Spatiotemporal Analysis of Verbal Working Memory With 31-Channel EEG: Comparative Insights From 63-Channel EEG

This study investigates the spatiotemporal dynamics and involvement of subcortical regions, such as the hippocampus, amygdala, and thalamus during verbal working memory (vWM) stages. These areas are key for memory and early indicators of neurodegenerative diseases and aging. While previous research primarily utilized magnetoencephalography (MEG), intracranial electroencephalogram (iEEG), or fMRI, this study employs a surface electroencephalography (EEG) system to investigate subcortical dynamics during vWM. In particular, standardized low-resolution brain electromagnetic topography (sLORETA)-based brain source localization (BSL) is utilized to study temporal information of the target regions of interest (ROIs) during vWM stages in healthy subjects. The spatiotemporal data from sLORETA is further analyzed to understand the performance metrics of each ROI in vWM stage classification. A preliminary comparison of 31- and 63-channel EEG systems is presented for studying the variation in spatiotemporal information of subcortical regions. The 31-channel system, despite being the low-density configuration, effectively captures subcortical dynamics, particularly in the hippocampus. It is to be noted that the activation in subcortical regions begins around 180-200 ms after task onset. The left hippocampus is active during encoding and retrieval, while the right hippocampus shows broader activation before probe responses. Such activation dynamics and load processing studies are possible owing to the high-temporal-resolution feature of EEG. Additionally, time-frequency analysis is presented for load processing that provides insights into memory encoding and recall correlation. It is observed that the subcortical spatiotemporal information improves the accuracy of the vWM stage classification. The current study of subcortical investigation using limited sensors finds application in neurodegenerative disease stage estimation where high-density EEG or fMRI faces patient tolerability constraints.