Neurophysiological Insights into Network Decoding of Working Memory: An Analysis of Human Brain Electrophysiological Signals

Purpose: The human brain is comprised of distinct regions, each contributing uniquely to behavioral control. The execution of even basic tasks necessitates synchronized activities among multiple brain regions. Fundamental cognitive functions hinge on the capacity to retain and flexibly manipulate information, a key role ascribed to Working Memory (WM). This study seeks to enhance our understanding of the neural mechanisms underlying WM and elucidate the coordinated neural activities spanning various brain regions. Materials and Methods: To achieve this objective, the invasively recorded electrophysiological activities from the Medial Temporal (MT) cortex of humans using a high number of electrodes were analyzed. The subjects did a verbal working memory task including three phases: encoding, maintenance, and retrieval. Phase synchronization between electrode signals in common frequency rhythms determined by Phase Locking Value (PLV) was used to create brain network graphs. Results: This study validates prior findings on neural synchronization in the hippocampus, entorhinal cortex, and amygdala during WM within the theta, alpha, and beta bands. Analysis of Phase Locking Value (PLV) dynamics during encoding and maintenance, reveals strong modulation in the theta, alpha, and beta rhythm. Notably, the PLV of theta oscillation between channels within the posterior hippocampal region was significantly reduced during maintenance. Conversely, PLV of theta-alpha rhythms between the Anterior Hippocampal region (AHL) and amygdala/entorhinal cortex was significantly increased by WM. Conclusion: This study, for the first time, demonstrates the networks involved in WM within MT areas in the human brain. These findings underscore the frequency-specific intricacies in WM modulation, providing valuable insights into neural coordination during specific processing stages.