Authors :
Presenting Author: Elizabeth Espinal, PhD – Medical University of South Carolina
Akash Mishra, BS – Northwell Health
Sarah Johnson, BS – Northwell Health
Evangelia Chrysikou, PhD – Drexel University
Noah Markowitz, MS – Northwell Health
Ashesh Mehta, MD PhD – Northwell Health
Stephan Bickel, MD PhD – Northwell Health
Rationale:
Memory consolidation during sleep depends on the coordination of hippocampal ripples, thalamocortical spindles, cortical high-frequency oscillations (CHFOs), and slow oscillations. While animal work has outlined mechanisms, their interaction with memory outcomes in humans remains unclear. Intracranial EEG (iEEG) provides high-resolution access to these dynamics but has rarely been applied in targeted memory reactivation (TMR). We tested whether auditory cues during sleep modulate ripple, spindle, and CHFO activity, and whether such modulations predict memory.Methods:
Seven epilepsy patients with implanted iEEG electrodes completed a two-day TMR paradigm. Participants learned auditory–object associations, and a subset of sounds was replayed during sleep. Memory was assessed with pre- and post-sleep spatial recall tests. Behavioral outcomes were compared for cued versus uncued items. Neural analyses focused on hippocampal ripple rates, CHFO responses in visually responsive versus nonvisual electrodes, spindle activity in the same sites, and ripple–spindle–CHFO coupling relative to cue onset. Cluster-based permutation tests and nonparametric statistics evaluated temporal effects.Results:
Behaviorally, recall accuracy did not differ for cued versus uncued items, t(341) = 0.97, p = .33, Cohen’s d = 0.10. Neural data revealed state- and region-specific cue effects. Good performers showed elevated hippocampal ripple rates –500 to –200 ms before cue onset (p < .01), suggesting a receptive “on-state.” CHFOs increased significantly in visually responsive electrodes after cues (p < .01) but not in nonvisual sites (p = .29). Spindles in visual cortex showed a trend toward pre-cue enhancement (min p = .020). Ripple–spindle lag distributions shifted significantly post-cue (all p < .001), with visual electrodes exhibiting tighter clustering around spindle peaks (z = 8.89, p < .001). Exploratory analyses showed post-cue CHFO–spindle coordination in visual cortex correlated moderately with memory benefit (ρ = .60), though nonsignificant.
Conclusions:
Although behavioral benefits of TMR were not significant at the group level, iEEG revealed temporally precise neural markers of reactivation. Elevated pre-cue ripple rates in good performers suggest an “on-state” priming the hippocampus for consolidation, while post-cue CHFO increases and altered ripple–spindle timing highlight cortical–hippocampal coordination as mechanisms of cue-induced reactivation. Visual electrodes showed stronger and more specific coupling effects than nonvisual sites, consistent with sensory-specific consolidation pathways. These findings support active systems consolidation models, identify pre-cue ripple elevation as a potential biomarker of receptive memory states, and underscore the importance of cue timing in shaping memory outcomes.
Funding: National Science Foundation, Graduate Research Fellowship Program (NSF-GRFP)
American Psychological Association (APA) Dissertation Award