Systematically investigating sustained activity estimates of cognitive control processes in mixed blocked/event-related fMRI designs

Poster Session C - Sunday, April 14, 2024, 5:00 – 7:00 pm EDT, Sheraton Hall ABC

Rebecca Feldman¹, Joset A. Etzel¹, Todd S. Braver¹; ¹Washington University in St. Louis

Mixed -block/event-related fMRI designs have unique advantages for studying the neural mechanisms of cognition, in that they enable separation of sustained activation signals (i.e., persisting across trials) from those that are transient (i.e., event-related). Prior work has suggested that sustained activation patterns reflect the presence of cognitive control processes (e.g., task-set / goal maintenance) in fronto-parietal and cingulo-opercular brain networks. Nevertheless, the mixed design has been infrequently deployed in cognitive neuroscience studies, due its additional design and analysis constraints. Consequently, psychometric and reliability assessments of sustained activation estimates (i.e., compared with event-related) are lacking. Here, we conducted a systematic investigation of this type, capitalizing on data from the Dual Mechanisms of Cognitive Control (DMCC) project, a large-sample (N>100) fMRI study, in which mixed designs were utilized to estimate brain activity while each participant performed a battery of 4 different cognitive control tasks (AX-CPT, Cued Task-Switching, Stroop, Sternberg WM) in each of 3 conditions (Baseline, Proactive, Reactive). Extending and confirming claims made in the prior work, effect sizes were found to be small (~0.2) for sustained activation and were significantly less than event-related effects. Nevertheless, sustained activation patterns were consistent across tasks and strongest in cingulo-opercular and prefrontal regions, replicating prior findings. Individual differences in sustained activity were not highly reliable across tasks within these regions, but cross-task correlations were similar in magnitude to those for event-related activity. Additional analyses of sustained activity, focusing on test-retest reliability and twin-similarity are warranted and possible, by capitalizing on other unique DMCC dataset features.

Topic Area: METHODS: Neuroimaging