This title appears in the Scientific Report :
2023
Please use the identifier:
http://dx.doi.org/10.34734/FZJ-2024-01235 in citations.
Impact of data processing on DCM estimates of effective connectivity from task-evoked fMRI
Impact of data processing on DCM estimates of effective connectivity from task-evoked fMRI
Introduction. Effective connectivity (EC) refers to directional or causal influences among interacting neuronal populations or brain regions and can be estimated from functional magnetic resonance imaging (fMRI) data via dynamic causal modeling (DCM) (Friston et al., 2003). However, in contrast to f...
Saved in:
Personal Name(s): | Zhang, Shufei |
---|---|
Jung, Kyesam / Langner, Robert / Florin, Esther / Eickhoff, Simon / Popovych, Oleksandr (Corresponding author) | |
Contributing Institute: |
Gehirn & Verhalten; INM-7 |
Imprint: |
2023
|
DOI: |
10.34734/FZJ-2024-01235 |
Conference: | The 29th Annual Meeting of the Organization for Human Brain Mapping, Montreal (Canada), 2023-07-22 - 2023-07-26 |
Document Type: |
Poster |
Research Program: |
Personalized Recommendations for Neurodegenerative Disease Human Brain Project Specific Grant Agreement 3 Human Brain Project Specific Grant Agreement 2 Multilevel Brain Organization and Variability Computational Principles |
Link: |
OpenAccess |
Publikationsportal JuSER |
Introduction. Effective connectivity (EC) refers to directional or causal influences among interacting neuronal populations or brain regions and can be estimated from functional magnetic resonance imaging (fMRI) data via dynamic causal modeling (DCM) (Friston et al., 2003). However, in contrast to functional connectivity, the impact of data processing varieties (Carp, 2012) on DCM estimates of task-evoked EC has hardly ever been systematically addressed. We therefore investigated how task-evoked EC is affected by reasonable choices made for processing task fMRI data. Methods. Task-evoked EC was investigated for a spatial stimulus-response compatibility (SRC) task (Fitts & Deininger, 1954) in 271 subjects (123 females, 18-85 years old, mean age: 52.6 ± 16.5 years) recruited from the subject pool of the 1000BRAINS project (Caspers et al., 2014). We considered the impact of the following data processing conditions on the modulatory component of task-evoked EC: Global signal regression (Almgren et al., 2020; Power et al., 2017), block vs. event-related general linear model (GLM) design (Daunizeau et al., 2011; Petersen & Dubis, 2012), type of activation task contrast (Zeidman, Jafarian, Corbin, et al., 2019), and significance thresholdingapproach (Roels et al., 2015). Using DCM designed in accordance with the considered parameters of the data processing, we estimated individual and group-averaged task-evoked EC within the SRC brain network of 9 nodes related to spatial conflict processing [Fig. 1]. Using the Parametric Empirical Bayes (PEB) analysis (Zeidman, Jafarian, Seghier, et al., 2019), we evaluated and compared the group-mean task-evoked EC patterns and between-group differencesin the task-evoked EC for any two of the considered conditions of the data processing (with vs. without GSR, event-related vs. block designs, corrected vs. uncorrected thresholding, and incompatible+compatible vs. incompatible contrasts). Results. We observed strongly varying patterns of the group-averaged EC depending on data processing choices. In particular, task-evoked EC was significantly impacted by GLM design (event-related or block) and type of activation contrast (incompatible task contrast vs. incompatible + compatible task contrast) [Fig. 2]. On the other hand, EC was little affected by globalsignal regression and the type of significance thresholding. The PEB analyses showed that more EC edges were significantly modulated by the task conditionsfor the event-related GLM than for the block one. Furthermore, the variation of the activation contrast induced more changes to the task-evoked EC for the block GLM than for the event-related one [Fig. 2].Conclusions. Our results demonstrate that different reasonable data processing choices can substantially alter the task-evoked EC as estimated by DCM. In particular, the event-related GLM design appears to be more responsive to task-evoked modulations of EC than the block design. On the other hand, the latter GLM design is more sensitive to the type of activation contrast than the event-related design. These choices should thus be made with care and, whenever possible, varied across parallel analyses to evaluate their impact and identify potential convergence for robust outcomes of the data analysis and neuroscientific interpretation of the estimated connectivity patterns. |