Optimizing Preprocessing Pipeline for Vector Autoregressive-Based Effective Connectivity Analysis
Poster No:
1192
Submission Type:
Abstract Submission
Authors:
Jinho Bae1, Hee Soo Park1, Joonmin Lee1, Seoyeong Ha1, Hang Joon Jo2
Institutions:
1Hanyang Univ., Seoul, Seoul Metropolitan City, 2College of Medicine, Hanyang Univ., Seoul, Seoul Metropolitan City
First Author:
Co-Author(s):
Introduction:
In effective connectivity (EC) analysis, several preprocessing steps are necessary, and adjustments to the specific parameters within these steps can affect the results. In this study, we compared the effects of different choices for smoothing techniques, time series extraction methods, optimal lag order selection, and the primary analytical approach on EC outcomes. By examining these variations, we aimed to assess how each factor may influence the results of EC analysis.
Methods:
RS fMRI data (n=58) were randomly obtained from the FCON-1000 project (Biswal et al., 2010). The target of the EC analysis was the visuomotor network, and the regions of interest (ROIs) within this network were derived from the FreeSurfer segmentation results of each subject (Fischl, 2012). Preprocessing was performed using the AFNI software packages (Cox, 1996). We manipulated the following conditions during the preprocessing: (i) application of smoothing to the EPI images (unsmoothed / isotropic Gaussian smoothing with FWHM=6mm), (ii) time series extraction methods (averaging the signal within each ROI / selecting the first principal component / creating a time series that explains 95% of the variance within each ROI), and (iii) primary EC analysis methods (vector autoregressive (VAR) / Granger causality analysis) (Haslbeck et al., 2021; Seth et al., 2015). For the VAR method, the optimal lag order was determined by incrementally increasing the lag order and selecting the model with the lowest information criterion (IC) value. Four commonly used ICs-Akaike Information Criterion (AIC), Hannan-Quinn Information Criterion (HQIC), Final Prediction Error (FPE), and Bayesian Information Criterion (BIC)-were computed (Cavanaugh & Neath, 2019; Jayakumar & Sulthan, 2013; Karimi, 2007; Neath & Cavanaugh, 2012). Since the lowest IC value varied across subjects, the data were modeled by considering all lag orders from the minimum to the maximum lag for group analysis. To compare the differences across methods, ANOVA and t-tests were conducted to examine the directionality, strength, and lag effects of EC, in comparison to existing physiological and anatomical knowledge of the visuomotor network. Monte Carlo simulations were employed to address multiple comparisons.
Results:
Differences in EC patterns were observed depending on the smoothing conditions, time series extraction methods, and lag order (Figs. 1 and 2). Specifically, significant differences in EC patterns were found between the most recent lag order (lag 1) and the optimal lag order (lag 9), with and without the smoothing process, within each time series extraction method (Figs. 1A and 2A). Furthermore, within both the unsmoothed and smoothed conditions, significant differences in EC patterns were identified across the three time series extraction methods for both lag order 1 and lag order 9 (Figs. 1B and 2B).
Conclusions:
Our findings emphasize that EC analysis is highly influenced by preprocessing choices, including smoothing, time series extraction methods, and lag order selection in VAR modeling. For the visuomotor network, the PC95% method produced the robust and reliable results, effectively compensating for ROI size differences and capturing broader neural variability. Smoothing was found to be undesirable, as it introduces unwanted signals from outside the ROI (Jo et al., 2007). Additionally, selecting the optimal lag order, rather than the most recent lag (lag 1), yielded more accurate EC patterns, especially for datasets with longer time series (e.g., > 200 time points). In summary, careful selection of preprocessing steps-avoiding smoothing, using the PC95% method, and determining the optimal lag order-significantly enhances the accuracy and reliability of EC analysis in brain networks.
Modeling and Analysis Methods:
Connectivity (eg. functional, effective, structural) 1
fMRI Connectivity and Network Modeling 2
Keywords:
Other - Effective connectivity (EC); Vector autoregressive (VAR) modeling; Smoothing; Time series extraction; Optimal lag order; Information criteria
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.
Cavanaugh, J. E., & Neath, A. A. (2019). The Akaike information criterion: Background, derivation, properties, application, interpretation, and refinements. Wiley Interdisciplinary Reviews: Computational Statistics, 11(3), e1460.
Cox, R. W. (1996). AFNI: software for analysis and visualization of functional magnetic resonance neuroimages. Computers and Biomedical research, 29(3), 162-173.
Fischl, B. (2012). FreeSurfer. Neuroimage, 62(2), 774-781.
Haslbeck, J. M., Bringmann, L. F., & Waldorp, L. J. (2021). A tutorial on estimating time-varying vector autoregressive models. Multivariate behavioral research, 56(1), 120-149.
Jayakumar, G. D. S., & Sulthan, A. (2013). HETEROSCEDASTICITY IN SURVEY DATA AND MODEL SELECTION BASED ON WEIGHTED HANNAN-QUINN INFORMATION CRITERION. Journal of Reliability and Statistical Studies, 17-40.
Jo, H. J., Lee, J.-M., Kim, J.-H., Shin, Y.-W., Kim, I.-Y., Kwon, J. S., & Kim, S. I. (2007). Spatial accuracy of fMRI activation influenced by volume-and surface-based spatial smoothing techniques. Neuroimage, 34(2), 550-564.
Karimi, M. (2007). A corrected FPE criterion for autoregressive processes. 2007 15th European Signal Processing Conference,
Neath, A. A., & Cavanaugh, J. E. (2012). The Bayesian information criterion: background, derivation, and applications. Wiley Interdisciplinary Reviews: Computational Statistics, 4(2), 199-203.
Seth, A. K., Barrett, A. B., & Barnett, L. (2015). Granger causality analysis in neuroscience and neuroimaging. Journal of Neuroscience, 35(8), 3293-3297.