Mesoscale Functional Architecture of the Human Hippocampus During Memory Recognition
Abstract number :
1.28
Submission category :
3. Neurophysiology / 3G. Computational Analysis & Modeling of EEG
Year :
2025
Submission ID :
175
Source :
www.aesnet.org
Presentation date :
12/6/2025 12:00:00 AM
Published date :
Authors :
Presenting Author: Weichen Huang, PhD – Stanford University
Josef Parvizi, MD, PhD – Stanford University School of Medicine
Rationale: The hippocampus (HPC) plays an important role in memory recognition, characterized by distinct electrophysiological signatures at the mesoscale level such as low-frequency activity (LFA) encompassing theta band, high-frequency activity (HFA) encompassing high-gamma, and ripple events. However, it remains unknown how these activities are generated by different subpopulation of neurons or distinct HPC subregions.
Methods: In this study, we analyzed data from 26 patients with intracranial electroencephalography (iEEG, 187 HPC sites, 94 sites in the left hemisphere, median value is 6 sites across patients) while performing a verbal old–new memory recognition task for abstract and concrete words presented minutes or a day before. Combining data from immediate and delayed recognition tests, and measuring changes in the aperiodic exponent during recognition trials, we identified three groups of recording sites based on the changes of aperiodic exponent compared to the pre-stimulus baseline.
Results: Those sites showing an increase in the aperiodic exponent (n = 79 electrodes from 18 patients) with significantly higher HFA power and ripple rates for hit trials compared to miss, correct rejection, and false alarm trials during both post-stimulus onset and pre-response periods. These sites also showed increased LFA power during the post-onset period but a decrease during the pre-response period for hit trials compared to other conditions. In contrast, electrodes with a decrease in exponent (n = 13 electrodes from 9 patients) showed no significant memory-related differences; however, they did exhibit a significant increase in ripple rate during the pre-response period regardless of memory condition. Electrodes with no significant change in exponent (n = 95 electrodes from 23 patients) showed partial memory-related effects across specific conditions, reflected in LFA power and ripple rates but not in HFA power. Anatomically, the three electrode groups were not restricted to specific hippocampal subregions but were instead distributed in a mosaic-like pattern throughout the hippocampus.
Conclusions: Our findings offer new insights into the mesoscale functional organization of the human hippocampus during memory recognition, indicating that distinct neural populations within the hippocampus play specialized roles while being interwoven and unevenly distributed across its structure.
Funding: The study was supported by research grants from the US National Institute of Neurological Disorders and Stroke (R01NS078396) to JP.
Neurophysiology