(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



Why does invasive brain stimulation sometimes improve memory and sometimes impair it? [1]

['Uma R. Mohan', 'Surgical Neurology Branch', 'Ninds', 'National Institutes Of Health', 'Bethesda', 'Maryland', 'United States Of America', 'Joshua Jacobs', 'Department Of Biomedical Engineering', 'Columbia University']

Date: 2024-10

Episodic memory is a critical aspect of everyday life that allows us to recall events, make decisions, and navigate our environments. It is estimated that 1 in 25 Americans suffer from episodic memory loss from Alzheimer ‘s disease, dementia, or following a traumatic brain injury. Clinicians and researchers have recently tried to help supplement or repair memory in individuals with episodic memory disorders using electrical brain stimulation. However, although there is evidence of invasive brain stimulation improving memory, studies also show that it can have variable effects, and sometimes cause memory impairment. In this Essay, we investigate these differences and try to explain this variability. We propose that key aspects of the memory network are heterogeneous across individuals, requiring customized stimulation protocols to create successful brain stimulation systems for memory enhancement. We provide a plan for applying brain stimulation while overcoming this variability by using closed-loop stimulation protocols that are customized according to a person’s own anatomical structure, functional connectivity, and memory-related electrophysiological patterns.

This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.

In this Essay, we discuss why the behavioral and physiological effects of invasive brain stimulation are so variable across studies and individuals, explaining how fixed approaches for memory enhancement produce inconsistent results. In this context, we explain why a more flexible approach that accommodates complex and heterogenous brain patterns of individuals is necessary for consistent memory improvement. Finally, we describe a framework for tailoring customized stimulation for memory enhancement based on each person’s individually mapped physiology, with a focus on the hippocampal–cortical network.

One approach that can be employed is to stimulate in a way that would recreate or supplement the brain’s natural memory-related signals in order to improve the brain’s own natural memory encoding or retrieval state. In this approach, stimulation is often targeted at the hippocampus, a critical structure for memory. Individuals are stimulated with a fixed set of stimulation pulses designed to broadly excite hippocampal neural activity [ 23 – 27 , 30 , 33 ] or reinstate the hippocampus’ own endogenous 4- to 8-Hz theta rhythm. Because theta oscillations are linked to memory encoding and synaptic plasticity ( Table 1 ) [ 37 , 38 , 42 ], researchers hypothesized that theta-enhancing stimulation would improve memory accuracy. Although this method seems logical, the effects on memory performance from hippocampal stimulation are mixed, perhaps reflecting our lack of understanding of how memories are encoded.

( a ) Different stimulation locations, frequencies, amplitudes, durations, and pulse widths have been shown to have varying effects on neural activity and in turn on memory function. When developing a stimulation protocol designed to modulate a specific memory-related neural signal, it is critical to first determine how the combination of these parameters will alter electrophysiology [ 37 , 42 – 52 ]. ( b ) Separate from the parameters that determine how current is delivered, the timing, or state in which an individual is in, impacts the effect of stimulation on memory. These can be considered in 2 different ways: timing with respect to the targeted memory process determined by the structure of the memory task [ 22 , 25 – 30 , 32 , 36 , 39 , 30 , 41 , 42 ], and timing from moment-to-moment, i.e., whether an individual’s neural signals show that they are in a good or bad memory processing state [ 36 , 53 ].

Previous studies on the use of brain stimulation for memory enhancement generally used fixed stimulation protocols across individuals. These studies produced wide-ranging outcomes, with some reporting impaired memory performance from stimulation [ 22 – 31 ], and others showing enhancement [ 30 , 32 – 42 ] ( Table 1 ). Across these studies, electrical brain stimulation was applied with various ranges of parameters, with substantial differences in location, frequency, duration, amplitude, and timing ( Fig 1 ). As well as technical challenges, a significant scientific challenge is that we do not yet have a complete characterization of the neural and electrophysiological correlates of memory. Researchers have identified a number of different electrophysiological signals that correlate with memory encoding, such as theta-band oscillations, which are potential targets for enhancement with stimulation. However, we do not know which electrophysiological signals are most directly and causally relevant for forming new memories. This uncertainty regarding the neural basis of human memory encoding makes it hard to precisely design stimulation protocols that are optimized to drive memory enhancement.

Direct electrical brain stimulation is an effective treatment for a number of neurological and behavioral disorders, including Parkinson ‘s disease, essential tremor, dystonia, and epileptic seizures [ 1 – 9 ]. Building on this success, in the past decades researchers have expanded the scope of invasive brain stimulation, using it to help patients with a much broader range of neuropsychiatric and cognitive disorders, including major depressive disorder [ 10 , 11 ], obsessive compulsive disorder [ 12 ], anorexia nervosa [ 13 ], addiction [ 14 , 15 ], schizophrenia [ 16 , 17 ], and memory disorders such as Alzheimer’s disease [ 18 , 19 ]. However, unlike the earlier success of deep brain stimulation for motor disorders, these efforts at using brain stimulation to treat neuropsychiatric and cognitive disorders have produced inconsistent effects [ 2 , 3 , 20 , 21 ]. But why is this the case?

Why are the effects of invasive brain stimulation so variable?

Given the importance of memory for everyday life, researchers have tried a number of approaches to apply electrical stimulation to enhance memory. Here, we focus on direct electrical stimulation with surgically implanted electrodes. Researchers have conducted studies with various types of stimulation, from continuous stimulation with direct currents to bursts of charge-balanced pulse trains with different parameters. Studies have also tested the effects of current applied at different times relative to behavior and across a range of target areas including both gray and white matter [37,42]. These different approaches and parameters have widely varying effects and need to be understood if we are to use invasive brain stimulation to effectively improve memory (Table 1).

How does the precise location of stimulation affect its impact on memory? Memory is a complex process that involves a broad network of brain regions. Many studies have attempted to improve human memory by applying stimulation to different elements of the memory network [54], with the aim of enhancing synaptic plasticity and memory formation; however, a wide range of behavioral effects have been observed. In general, the behavioral effect of applying electrical stimulation to a region often corresponds with the broader functional role of that region (Fig 1A). For instance, stimulation in sensory regions induces perceptual phosphenes [55], the sensation of light without light actually entering the eye. However, the effects of stimulation even in a particular brain region can vary dramatically according to the specific positioning of the stimulating electrode(s). In some cases, stimulation at nearby sites within a region can produce highly variable, even opposite [56–59], behavioral effects. For instance, when direct electrical stimulation was applied to a specific cortical location, a patient spontaneously recalled memories from high school; however, when stimulation was applied at all neighboring locations it did not evoke high school-related memories [59]. To explain why the effects of invasive brain stimulation are so spatially specific, it is helpful to consider multi-scale interactions between the network and local physiology because stimulating different combinations of neuron populations and pathways can have complex network-level effects beyond the stimulation site [60]. The importance of the precise stimulation location is most evident from a set of recent memory-modulation studies that measured the positioning of individual stimulation electrodes with respect to white matter pathways (Fig 1A). White matter in the brain consists of bundles of axons that create structural pathways for communication between brain regions [61]. Due to the interconnection of regions by white matter tracts, stimulating nearby specific bundles can have particularly strong impacts. Two studies reported greater improvement in memory performance from stimulating in or close to the white matter pathways of the medial temporal lobe (MTL) rather than in gray matter (Table 1) [37,62]. This trend of different stimulation effects in white matter versus gray matter were also observed in the lateral temporal cortex [63]. These findings could be explained by the enhanced ability of white matter stimulation to modulate both local and distant neural activity [43,51,63,64], thus allowing for the regulation of broader networks that drive cognitive states [44–47,65,66]. Further, when stimulating gray matter directly, currents can impact neighboring cell bodies, and the effects are likely to be inhibitory and spatially limited compared to white matter [52,67,68]. Thus, to improve targeting of white matter with electrical stimulation, recent modeling work shows promise for selecting stimulation locations based on individual patient tractography [69,70]. Despite some apparent successes, merely stimulating the white matter elements of the memory network does not always lead to memory enhancement. For example, 2 studies [27,32] both targeted the white matter of the hippocampus and entorhinal cortex with 50-Hz stimulation and reported opposite effects on memory. This variability highlights the fact that effective stimulation for memory enhancement requires more than simply identifying a particular anatomical target. Therefore, even when applying stimulation in an optimal location, one may need to consider other aspects of the stimulation parameters with respect to task design, the type of memory process, and ongoing neural dynamics, because the wrong type of stimulation can be disruptive, even when applied even to the right target location (Table 1 and Fig 1B).

Which stimulation parameters are most effective for memory modulation? Electrical stimulation can be applied to the brain with countless combinations of parameters, including different frequencies, amplitudes, and pulse-burst rates (Table 1 and Fig 1A). Identifying the optimal stimulation parameters is critical. The 2 stimulation parameters that often have the strongest effects on underlying neural activity are the frequency and amplitude. Generally, applying low-frequency stimulation inhibits local neural activity, while high-frequency stimulation (100 Hz and above) can both excite or inhibit neural activity depending on the target region (Fig 2A) [43,51,71,72]. In contrast to frequency, where effects can be positive or negative, the effect of amplitude is generally simpler. At a given stimulation site, increasing amplitude increases the magnitude of the effect that occurred from lower-amplitude stimulation (Fig 2B [51,64,72]. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 2. Percent of recording electrodes excited or inhibited by stimulation at different frequencies, amplitudes, and types of electrodes. (a) Percent of recording electrodes showing significant high frequency activity increases and decreases for each stimulation frequency for depth and surface stimulation sites. Adapted from [43]. (b) Increase in inhibitory and excitatory effects determined by spiking rate with stimulation amplitude. Adapted from [50]. https://doi.org/10.1371/journal.pbio.3002894.g002 Beyond these general trends, there is additional variability across regions in the effect of stimulation frequencies that differ in relation to the functional role of the local circuit. In some brain regions, stimulation at particular frequencies can improve memory by up-regulating specific local processes. There is evidence of memory improvement from stimulation at higher frequencies [43], bursts of stimulation at 4- to 8-Hz theta frequencies [37,42,73], and lower amplitudes applied in deep structures. In particular, key hippocampal signals, such as ripple oscillations and 4- to 8-Hz theta rhythms, are vital for synaptic plasticity and memory in encoding, consolidation, and retrieval of information in spatial navigation, working memory, and long-term memory formation [74–77]. There is evidence that bursts of stimulation at theta frequencies can improve memory, likely by entraining and enhancing the brain’s own theta-band oscillations [78,79]. Meanwhile, 200-Hz stimulation in the temporal cortex has been used to target the spectral tilt in the power spectrum, which represents the prevalence of high and low frequency components in a neural signal. This biomarker, which is associated with memory-related increases in high-frequency activity, is thought to represent synaptic excitation/inhibition balance [80–83]. The ability of these contrasting approaches to elicit comparable memory enhancements demonstrates that there are multiple distinct neurophysiological signatures of memory processes that form effective stimulation targets [76]. An important challenge going forward is to identify the optimal neural signal to modulate in cases where there are multiple potential stimulation targets at different frequencies. In addition to frequency and amplitude, the pulse width and waveform also are important factors in controlling how stimulation affects the brain (Fig 1A) [84,85]. Most memory modulation studies have applied charge-balanced, biphasic, rectangular pulses (Table 1). These rectangle-shaped voltages patterns, especially when delivered continuously for many seconds, are quite different from the brain’s native electrical patterns. Although research is limited, there is reason to think that alternative stimulation waveforms, such as sinusoids [67], may be helpful because the currents may more closely align with endogenous electrophysiology. An additional useful approach may be varying the width of stimulation pulses, as varying pulse widths may elicit the ability to recruit different types and sizes of neurons [86]. In recent years, researchers have identified a number of new, important electrophysiological patterns that may be useful targets for memory enhancement in humans, such as sharp-wave ripples and traveling waves of memory-related oscillations [87]. Up-regulating these signals with naturalistic stimulation may be more effective for memory improvement. Along these lines, one study experimented with stimulation that more closely resembled actual electrophysiological rhythms, with sine-wave-shaped stimulation delivered in-phase and anti-phase inducing memory enhancement and impairment, respectively [33]. A more recent study demonstrated impairment of memory retrieval by targeting separate nodes of the memory retrieval network with precisely timed theta-burst stimulation [29]. The specific phase lag that impaired memory the most varied depending on the stimulation target nodes of the network, emphasizing the importance of precisely timing stimulation to match the brain’s intrinsic signals. Thus, the efficacy of stimulation for cognitive enhancement may depend on finding the optimal combination of waveform shape, timing, frequency, amplitude, pulse width, and temporal patterning for a particular target brain region. Together, these studies open pathways for future work, such as designing stimulation waveforms to modulate endogenous neural signals and determining how to best deliver stimulation at multiple locations simultaneously.

[END]
---
[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3002894

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.