Asynchronous Distributed Multielectrode Microstimulation Reduces Seizures in the Dorsal Tetanus Toxin Model of Temporal Lobe Epilepsy

SAFE-OPT: a Bayesian optimization algorithm for learning optimal deep brain stimulation parameters with safety constraints

Objective.To treat neurological and psychiatric diseases with deep brain stimulation (DBS), a trained clinician must select parameters for each patient by monitoring their symptoms and side-effects in a months-long trial-and-error process, delaying optimal clinical outcomes. Bayesian optimization has been proposed as an efficient method to quickly and automatically search for optimal parameters. However, conventional Bayesian optimization does not account for patient safety and could trigger unwanted or dangerous side-effects.Approach.In this study we develop SAFE-OPT, a Bayesian optimization algorithm designed to learn subject-specific safety constraints to avoid potentially harmful stimulation settings during optimization. We prototype and validate SAFE-OPT using a rodent multielectrode stimulation paradigm which causes subject-specific performance deficits in a spatial memory task. We first use data from an initial cohort of subjects to build a simulation where we design the best SAFE-OPT configuration for safe and accurate searchingin silico. Main results.We then deploy both SAFE-OPT and conventional Bayesian optimization without safety constraints in new subjectsin vivo, showing that SAFE-OPT can find an optimally high stimulation amplitude that does not harm task performance with comparable sample efficiency to Bayesian optimization and without selecting amplitude values that exceed the subject's safety threshold.Significance.The incorporation of safety constraints will provide a key step for adopting Bayesian optimization in real-world applications of DBS.

Effects of acute hippocampal stimulation in the nonhuman primate penicillin model of temporal lobe seizures

Asynchronous distributed multielectrode stimulation (ADMES) is a novel approach to deep brain stimulation for medication resistant temporal lobe epilepsy that has shown promise in rodent and in vitro seizure models. To further evaluate its effects on a pre-clinical model, we characterized the effect of unilateral ADMES in an NHP model of temporal lobe seizures induced by intra-hippocampal injection of penicillin (PCN). Four non-human primates were used for this study in two contemporaneous cohorts. One cohort (n = 3 hemispheres) was implanted with the Medtronic RC + S stimulation (GIN cohort) and recording system connected to two 4-contact ring electrodes to evaluate three unilateral stimulation patterns: 7 Hz Ring ADMES, 20 Hz Dual Ring, and 125 Hz Dual Ring (analog of clinical stimulation). In an additional cohort (EPC cohort, n = 2), two 12-contact segmented electrodes were implanted in the right hippocampus and connected to an externalized recording and stimulation system to allow more flexibility in the stimulation pattern. In this second cohort, 4 variations of stimulation were evaluated (7 Hz Full ADMES, 7 Hz Ring ADMES, 31 Hz Wide Ring, and 31 Hz Dual Ring). In the GIN cohort, we found an increase in seizure frequency and time spent in seizure during the 7 Hz Ring ADMES stimulation compared to the respective post-stimulation. A similar post-stimulation effect was found in the EPC cohort. We also found an increase in seizure frequency during the 7Hz full ADMES compared to the respective post-stimulation. However, we did not find a difference between pre-stimulation and stimulation conditions suggesting a possible post stimulation effect of the 7Hz hippocampal stimulation. In conclusion, in the NHP PCN model of temporal lobe seizures, acute asynchronous hippocampal stimulation was not therapeutic, however, our findings related to the post-stimulation effect can support future studies using hippocampal stimulation for the treatment of temporal lobe epilepsy.

Mimicking CA3 Temporal Dynamics Controls Limbic Ictogenesis

Mesial temporal lobe epilepsy (MTLE) is the most common partial complex epilepsy in adults and the most unresponsive to medications. Electrical deep brain stimulation (DBS) of the hippocampus has proved effective in controlling seizures in epileptic rodents and in drug-refractory MTLE patients. However, current DBS paradigms implement arbitrary fixed-frequency or patterned stimuli, disregarding the temporal profile of brain electrical activity. The latter, herein included hippocampal spontaneous firing, has been shown to follow lognormal temporal dynamics. Here, we present a novel paradigm to devise DBS protocols based on stimulation patterns fashioned as a surrogate brain signal. We focus on the interictal activity originating in the hippocampal subfield CA3, which has been shown to be anti-ictogenic. Using 4-aminopyridine-treated hippocampus-cortex slices coupled to microelectrode array, we pursue three specific aims: (1) address whether lognormal temporal dynamics can describe the CA3-driven interictal pattern, (2) explore the possibility of restoring the non-seizing state by mimicking the temporal dynamics of this anti-ictogenic pattern with electrical stimulation, and (3) compare the performance of the CA3-surrogate against periodic stimulation. We show that the CA3-driven interictal activity follows lognormal temporal dynamics. Further, electrical stimulation fashioned as a surrogate interictal pattern exhibits similar efficacy but uses less pulses than periodic stimulation. Our results support the possibility of mimicking the temporal dynamics of relevant brain signals as a straightforward DBS strategy to ameliorate drug-refractory epilepsy. Further, they herald a paradigm shift in neuromodulation, wherein a compromised brain signal can be recreated by the appropriate stimuli distribution to bypass trial-and-error studies and attain physiologically meaningful DBS operating modes.

Animal Models of Epilepsy: A Phenotype-oriented Review

Epilepsy is a serious neurological disorder characterized by abnormal, recurrent, and synchronous discharges in the brain. Long-term recurrent seizure attacks can cause serious damage to brain function, which is usually observed in patients with temporal lobe epilepsy. Controlling seizure attacks is vital for the treatment and prognosis of epilepsy. Animal models, such as the kindling model, which was the most widely used model in the past, allow the understanding of the potential epileptogenic mechanisms and selection of antiepileptic drugs. In recent years, various animal models of epilepsy have been established to mimic different seizure types, without clear merits and demerits. Accordingly, this review provides a summary of the views mentioned above, aiming to provide a reference for animal model selection.

Dependence of Working Memory on Coordinated Activity Across Brain Areas

Neural signatures of working memory (WM) have been reported in numerous brain areas, suggesting a distributed neural substrate for memory maintenance. In the current manuscript we provide an updated review of the literature focusing on intracranial neurophysiological recordings during WM in primates. Such signatures of WM include changes in firing rate or local oscillatory power within an area, along with measures of coordinated activity between areas based on synchronization between oscillations. In comparing the ability of various neural signatures in any brain area to predict behavioral performance, we observe that synchrony between areas is more frequently and robustly correlated with WM performance than any of the within-area neural signatures. We further review the evidence for alteration of inter-areal synchrony in brain disorders, consistent with an important role for such synchrony during behavior. Additionally, results of causal studies indicate that manipulating synchrony across areas is especially effective at influencing WM task performance. Each of these lines of research supports the critical role of inter-areal synchrony in WM. Finally, we propose a framework for interactions between prefrontal and sensory areas during WM, incorporating a range of experimental findings and offering an explanation for the observed link between intra-areal measures and WM performance.

Combination of Matching Responsive Stimulations of Hippocampus and Subiculum for Effective Seizure Suppression in Temporal Lobe Epilepsy

Responsive neural stimulation (RNS) is considered a promising neural modulation therapy for refractory epilepsy. Combined stimulation on different targets may hold great promise for improving the efficacy of seizure control since neural activity changed dynamically within associated brain targets in the epileptic network. Three major issues need to be further explored to achieve better efficacy of combined stimulation: (1) which nodes within the epileptogenic network should be chosen as stimulation targets? (2) What stimulus frequency should be delivered to different targets? and (3) Could the efficacy of RNS for seizure control be optimized by combined different stimulation targets together? In our current study, Granger causality (GC) method was applied to analyze epileptogenic networks for finding key targets of RNS. Single target stimulation (100 μA amplitude, 300 μs pulse width, 5s duration, biphasic, charge-balanced) with high frequency (130 Hz, HFS) or low frequency (5 Hz, LFS) was firstly delivered by our lab designed RNS systems to CA3, CA1, subiculum (SUB) of hippocampi, and anterior nucleus of thalamus (ANT). The efficacy of combined stimulation with different groups of frequencies was finally assessed to find out better combined key targets with optimal stimulus frequency. Our results showed that stimulation individually delivered to SUB and CA1 could shorten the average duration of seizures. Different stimulation frequencies impacted the efficacy of seizure control, as HFS delivered to CA1 and LFS delivered to SUB, respectively, were more effective for shortening the average duration of electrographic seizure in Sprague-Dawley rats (n = 3). Moreover, the synchronous stimulation of HFS in CA1 combined with LFS in SUB reduced the duration of discharge significantly in rats (n = 6). The combination of responsive stimulation at different targets may be an inspiration to optimize stimulation therapy for epilepsy.

The impact of synchronous versus asynchronous electrical stimulation in artificial vision

Visual prosthesis devices designed to restore sight to the blind have been under development in the laboratory for several decades. Clinical translation continues to be challenging, due in part to gaps in our understanding of critical parameters such as how phosphenes, the electrically-generated pixels of artificial vision, can be combined to form images. In this review we explore the effects that synchronous and asynchronous electrical stimulation across multiple electrodes have in evoking phosphenes. Understanding how electrical patterns influence phosphene generation to control object binding and perception of visual form is fundamental to creation of a clinically successful prosthesis.

An Overview of Studies Demonstrating that ex vivo Neuronal Networks Display Multiple Complex Behaviors: Emergent Properties of Nearest-Neighbor Interactions of Excitatory and Inhibitory Neurons

The responsiveness of the human nervous system ranges from the basic sensory interpretation and motor regulation to so-called higher-order functions such as emotion and consciousness. Aspects of higher-order functions are displayed by other mammals and birds. In efforts to understand how neuronal interaction can generate such a diverse functionality, murine embryonic cortical neurons were cultured on Petri dishes containing multi-electrode arrays that allowed recording and stimulation of neuronal activity. Despite the lack of major architectural features that govern nervous system development in situ, this overview of multiple studies demonstrated that these 2-dimensional ex vivo neuronal networks nevertheless recapitulate multiple key aspects of nervous system development and activity in situ, including density-dependent, the spontaneous establishment of a functional network that displayed complex signaling patterns, and responsiveness to environmental stimulation including generation of appropriate motor output and long-term potentiation. These findings underscore that the basic interplay of excitatory and inhibitory neuronal activity underlies all aspects of nervous system functionality. This reductionist system may be useful for further examination of neuronal function under developmental, homeostatic, and neurodegenerative conditions.

Toward Closed-Loop Electrical Stimulation of Neuronal Systems: A Review

Biological neuronal cells communicate using neurochemistry and electrical signals. The same phenomena also allow us to probe and manipulate neuronal systems and communicate with them. Neuronal system malfunctions cause a multitude of symptoms and functional deficiencies that can be assessed and sometimes alleviated by electrical stimulation. Our working hypothesis is that real-time closed-loop full-duplex measurement and stimulation paradigms can provide more in-depth insight into neuronal networks and enhance our capability to control diseases of the nervous system. In this study, we review extracellular electrical stimulation methods used in in vivo, in vitro, and in silico neuroscience research and in the clinic (excluding methods mainly aimed at neuronal growth and other similar effects) and highlight the potential of closed-loop measurement and stimulation systems. A multitude of electrical stimulation and measurement-based methods are widely used in research and the clinic. Closed-loop methods have been proposed, and some are used in the clinic. However, closed-loop systems utilizing more complex measurement analysis and adaptive stimulation systems, such as artificial intelligence systems connected to biological neuronal systems, do not yet exist. Our review promotes the research and development of intelligent paradigms aimed at meaningful communications between neuronal and information and communications technology systems, "dialogical paradigms," which have the potential to take neuroscience and clinical methods to a new level.

Neurostimulation stabilizes spiking neural networks by disrupting seizure-like oscillatory transitions

An improved understanding of the mechanisms underlying neuromodulatory approaches to mitigate seizure onset is needed to identify clinical targets for the treatment of epilepsy. Using a Wilson–Cowan-motivated network of inhibitory and excitatory populations, we examined the role played by intrinsic and extrinsic stimuli on the network’s predisposition to sudden transitions into oscillatory dynamics, similar to the transition to the seizure state. Our joint computational and mathematical analyses revealed that such stimuli, be they noisy or periodic in nature, exert a stabilizing influence on network responses, disrupting the development of such oscillations. Based on a combination of numerical simulations and mean-field analyses, our results suggest that high variance and/or high frequency stimulation waveforms can prevent multi-stability, a mathematical harbinger of sudden changes in network dynamics. By tuning the neurons’ responses to input, stimuli stabilize network dynamics away from these transitions. Furthermore, our research shows that such stabilization of neural activity occurs through a selective recruitment of inhibitory cells, providing a theoretical undergird for the known key role these cells play in both the healthy and diseased brain. Taken together, these findings provide new vistas on neuromodulatory approaches to stabilize neural microcircuit activity.

Optimizing neuromodulation based on surrogate neural states for seizure suppression in a rat temporal lobe epilepsy model

OBJECTIVE

Developing a new neuromodulation method for epilepsy treatment requires a large amount of time and resources to find effective stimulation parameters and often fails due to inter-subject variability in stimulation effect. As an alternative, we present a novel data-driven surrogate approach which can optimize the neuromodulation efficiently by investigating the stimulation effect on surrogate neural states.

APPROACH

Medial septum (MS) optogenetic stimulation was applied for modulating electrophysiological activities of the hippocampus in a rat temporal lobe epilepsy model. For the new approach, we implemented machine learning techniques to describe the pathological neural states and to optimize the stimulation parameters. Specifically, first, we found neural state surrogates to estimate a seizure susceptibility based on hippocampal local field potentials. Second, we modulated the neural state surrogates in a desired way with the subject-specific optimal stimulation parameters found by in vivo Bayesian optimization. Finally, we tested whether modulating the neural state surrogates affected seizure frequency.

MAIN RESULTS

We found two neural state surrogates: The first was hippocampal theta power by considering its well-known relationship with epilepsy, and the second was the output of pre-ictal state model (PriSM) which was built by characterizing the hippocampal activity during the pre-ictal period. The optimal stimulation parameters found by Bayesian optimization outperformed the other parameters in terms of modulating the surrogates toward anti-seizure neural state. When treatment efficacy was tested, the subject-specific optimal parameters for increasing theta power were more effective to suppress seizures than fixed stimulation parameter (7 Hz). However, modulation of the other neural state surrogate, PriSM, did not suppress seizures.

SIGNIFICANCE

The surrogate approach can save enormous time and resources to find subject-specific optimal stimulation parameters which can effectively modulate neural states and further improve therapeutic effectiveness. This approach can also be used for improving neuromodulation treatment of other neurological or psychiatric diseases.

Perspectives on the current developments with neuromodulation for the treatment of epilepsy

Introduction: As deep brain stimulation revolutionized the treatment of movement disorders in the late 80s, neuromodulation in the treatment of epilepsy will undoubtedly undergo transformative changes in the years to come with the exponential growth of technological development moving into mainstream practice; the appearance of companies such as Facebook, Google, Neuralink within the realm of brain-computer interfaces points to this trend.Areas covered: This perspective piece will talk about the history of brain stimulation in epilepsy, current-approved treatments, technical developments and the future of neurostimulation.Expert opinion: Further understanding of the brain alongside machine learning and innovative technology will be the future of neuromodulation for the treatment of epilepsy. All of these innovations and advances should pave the way toward overcoming the vexing underutilization of surgery in the therapeutic armamentarium against medically refractory seizures, given the implicit advantage of a neuromodulatory rather than neurodestructive approach.

Tissue Response to Neural Implants: The Use of Model Systems Toward New Design Solutions of Implantable Microelectrodes

The development of implantable neuroelectrodes is advancing rapidly as these tools are becoming increasingly ubiquitous in clinical practice, especially for the treatment of traumatic and neurodegenerative disorders. Electrodes have been exploited in a wide number of neural interface devices, such as deep brain stimulation, which is one of the most successful therapies with proven efficacy in the treatment of diseases like Parkinson or epilepsy. However, one of the main caveats related to the clinical application of electrodes is the nervous tissue response at the injury site, characterized by a cascade of inflammatory events, which culminate in chronic inflammation, and, in turn, result in the failure of the implant over extended periods of time. To overcome current limitations of the most widespread macroelectrode based systems, new design strategies and the development of innovative materials with superior biocompatibility characteristics are currently being investigated. This review describes the current state of the art of in vitro, ex vivo, and in vivo models available for the study of neural tissue response to implantable microelectrodes. We particularly highlight new models with increased complexity that closely mimic in vivo scenarios and that can serve as promising alternatives to animal studies for investigation of microelectrodes in neural tissues. Additionally, we also express our view on the impact of the progress in the field of neural tissue engineering on neural implant research.

A characterization of epileptogenesis presented in hippocampal neural activity in a rat tetanus toxin model

We built a regression model to describe the progress of epileptogenesis in a rat intrahippocampal tetanus toxin (TeNT) epilepsy model by identifying informative neural features from hippocampal local field potentials (LFPs). The LFPs were recorded from the awake and freely behaving animals during the latent period and the active-seizure period. Frequency domain neural features including power spectral density, coherence and phase coherence were calculated from the hippocampal LFPs. A least angle regression with elastic net regularization (LARS-ENR) model successfully predicted a relative day from the first seizure in multiple rats (R²_test = 0.724±0.025). By leveraging a characteristic of LARS-ENR which reduces unnecessary features, we identified the neural features related to epileptogenesis in a TeNT model.

Bayesian Optimization of Asynchronous Distributed Microelectrode Theta Stimulation and Spatial Memory

There is a great need for an electrical stimulation therapy to treat medication-resistant, surgically ineligible epileptic patients that successfully reduces seizure incidence with minimal side effects. Critical to advancing such therapies will be identifying the trade-offs between therapeutic efficacy and side effects. One novel treatment developed in the tetanus toxin rat model of mesial temporal lobe epilepsy, asynchronous distributed microelectrode stimulation (ADMETS) in the hippocampus has been shown to significantly reduce seizure frequency. However, our results have demonstrated that ADMETS has a negative effect on spatial memory that scales with the amplitude of stimulation. Given the high dimensional space of possible stimulation parameters, it is difficult to construct a mapping from variations in stimulation to behavioral effect. In this project, we present a novel, principled approach using closed-loop Bayesian optimization to tune stimulation that successfully maximize a desired objective - performance on a spatial memory assay.

Evolving Applications, Technological Challenges and Future Opportunities in Neuromodulation: Proceedings of the Fifth Annual Deep Brain Stimulation Think Tank

The annual Deep Brain Stimulation (DBS) Think Tank provides a focal opportunity for a multidisciplinary ensemble of experts in the field of neuromodulation to discuss advancements and forthcoming opportunities and challenges in the field. The proceedings of the fifth Think Tank summarize progress in neuromodulation neurotechnology and techniques for the treatment of a range of neuropsychiatric conditions including Parkinson's disease, dystonia, essential tremor, Tourette syndrome, obsessive compulsive disorder, epilepsy and cognitive, and motor disorders. Each section of this overview of the meeting provides insight to the critical elements of discussion, current challenges, and identified future directions of scientific and technological development and application. The report addresses key issues in developing, and emphasizes major innovations that have occurred during the past year. Specifically, this year's meeting focused on technical developments in DBS, design considerations for DBS electrodes, improved sensors, neuronal signal processing, advancements in development and uses of responsive DBS (closed-loop systems), updates on National Institutes of Health and DARPA DBS programs of the BRAIN initiative, and neuroethical and policy issues arising in and from DBS research and applications in practice.

The role of the basal ganglia in the control of seizure

Epilepsy is a network disorder and each type of seizure involves distinct cortical and subcortical network, differently implicated in the control and propagation of the ictal activity. The role of the basal ganglia has been revealed in several cases of focal and generalized seizures. Here, we review the data that show the implication of the basal ganglia in absence, temporal lobe, and neocortical seizures in animal models (rodent, cat, and non-human primate) and in human. Based on these results and the advancement of deep brain stimulation for Parkinson's disease, basal ganglia neuromodulation has been tested with some success that can be equally seen as promising or disappointing. The effect of deep brain stimulation can be considered promising with a 76% in seizure reduction in temporal lobe epilepsy patients, but also disappointing, since only few patients have become seizure free and the antiepileptic effects have been highly variable among patients. This variability could probably be explained by the heterogeneity among the patients included in these clinical studies. To illustrate the importance of specific network identification, electrophysiological activity of the putamen and caudate nucleus has been recorded during penicillin-induced pre-frontal and motor seizures in one monkey. While an increase of the firing rate was found in putamen and caudate nucleus during pre-frontal seizures, only the activity of the putamen cells was increased during motor seizures. These preliminary results demonstrate the implication of the basal ganglia in two types of neocortical seizures and the necessity of studying the network to identify the important nodes implicated in the propagation and control of each type of seizure.

A machine learning approach to characterizing the effect of asynchronous distributed electrical stimulation on hippocampal neural dynamics in vivo

Asynchronous distributed microelectrode theta stimulation (ADMETS) of the hippocampus has been shown to reduce seizure frequency in the tetanus toxin rat model of mesial temporal lobe epilepsy suggesting a hypothesis that ADMETS induces a seizure resistant state. Here we present a machine learning approach to characterize the nature of neural state changes induced by distributed stimulation. We applied the stimulation to two animals under sham and ADMETS conditions and used a combination of machine learning techniques on intra-hippocampal recordings of Local Field Potentials (LFPs) to characterize the difference in the neural state between sham and ADMETS. By iteratively fitting a logistic regression with data from the inter-stimulation interval under sham and ADMETS condition we found that the classification performance improves for both animals until 90s post stimulation before leveling out at AUC of 0.64 ± 0.2 and 0.67 ± 0.02 when all inter-stimulation data is included. The models for each animal were re-fit using elastic net regularization to force many of the model coefficients to 0, identifying those that do not optimally contribute to the classifier performance. We found that there is significant variation in the non-zero coefficients between animals (p <; 0.01), suggesting that the ADMETS induced state is represented differently between subject. These findings lay the foundation for using machine learning to robustly and quantitatively characterize neural state.

The influence of the pre-stimulation neural state on the post-stimulation neural dynamics via distributed microstimulation of the hippocampus

In this study we investigated how the neural state influences how the brain responds to electrical stimulation using a 16-channel microelectrode array with 8 stimulation and recording channels implanted in the rat hippocampus. In two experiments we identified the stimulation threshold at which the brain changes to an afterdischarge state. In one experiment a range of suprathreshold stimulations were applied, and in another the stimulation was not changed. The neural state was measured by the power spectral density prior to stimulation. In the first experiment, these measures and the stimulation parameters were used as features, either together or separately, for training a Support Vector Machine (SVM) classifier to predict whether the stimulation would produce an afterdischarge. In the second experiment, recursive feature elimination was used to iteratively remove the neural state features from the recording channels that had the least impact on the overall accuracy. In the first experiment 43 stimulations elicited 26 afterdischarges. In predicting the post-stimulation state-change (afterdischarge vs. no afterdischarge) the feature space of only neural state had a higher accuracy (67.4%) than when combined with the stimulation parameters (65.1%) or the stimulation parameters alone (58.1%). The overall classification results from both feature spaces containing the neural state were non-independent (chi-squared p <; 0.01). In the second experiment, the channels that were the least predictive were those on the more distal ends of the recording electrode, and the most predictive were clustered in the center of the electrode. Additionally, the accuracy increased when 4 channels were removed. The findings from these experiments suggest that both the pre-stimulation state and the spatial properties from where it is measured can play a role in how neural stimulation can induce functional changes in the hippocampal networks.