Functionally distinct high and low theta oscillations in the human hippocampus.

Based on rodent models, researchers have theorized that the hippocampus supports episodic memory and navigation via the theta oscillation, a ~4-10 Hz rhythm that coordinates brain-wide neural activity. However, recordings from humans have indicated that hippocampal theta oscillations are lower in frequency and less prevalent than in rodents, suggesting interspecies differences in theta's function. To characterize human hippocampal theta, we examine the properties of theta oscillations throughout the anterior-posterior length of the hippocampus as neurosurgical subjects performed a virtual spatial navigation task. During virtual movement, we observe hippocampal oscillations at multiple frequencies from 2 to 14 Hz. The posterior hippocampus prominently displays oscillations at ~8-Hz and the precise frequency of these oscillations correlates with the speed of movement, implicating these signals in spatial navigation. We also observe slower ~3 Hz oscillations, but these signals are more prevalent in the anterior hippocampus and their frequency does not vary with movement speed. Our results converge with recent findings to suggest an updated view of human hippocampal electrophysiology. Rather than one hippocampal theta oscillation with a single general role, high- and low-frequency theta oscillations, respectively, may reflect spatial and non-spatial cognitive processes.

Memory retrieval modulates spatial tuning of single neurons in the human entorhinal cortex.

The medial temporal lobe is critical for both spatial navigation and memory. Although single neurons in the medial temporal lobe activate to represent locations in the environment during navigation, how this spatial tuning relates to memory for events involving those locations remains unclear. We examined memory-related changes in spatial tuning by recording single-neuron activity from neurosurgical patients performing a virtual-reality object-location memory task. We identified 'memory-trace cells' with activity that was spatially tuned to the retrieved location of the specific object that participants were cued to remember. Memory-trace cells in the entorhinal cortex, in particular, encoded discriminable representations of different memories through a memory-specific rate code. These findings indicate that single neurons in the human entorhinal cortex change their spatial tuning to target relevant memories for retrieval.

Stimulation of the Posterior Cingulate Cortex Impairs Episodic Memory Encoding.

Neuroimaging experiments implicate the posterior cingulate cortex (PCC) in episodic memory processing, making it a potential target for responsive neuromodulation strategies outside of the hippocampal network. However, causal evidence for the role that PCC plays in memory encoding is lacking. In human female and male participants (N = 17) undergoing seizure mapping, we investigated functional properties of the PCC using deep brain stimulation (DBS) and stereotactic electroencephalography. We used a verbal free recall paradigm in which the PCC was stimulated during presentation of half of the study lists, whereas no stimulation was applied during presentation of the remaining lists. We investigated whether stimulation affected memory and modulated hippocampal activity. Results revealed four main findings. First, stimulation during episodic memory encoding impaired subsequent free recall, predominantly for items presented early in the study lists. Second, PCC stimulation increased hippocampal gamma-band power. Third, stimulation-induced hippocampal gamma power predicted the magnitude of memory impairment. Fourth, functional connectivity between the hippocampus and PCC predicted the strength of the stimulation effect on memory. Our findings offer causal evidence implicating the PCC in episodic memory encoding. Importantly, the results indicate that stimulation targeted outside of the temporal lobe can modulate hippocampal activity and impact behavior. Furthermore, measures of connectivity between brain regions within a functional network can be informative in predicting behavioral effects of stimulation. Our findings have significant implications for developing therapies to treat memory disorders and cognitive impairment using DBS.SIGNIFICANCE STATEMENT Cognitive impairment and memory loss are critical public health challenges. Deep brain stimulation (DBS) is a promising tool for developing strategies to ameliorate memory disorders by targeting brain regions involved in mnemonic processing. Using DBS, our study sheds light on the lesser-known role of the posterior cingulate cortex (PCC) in memory encoding. Stimulating the PCC during encoding impairs subsequent recall memory. The degree of impairment is predicted by stimulation-induced hippocampal gamma oscillations and functional connectivity between PCC and hippocampus. Our findings provide the first causal evidence implicating PCC in memory encoding and highlight the PCC as a favorable target for neuromodulation strategies using a priori connectivity measures to predict stimulation effects. This has significant implications for developing therapies for memory diseases.

Gamma oscillations during episodic memory processing provide evidence for functional specialization in the longitudinal axis of the human hippocampus.

The question of whether the anterior and posterior hippocampus serve different or complementary functional roles during episodic memory processing has been motivated by noteworthy findings in rodent experiments and from noninvasive studies in humans. Researchers have synthesized these data to postulate several models of functional specialization, However, the issue has not been explored in detail using direct brain recordings. We recently published evidence that theta power increases during episodic memory encoding occur in the posterior hippocampus in humans. In our current investigation we analyzed an expanded data set of 32 epilepsy patients undergoing stereo EEG seizure mapping surgery with electrodes precisely targeted to the anterior and posterior hippocampus simultaneously who performed an episodic memory task. Using a repeated measures design, we looked for an interaction between encoding versus retrieval differences in gamma oscillatory power and anterior versus posterior hippocampal location. Our findings are consistent with a recently articulated model (the HERNET model) favoring posterior hippocampal activation during retrieval related processing. We also tested for encoding versus retrieval differences in the preferred gamma frequency band (high versus low gamma oscillations) motivated by published rodent data.

Comparison of logistic regression, support vector machines, and deep learning classifiers for predicting memory encoding success using human intracranial EEG recordings.

OBJECTIVE: We sought to test the performance of three strategies for binary classification (logistic regression, support vector machines, and deep learning) for the problem of predicting successful episodic memory encoding using direct brain recordings obtained from human stereo EEG subjects. We also sought to test the impact of applying t-distributed stochastic neighbor embedding (tSNE) for unsupervised dimensionality reduction, as well as testing the effect of reducing input features to a core set of memory relevant brain areas. This work builds upon published efforts to develop a closed-loop stimulation device to improve memory performance. APPROACH: We used a unique data set consisting of 30 stereo EEG patients with electrodes implanted into a core set of five common brain regions (along with other areas) who performed the free recall episodic memory task as brain activity was recorded. Using three different machine learning strategies, we trained classifiers to predict successful versus unsuccessful memory encoding and compared the difference in classifier performance (as measured by the AUC) at the subject level and in aggregate across modalities. We report the impact of feature reduction on the classifiers, including reducing the number of input brain regions, frequency bands, and the impact of tSNE. RESULTS: Deep learning classifiers outperformed both support vector machines (SVM) and logistic regression (LR). A priori selection of core brain regions also improved classifier performance for LR and SVM models, especially when combined with tSNE. SIGNIFICANCE: We report for the first time a direct comparison among traditional and deep learning methods of binary classification to the problem of predicting successful memory encoding using human brain electrophysiological data. Our findings will inform the design of brain machine interface devices to affect memory processing.

Theta band power increases in the posterior hippocampus predict successful episodic memory encoding in humans.

Functional differences in the anterior and posterior hippocampus during episodic memory processing have not been examined in human electrophysiological data. This is in spite of strong evidence for such differences in rodent data, including greater place cell specificity in the dorsal hippocampus, greater sensitivity to the aversive or motivational content of memories in ventral regions, connectivity analyses identifying preferential ventral hippocampal connections with the amygdala, and gene expression analyses identifying a dorsal-ventral gradient. We asked if memory-related oscillatory patterns observed in human hippocampal recordings, including the gamma band and slow-theta (2.5-5 Hz) subsequent memory effects, would exhibit differences along the longitudinal axis and between hemispheres. We took advantage of a new dataset of stereo electroencephalography patients with simultaneous, robotically targeted anterior, and posterior hippocampal electrodes to directly compare oscillatory subsequent memory effects during item encoding. This same data set allowed us to examine left-right connectivity and hemispheric differences in hippocampal oscillatory patterns. Our data suggest that a power increase during successful item encoding in the 2.5-5 Hz slow-theta frequency range preferentially occurs in the posterior hippocampus during the first 1,000 ms after item presentation, while a gamma band power increase is stronger in the dominant hemisphere. This dominant-nondominant pattern in the gamma range appears to reverse during item retrieval, however. Intrahippocampal phase coherence was found to be stronger during successful item encoding. Our phase coherence data are also consistent with existing reports of a traveling wave for theta oscillations propagating along the septotemporal (longitudinal) axis of the human hippocampus. We examine how our findings fit with theories of functional specialization along the hippocampal axis.

Similar patterns of neural activity predict memory function during encoding and retrieval.

Neural networks that span the medial temporal lobe (MTL), prefrontal cortex, and posterior cortical regions are essential to episodic memory function in humans. Encoding and retrieval are supported by the engagement of both distinct neural pathways across the cortex and common structures within the medial temporal lobes. However, the degree to which memory performance can be determined by neural processing that is common to encoding and retrieval remains to be determined. To identify neural signatures of successful memory function, we administered a delayed free-recall task to 187 neurosurgical patients implanted with subdural or intraparenchymal depth electrodes. We developed multivariate classifiers to identify patterns of spectral power across the brain that independently predicted successful episodic encoding and retrieval. During encoding and retrieval, patterns of increased high frequency activity in prefrontal, MTL, and inferior parietal cortices, accompanied by widespread decreases in low frequency power across the brain predicted successful memory function. Using a cross-decoding approach, we demonstrate the ability to predict memory function across distinct phases of the free-recall task. Furthermore, we demonstrate that classifiers that combine information from both encoding and retrieval states can outperform task-independent models. These findings suggest that the engagement of a core memory network during either encoding or retrieval shapes the ability to remember the past, despite distinct neural interactions that facilitate encoding and retrieval.