Biology subjects

Sweeney-Reed, C. M.

Publications and source records attributed to Sweeney-Reed, C. M..

5 recordsLinked to original sources

bioRxiv2026

Dissociable thalamic oscillatory mechanisms support motor sequence learning

The ventrointermediate nucleus of the thalamus (VIM) is implicated in motor sequence learning, yet the underlying neural mechanisms remain unclear. We recorded intracranial activity from the human VIM during a serial reaction time task to determine how neural dynamics support learning. Participants responded faster during repeating than randomized sequences. Beta-to-low-gamma activity was greater during repeating sequences and elevated relative to the prestimulus baseline, consistent with emergence and stabilization of learned motor representations. In contrast, beta-phase modulation of high-frequency activity decreased progressively from rest to random to repeated sequence execution. Stronger phase-amplitude coupling was associated with faster responses, reaching significance in the random condition. These findings reveal a dissociation between power and cross-frequency coupling. Together they suggest that thalamic dynamics contribute to motor learning through multiple mechanisms: beta-band power reflects the emergence of learned motor representations, whereas beta-high-frequency coupling is enhanced when the context is less predictable.

neuroscience↗

bioRxiv2024

Anterior thalamus functional connectivity with cortical regions underpinning prospective memory: a 7-Tesla fMRI seed-based study

Prospective memory, or memory for future intentions, engages particular cortical regions. Lesion studies also implicate the thalamus, with prospective memory deterioration following thalamic stroke. Neuroimaging, anatomical, and lesion studies suggest the ANT (anterior nuclei of the thalamus) in particular are involved in episodic memory, with electrophysiological studies suggesting an active role in selecting neural assemblies underlying particular memory traces. Here we hypothesized that the ANT are engaged in realizing prospectively-encoded intentions, detectable using ultra-high-field strength functional MRI. Using a within-subject design, participants (N = 14; age 20-35 years) performed an ongoing n-back working memory task with two cognitive loads, each with and without a prospective memory component, during 7-Tesla functional MRI. Seed-to-voxel whole brain functional connectivity analyses were performed to establish whether including a prospective memory component in an ongoing task results in greater connectivity between ANT and cortical regions engaged in prospective memory. Repeated measures ANOVAs were applied to behavioral and connectivity measures, with the factors Task Type (with prospective memory or not) and N-Back (2-back or 3-back). Response accuracy was greater and reaction times faster without the prospective memory component, and accuracy was higher in the 2-than 3-back condition. Task Type had a main effect on connectivity with an ANT seed, with greater ANT-DLPFC (dorsolateral prefrontal cortex) and ANT-STG (superior temporal gyrus) connectivity when including a prospective memory component. Post hoc testing based on a significant interaction showed greater ANT-DLPFC connectivity (p-FWE = 0.007) when prospective memory was included with the low and ANT-STG connectivity (p-FWE = 0.019) with the high cognitive load ongoing task. Direct comparison showed greater functional connectivity between these areas and the ANT than DMNT (dorsomedial nucleus of the thalamus) during prospective remembering. Enhanced ANT-DLPFC connectivity, a brain region with an established role in strategic monitoring for prospective memory cues, arose with a low cognitive load ongoing task that enabled monitoring. This connectivity was significantly less on direct comparison with increasing the cognitive load of the ongoing task without prospective memory, suggesting specificity for prospective memory. Greater ANT-STG connectivity on prospective memory inclusion in the higher cognitive load ongoing task fits with reported STG activation on prospective memory through spontaneous retrieval. Lower connectivity on direct comparison with a DMNT seed suggests ANT specificity. The findings fit with a coordinating role for the ANT in prospective remembering. Given the small sample, these findings should be considered preliminary, with replication required.

neuroscience↗

bioRxiv2023

Cholinergic modulation of motor sequence learning

The cholinergic system plays a key role in motor function, but whether pharmacological modulation of cholinergic activity affects motor sequence learning is unknown. The acetylcholine receptor antagonist biperiden, an established treatment in movement disorders, reduces attentional modulation, but whether it influences motor sequence learning is not clear. Using a randomized, double-blind placebo-controlled crossover design, we tested thirty healthy young participants and show that biperiden impairs production of sequential finger movements following a fixed but not a random sequence. A similar interaction was observed in widespread oscillatory broadband power changes (4-25 Hz) in the motor sequence learning network after receiving biperiden, with greater power in the theta, alpha, and beta bands over ipsilateral motor and bilateral parietal-occipital areas. The reduced theta power during a fixed compared to random sequence, likely reflecting disengagement of top-down attention to sensory processes, was disrupted by biperiden. The alpha synchronization during learned sequences, reflecting sensory gating and lower visuospatial attention requirements for the learned, compared with visuomotor responses to a random sequence, was greater after biperiden, potentially reflecting excessive visuospatial attention reduction following biperiden, also affecting visuomotor responding required to enable sequence learning. Beta oscillations facilitate sequence learning by integrating visual and somatosensory inputs, stabilizing learned sequences, and promoting prediction of the next stimulus. The beta synchronization after biperiden fits with a disruption of the selective visuospatial attention enhancement associated with initial sequence learning. These findings highlight the role of cholinergic processes in motor sequence learning.

neuroscience↗

bioRxiv2023

Ventrointermediate thalamic stimulation improves motor learning in humans

Ventrointermediate thalamic stimulation (VIM-DBS) modulates oscillatory activity in a cortical network including primary motor cortex, premotor cortex, and parietal cortex. Here we show that, beyond the beneficial effects of VIM-DBS on motor execution, this form of invasive stimulation facilitates production of sequential finger movements that follow a fixed sequence but not when the sequence is selected at random. These results highlight the role of thalamo-cortical activity in motor learning.

neuroscience↗

bioRxiv2019

Convolutional Neural Networks for Decoding of Covert Attention Focus and Saliency Maps for EEG Feature Visualization

ObjectiveConvolutional neural networks (CNNs) have proven successful as function approximators and have therefore been used for classification problems including electroencephalography (EEG) signal decoding for brain-computer interfaces (BCI). Artificial neural networks, however, are considered black boxes, because they usually have thousands of parameters, making interpretation of their internal processes challenging. Here we systematically evaluate the use of CNNs for EEG signal decoding and investigate a method for visualizing the CNN model decision process.\n\nApproachWe developed a CNN model to decode the covert focus of attention from EEG event-related potentials during object selection. We compared the CNN and the commonly used linear discriminant analysis (LDA) classifier performance, applied to datasets with different dimensionality, and analyzed transfer learning capacity. Moreover, we validated the impact of single model components by systematically altering the model. Furthermore, we investigated the use of saliency maps as a tool for visualizing the spatial and temporal features driving the model output.\n\nMain resultsThe CNN model and the LDA classifier achieved comparable accuracy on the lower-dimensional dataset, but CNN exceeded LDA performance significantly on the higher-dimensional dataset (without hypothesis-driven preprocessing), achieving an average decoding accuracy of 90.7% (chance level = 8.3%). Parallel convolutions, tanh or ELU activation functions, and dropout regularization proved valuable for model performance, whereas the sequential convolutions, ReLU activation function, and batch normalization components, reduced accuracy or yielded no significant difference. Saliency maps revealed meaningful features, displaying the typical spatial distribution and latency of the P300 component expected during this task.\n\nSignificanceFollowing systematic evaluation, we provide recommendations for when and how to use CNN models in EEG decoding. Moreover, we propose a new approach for investigating the neural correlates of a cognitive task by training CNN models on raw high-dimensional EEG data and utilizing saliency maps for relevant feature extraction.

neuroscience↗