment
Modeling and Analysis Methods:
Activation (eg. BOLD task-fMRI) 1
Univariate Modeling 2
Keywords:
Design and Analysis
FUNCTIONAL MRI
Modeling
Univariate
Other - Monetary Incentive Delay Task
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?
Are you Internal Review Board (IRB) certified? Please note: Failure to have IRB, if applicable will lead to automatic rejection of abstract.
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.
Esteban, O., Markiewicz, C. J., Goncalves, M., Provins, C., Kent, J. D., DuPre, E., Salo, T., Ciric, R., Pinsard, B., Blair, R. W., Poldrack, R. A., & Gorgolewski, K. J. (2022). fMRIPrep: A robust preprocessing pipeline for functional MRI [Computer software]. Zenodo. https://doi.org/10.5281/zenodo.7117719
Feczko, E., Conan, G., Marek, S., Tervo-Clemmens, B., Cordova, M., Doyle, O., Earl, E., Perrone, A., Sturgeon, D., Klein, R., Harman, G., Kilamovich, D., Hermosillo, R., Miranda-Dominguez, O., Adebimpe, A., Bertolero, M., Cieslak, M., Covitz, S., Hendrickson, T., … Fair, D. A. (2021). Adolescent Brain Cognitive Development (ABCD) Community MRI Collection and Utilities (p. 2021.07.09.451638). bioRxiv. https://doi.org/10.1101/2021.07.09.451638
Hagler, D. J., Hatton, S., Cornejo, M. D., Makowski, C., Fair, D. A., Dick, A. S., Sutherland, M. T., Casey, B. J., Barch, D. M., Harms, M. P., Watts, R., Bjork, J. M., Garavan, H. P., Hilmer, L., Pung, C. J., Sicat, C. S., Kuperman, J., Bartsch, H., Xue, F., … Dale, A. M. (2019). Image processing and analysis methods for the Adolescent Brain Cognitive Development Study. NeuroImage, 202, 116091. https://doi.org/10.1016/j.neuroimage.2019.116091
Knutson, B., Adams, C. M., Fong, G. W., & Hommer, D. (2001). Anticipation of Increasing Monetary Reward Selectively Recruits Nucleus Accumbens. Journal of Neuroscience, 21(16), RC159–RC159. https://doi.org/10.1523/JNEUROSCI.21-16-j0002.2001
Liu, T. T., Frank, L. R., Wong, E. C., & Buxton, R. B. (2001). Detection Power, Estimation Efficiency, and Predictability in Event-Related fMRI. NeuroImage, 13(4), 759–773. https://doi.org/10.1006/nimg.2000.0728
Poldrack, R. A., Baker, C. I., Durnez, J., Gorgolewski, K. J., Matthews, P. M., Munafò, M. R., Nichols, T. E., Poline, J.-B., Vul, E., & Yarkoni, T. (2017). Scanning the horizon: Towards transparent and reproducible neuroimaging research. Nature Reviews Neuroscience, 18(2), Article 2. https://doi.org/10.1038/nrn.2016.167