Cue switching but not task switching drives fronto-parietal activity in cued task switching paradigm
Poster No:
930
Submission Type:
Abstract Submission
Authors:
Patrick Bissett1, Sunjae Shim1, Jaime Ali Rios1, Jeanette Mumford2, Russell Poldrack1
Institutions:
1Stanford University, Stanford, CA, 2Stanford University, Madison, WI
First Author:
Co-Author(s):
Introduction:
Task switching incurs behavioral costs, and in the lab it is commonly measured with the cued task switch task, which involves cueing the subject to one of two tasks (e.g., magnitude judgment vs. odd-even judgment, see Figure 1), followed by a speeded response to a probe. Task switch cost is commonly ascribed to reconfiguring a task set, which is an "appropriate configuration of mental resources" (Monsell, 2003). A theoretical alternative is that the process of encoding a new cue drives the task switch cost, perhaps from cue repetition benefits (Logan & Bundesen, 2003).
Task-set reconfiguration can be distinguished from cue-related theories by introducing two cues per task (e.g., magnitude & high/low, odd/even & parity), which allows separation of the cost of switching tasks (measured by task switch - cue switch) from switching cues (measured by cue switch - cue stay). Task set reconfiguration theories predicting large task switch cost with little or no cue switch cost, and cue-related theories predict large cue switch cost with little or no task switch cost. When separated behaviorally, cue switch costs are often significant and large, and sometimes there are no additional task switch costs (Logan & Bundesen, 2003; Mayr & Kliegl, 2003).
Task switch fMRI studies commonly find a fronto-parietal network (Ruge et al., 2013), but the literature is dominated by tasks with 1 cue per task that cannot distinguish cue-switch and task switch activity. In the few fMRI studies with 2 cues per task, fronto-parietal activity was found in the task switch contrast but not the cue switch contrast (De Baene & Brass, 2011), consistent with this activity instantiating task set reconfiguration. However, this work did not account for RT differences between conditions (Mumford et al., 2023), which may be driving spurious contrast differences.
In two independent datasets with two cues per task, while accounting for reaction time confounds, we evaluate whether the fronto-parietal fMRI network observed in task switching is responsive to cue switching or task switching.
Task-set reconfiguration can be distinguished from cue-related theories by introducing two cues per task (e.g., magnitude & high/low, odd/even & parity), which allows separation of the cost of switching tasks (measured by task switch - cue switch) from switching cues (measured by cue switch - cue stay). Task set reconfiguration theories predicting large task switch cost with little or no cue switch cost, and cue-related theories predict large cue switch cost with little or no task switch cost. When separated behaviorally, cue switch costs are often significant and large, and sometimes there are no additional task switch costs (Logan & Bundesen, 2003; Mayr & Kliegl, 2003).
Task switch fMRI studies commonly find a fronto-parietal network (Ruge et al., 2013), but the literature is dominated by tasks with 1 cue per task that cannot distinguish cue-switch and task switch activity. In the few fMRI studies with 2 cues per task, fronto-parietal activity was found in the task switch contrast but not the cue switch contrast (De Baene & Brass, 2011), consistent with this activity instantiating task set reconfiguration. However, this work did not account for RT differences between conditions (Mumford et al., 2023), which may be driving spurious contrast differences.
In two independent datasets with two cues per task, while accounting for reaction time confounds, we evaluate whether the fronto-parietal fMRI network observed in task switching is responsive to cue switching or task switching.
Methods:
Dataset 1 (Bissett et al., 2023): 88 fMRI participants completed cued task switching with multiband 8, TR = 0.68s, and 2.2mm iso voxels.
Dataset 2: 18 fMRI participants each completed 5 sessions of cued task switching with multi-band 3, 3 echoes, TR = 1.49s, and 2.8mm iso voxels.
Data were quality assured using MRIQC (Esteban et al., 2017) and pre-processed using Tedana (only Dataset 2; DuPre, Salo et al., 2021) and fMRIPrep (Esteban et al., 2019).
First-level models were built that coded for the key conditions and reaction time (Mumford et al., 2023). For Dataset 2, fixed effects maps were created averaging across all 5 sessions. Statistical maps were created with Randomise TFCE values above 0.95 (equivalent to p < 0.05) with 5000 permutations. Conjunction maps are the binarized, voxel-wise intersection of the positive voxels in the TFCE maps.
Dataset 2: 18 fMRI participants each completed 5 sessions of cued task switching with multi-band 3, 3 echoes, TR = 1.49s, and 2.8mm iso voxels.
Data were quality assured using MRIQC (Esteban et al., 2017) and pre-processed using Tedana (only Dataset 2; DuPre, Salo et al., 2021) and fMRIPrep (Esteban et al., 2019).
First-level models were built that coded for the key conditions and reaction time (Mumford et al., 2023). For Dataset 2, fixed effects maps were created averaging across all 5 sessions. Statistical maps were created with Randomise TFCE values above 0.95 (equivalent to p < 0.05) with 5000 permutations. Conjunction maps are the binarized, voxel-wise intersection of the positive voxels in the TFCE maps.
Results:
We observed significant RT cue switch costs in dataset 1 (M=38ms) and dataset 2 (M=63ms) as well as task switch costs in dataset 1 (M = 25ms) and dataset 2 (M = 22ms) (Four one-sample t-tests vs. 0 each had p's<.001). Paired-sample, two-tailed t-tests showed that cue switch costs were significantly larger than task switch costs in dataset 2 (p<.001) but not dataset 1 (p=.19).
In our conjunction maps, we found left lateralized dlPFC and parietal activity in the cue switch contrast (Figure 2) but no activity in task switch contrast (not shown).
In our conjunction maps, we found left lateralized dlPFC and parietal activity in the cue switch contrast (Figure 2) but no activity in task switch contrast (not shown).
Conclusions:
We show the commonly observed fronto-parietal network in cued task switching is not the result of switching tasks, but is instead capturing a more general process of encoding a new cue. This challenges the widespread linking proposition between frontal-parietal activity in task switching and the endogenous act of control of "task set reconfiguration", and is more consistent with cue-repetition benefits (Logan & Bundesen, 2003), perhaps from stimulus priming.
Higher Cognitive Functions:
Executive Function, Cognitive Control and Decision Making 1
Modeling and Analysis Methods:
Activation (eg. BOLD task-fMRI) 2
Keywords:
Cognition
FUNCTIONAL MRI
Open Data
Univariate
Other - cognitive control; task switching; cue switching; RT modeling; conjunction analyses
1|2Indicates the priority used for review
Provide references using author date format
De Baene, W., & Brass, M. (2011). Cue-switch effects do not rely on the same neural systems as task-switch effects. Cognitive, Affective, and Behavioral Neuroscience, 11, 600-607.
Du Pre, E., Salo, T., Ahmed, Z., Bandettini, P. A., Bottenhorn, K. L., Caballero-Gauden, C., …., & Handwerker, D. A. (2021). TE-dependent analysis of multi-echo fMRI with tedana. The Journal of Open Source Software, 6(66), 3669.
Esteban, O., Birman, D., Schaer, M., Koyejo, O. O., Poldrack, R. A., & Gorgolewski, K. J. (2017). MRIQC: Advancing the automatic prediction of image quality in MRI from unseen sites. PloS one, 12, e0184661.
Esteban O., Markiewicz C. J., Blair R. W., Moodie C. A., Isik A. I., Erramuzpe A., … Gorgolewski K. J. (2019). fMRIPrep: a robust preprocessing pipeline for functional MRI. Nature Methods, 16, 111-116.
Logan, G. D., & Bundesen, C. (2003). Clever homunculus: Is there an endogenous act of control in the explicit task-cuing procedure? Journal of Experimental Psychology: Human Perception and Performance, 29(3), 575–599.
Mayr, U., & Kliegl, R. (2003). Differential effects of cue changes and task changes on task-set selection costs. Journal of Experimental Psychology: Learning, Memory, and Cognition, 29(3), 362-372.
Monsell, S. (2003). Task switching. Trends in Cognitive Sciences, 7(3), 134-140.
Mumford, J. A., Bissett, P. G., Jones, H. M., Shim, S., Rios, J. A. H., & Poldrack, R. A. (2023). The response time paradox in functional magnetic resonance imaging analyses. Nature Human Behaviour,
Ruge, H., Jamadar, S., Zimmerman, U., & Karayanidis, F. (2013). The many faces of preparatory control in task switching: Reviewing a decade of fMRI research. Human Brain Mapping, 34, 12-35.