Testing network properties of episodic memory using non-invasive brain stimulation

Abstract

Episodic memory depends on the hippocampus and its coordination with a distributed network of interconnected brain areas. Recent findings indicate that the function of this network can be altered using network-targeted transcranial magnetic stimulation (TMS). These stimulation experiments have identified increases in episodic memory and the network-wide coordination it requires. Network-target stimulation differs from the dominant framework for TMS experiments, in which stimulation has been considered as a focal virtual lesion. We offer a conceptual framework for important distinctions between network-wide and focal effects of stimulation on episodic memory and discuss factors that may influence the quality and quantity of stimulation effects. Findings from these experiments indicate that many properties of episodic memory can be effectively studied at the network level via noninvasive stimulation.

Keywords: Episodic memory; brain stimulation; fmri; hippocampus; recollection.

Conflict of interest statement

Conflict of interest The authors declare no conflict of interest.

Figures

Figure 1.. Effects of TMS on the episodic memory network

Episodic memory is supported by a network of regions which communicate and are thought to be coordinated via the hippocampus (HPC). Key regions include medial prefrontal cortex (mPFC), posterior cingulate cortex (PCC), precuneus (PCu), angular gyrus (AnG), retrosplenial cortex (RSc), and parahippocampal cortex (PHC), as well as others listed in the main text (not shown). (A) Stimulating one region within this network (lightning bolt) can influence its discrete contribution to memory. These local effects are typically disruptive and short-lived (up to 1 hr). In this example, stimulating angular gyrus leads to decreased local activity (red circle) and disruption of memory functions that depend on this area, such as vividness and confidence. (B) Stimulation of the same cortical region may also lead to network-wide effects, particularly when using parameters such as multiple-session stimulation that produce more distributed and/or longer-lasting effects. Network-wide effects involve relatively long-lasting (>24 hours) enhanced functional connectivity among network regions (red lines and arrows between network regions), and improvements in episodic memory functions that require coordination of hippocampus with its network. Local and network-wide effects may occur at the same time and may interact with one another. Certain experimental conditions may influence whether TMS leads to local or network-wide effects, or a combination of the two, as described in the main text.

Figure 2.. Experimental procedure for multi-day stimulation targeting the hippocampal network.

(A) The procedure used to define a TMS target involves assessing resting-state fMRI connectivity between the hippocampus and lateral parietal cortex (red arrow). Subject-specific left lateral parietal stimulation locations (blue activations, circled) are selected based on functional connectivity with anatomically defined hippocampal seeds (red activations). (B) Experimental design for multi-day stimulation studies. Each study involves one week of active stimulation (stim week) and one week of a control condition (control week), the order of which is counterbalanced. Active stimulation involves lateral parietal rTMS, while control conditions differ among studies, sometimes consisting of sham stimulation (reduced intensity) [34], other times of matched-intensity stimulation to separate out-of-network locations [9,29], and a combination of these control conditions for other studies [8,10,12,30]. At the first baseline session prior to the first stimulation session, resting state fMRI is measured to determine TMS target, and TMS motor threshold is assessed to determine TMS intensity. Baseline sessions for both weeks of stimulation involve both episodic memory and neuroimaging assessments (resting state fMRI, task-based fMRI, or EEG). Five consecutive days of stimulation or a control condition are administered following baseline assessments. Post treatment assessments occur approximately 24 hours after the final stimulation session, and again involve episodic memory and neuroimaging assessments.