Non-invasive tracking of hippocampal theta oscillations

Abstract

Hippocampal theta oscillations play a crucial role in the formation of episodic memories by binding multisensory information into coherent episodes. Brain stimulation studies suggest that memory performance can be modulated by targeting these oscillations, offering a potential way of treating memory disorders. However, to effectively engage these rhythms, precise knowledge about their presence, frequency, phase and location is required, which has been a challenge, particularly for non-invasive methods such as magnetoencephalography (MEG). Here, we first asked if hippocampal signals are detectable in MEG recordings. Using simultaneous MEG-intracranial EEG (iEEG) data, we show that invasively recorded hippocampal activity is reflected in MEG sensor signals, demonstrating the feasibility of detecting deep-brain activity non-invasively. Building on this, we introduce a MEG-based analysis pipeline to track hippocampal theta frequency over time. Applied across three independent datasets, the pipeline captured characteristic hippocampal theta frequency patterns in both rodents and humans, and MEG-derived hippocampal frequencies were consistent with those observed simultaneously in intracranial EEG. These findings provide evidence that MEG can reliably track individual hippocampal oscillatory dynamics, paving the way for future non-invasive closed-loop interventions that adapt stimulation frequency and timing to ongoing oscillations.