Causal Role of Theta and Alpha Oscillations in Output-gating

This study is a pilot, five-session study with transcranial magnetic stimulation (TMS), electroencephalography (EEG), and magnetic resonance imaging (MRI) to understand the neural oscillatory basis of output-gating. The first session of the experiment will be screening session, in which participants provide written consent to participate, screened for colorblindness, complete questionnaires, and perform a working memory task with retrospective cues. Participants will be invited back to the second session if they show a benefit to their working memory percent correct by use of the informative retro-cue relative to the uninformative neutral cue. This session will also be used to select the number of items that will be used in the working memory task for subsequent sessions. The criteria for difficulty titration is task performance between 60% and 85% correct for retro-cue trials and a benefit of at least 5% greater than neutral cue trials. Thus, different participants will perform the task with different numbers of items to be encoded into working memory. Titration of task difficulty as described here is critical for experiments that use causal manipulation (e.g. transcranial magnetic stimulation) to modulate performance. If participants are performing at ceiling (close to 100%) or at floor (close to random change), then any experimental manipulation of behavior is less likely to impact performance as the task is too "easy" or too "hard." For the second session of the experiment, participants perform the working memory retro-cue task while EEG is recorded. In addition, participants will complete a simple perception color task in which participants see a color and choose the matching color from the color circle. This task tests for the precision of perception throughout the color circle. The EEG data from the second session will be preprocessed and a Morlet wavelet convolution analysis will be conducted. The resulting spectrogram will be contrasted between the informative retro-cue and uninformative neutral cue to derive the theta frequency with peak amplitude in prefrontal cortex, and contrasted between a leftward and rightward retro-cue to derive the alpha frequency with peak amplitude in parietal cortex. These peak frequencies will be used for stimulation in the fourth and fifth session. In the third session, the investigators will acquire structural and functional MRI for each participant. The functional MRI data will be analyzed to identify regions in the anterior middle frontal gyrus and posterior intraparietal sulcus that are functionally connected within the frontal-parietal, "executive control," network. A previous meta-analysis of functional MRI studies found that the regions with peak retro-cue activity was at Montreal Neurological Institute coordinates (-40, 36, 28) for anterior middle frontal gyrus and (-38, -48, 44) for inferior intraparietal sulcus. Therefore, the investigators will constrain the search light to the anatomical landmarks and these coordinates. The center of mass in these regions will be used for targeting with TMS in the subsequent fourth and fifth sessions. In the fourth and fifth sessions, stimulation will be delivered at the timing relative to retro-cue, frequency, and spatial location based on previous localizers. During stimulation, the location of the TMS coil needs to be aligned to the targeted brain region with neuro-navigation software that records the accuracy of each TMS pulse relative to the target. On each trial, the stimulation type will be randomly selected, counter-balanced, and inter-mixed. The effects of rhythmic TMS are not expected to last for more than a few cycles beyond stimulation itself. Therefore, the experimental design randomly intermixes the stimulation type within every task block.

The study 'Theta Connectivity in Working Memory (STAR)', NCT05204381, investigated alpha and theta frequency neural oscillations in the context of cognitive control. Following study completion and during data analysis, a previously unrecognized error in the code of the randomization software used in the conduct of the trial was identified. As a function of the software error, group allocation did not occur as intended and the study design became, de facto and post hoc, a parallel arm design with participants receiving either alpha stimulation or theta stimulation in both stimulation sessions, as opposed to alpha stimulation then theta stimulation in a counterbalanced fashion. The outcomes reported herein of the resulting parallel arm design remain a scientifically valid and appropriate design to investigate the roles of alpha and theta oscillations in the context of cognitive control and additionally resulted in greater statistical power to detect a potential difference between groups.