A computational outlook on neurostimulation

Protein kinase A: A potential marker of sympathovagal imbalance in heart failure

Mitigation of cardiac autonomic dysregulation by neuromodulation technologies is emerging as a new therapeutic modality of heart failure (HF). This recent progress has necessitated the identification of a biomarker for the quantification of sympathovagal balance, the potential target of 'neuromodulation' strategies. The currently available autonomic nervous system assessment parameters do not truly reflect the sympathovagal balance of the ventricle. Protein kinase A (PKA) is an intracellular enzyme that plays a major role in the pathophysiology of functional and structural ventricular remodeling in HF. Interestingly, sympathetic and parasympathetic activations exert reciprocal influence on the activity of PKA. The current review attempts to evaluate the potential concept and feasibility of using in vitro assessment of PKA activity as a marker of sympathovagal balance in HF.

A systematic review of computational models for the design of spinal cord stimulation therapies: from neural circuits to patient-specific simulations

Seventy years ago, Hodgkin and Huxley published the first mathematical model to describe action potential generation, laying the foundation for modern computational neuroscience. Since then, the field has evolved enormously, with studies spanning from basic neuroscience to clinical applications for neuromodulation. Computer models of neuromodulation have evolved in complexity and personalization, advancing clinical practice and novel neurostimulation therapies, such as spinal cord stimulation. Spinal cord stimulation is a therapy widely used to treat chronic pain, with rapidly expanding indications, such as restoring motor function. In general, simulations contributed dramatically to improve lead designs, stimulation configurations, waveform parameters and programming procedures and provided insight into potential mechanisms of action of electrical stimulation. Although the implementation of neural models are relentlessly increasing in number and complexity, it is reasonable to ask whether this observed increase in complexity is necessary for improved accuracy and, ultimately, for clinical efficacy. With this aim, we performed a systematic literature review and a qualitative meta-synthesis of the evolution of computational models, with a focus on complexity, personalization and the use of medical imaging to capture realistic anatomy. Our review showed that increased model complexity and personalization improved both mechanistic and translational studies. More specifically, the use of medical imaging enabled the development of patient-specific models that can help to transform clinical practice in spinal cord stimulation. Finally, we combined our results to provide clear guidelines for standardization and expansion of computational models for spinal cord stimulation.

Design of an adaptable intrafascicular electrode (AIR) for selective nerve stimulation by model-based optimization

Peripheral nerve stimulation is being investigated as a therapeutic tool in several clinical scenarios. However, the adopted devices have restricted ability to obtain desired outcomes with tolerable off-target effects. Recent promising solutions are not yet employed in clinical practice due to complex required surgeries, lack of long-term stability, and implant invasiveness. Here, we aimed to design a neural interface to address these issues, specifically dimensioned for pudendal and sacral nerves to potentially target sexual, bladder, or bowel dysfunctions. We designed the adaptable intrafascicular radial electrode (AIR) through realistic computational models. They account for detailed human anatomy, inhomogeneous anisotropic conductance, following the trajectories of axons along curving and branching fascicles, and detailed biophysics of axons. The model was validated against available experimental data. Thanks to computationally efficient geometry-based selectivity estimations we informed the electrode design, optimizing its dimensions to obtain the highest selectivity while maintaining low invasiveness. We then compared the AIR with state-of-the-art electrodes, namely InterStim leads, multipolar cuffs and transversal intrafascicular multichannel electrodes (TIME). AIR, comprising a flexible substrate, surface active sites, and radially inserted intrafascicular needles, is designed to be implanted in a few standard steps, potentially enabling fast implants. It holds potential for repeatable stimulation outcomes thanks to its radial structural symmetry. When compared in-silico, AIR consistently outperformed cuff electrodes and InterStim leads in terms of recruitment threshold and stimulation selectivity. AIR performed similarly or better than a TIME, with quantified less invasiveness. Finally, we showed how AIR can adapt to different nerve sizes and varying shapes while maintaining high selectivity. The AIR electrode shows the potential to fill a clinical need for an effective peripheral nerve interface. Its high predicted performance in all the identified requirements was enabled by a model-based approach, readily applicable for the optimization of electrode parameters in any peripheral nerve stimulation scenario.

Autonomous optimization of neuroprosthetic stimulation parameters that drive the motor cortex and spinal cord outputs in rats and monkeys

Neural stimulation can alleviate paralysis and sensory deficits. Novel high-density neural interfaces can enable refined and multipronged neurostimulation interventions. To achieve this, it is essential to develop algorithmic frameworks capable of handling optimization in large parameter spaces. Here, we leveraged an algorithmic class, Gaussian-process (GP)-based Bayesian optimization (BO), to solve this problem. We show that GP-BO efficiently explores the neurostimulation space, outperforming other search strategies after testing only a fraction of the possible combinations. Through a series of real-time multi-dimensional neurostimulation experiments, we demonstrate optimization across diverse biological targets (brain, spinal cord), animal models (rats, non-human primates), in healthy subjects, and in neuroprosthetic intervention after injury, for both immediate and continual learning over multiple sessions. GP-BO can embed and improve "prior" expert/clinical knowledge to dramatically enhance its performance. These results advocate for broader establishment of learning agents as structural elements of neuroprosthetic design, enabling personalization and maximization of therapeutic effectiveness.

From disabled tourists to impaired cyborg tourists: What would it take to transform?

Despite the media reports and governments promoting tourism as a fundamental right for everyone, traveling is still not accessible for disabled people. This study has highlighted the need to make tourists with disabilities accessable for inaccessible destinations. Cyborg products in the form of technological implants can make tourists with disabilities accessable for inaccessible destinations. Since tourists with a mobility disability (TMD) will be one of the primary beneficiaries of technological implants, little is known about their acceptance of technological implants during traveling.Therefore, this study assesses the willingness of TMD to use technological implants during traveling through a qualitative research approach. The results from thematic analysis identified two main themes: the use of assistive devices during traveling with four sub-themes (freedom of traveling, physical and attitudinal barriers, cost, and additional assistance and battery issues), and drivers of impaired cyborg tourists with seven sub-themes (independence, improved well-being, convenience/ease of use, social inclusion, positive emotions, motivation, and other issues). The study contributed to the literature by introducing drivers of impaired cyborg tourists along with previously identified concepts. The results also provided implications for the stakeholders of the tourism industry.

Toward a cerebello-thalamo-cortical computational model of spinocerebellar ataxia

Computational neural network modelling is an emerging approach for optimization of drug treatment of neurological disorders and fine-tuning of rehabilitation strategies. In the current study, we constructed a cerebello-thalamo-cortical computational neural network model to simulate a mouse model of cerebellar ataxia (pcd^5J mice) by manipulating cerebellar bursts through reduction of GABAergic inhibitory input. Cerebellar output neurons were projected to the thalamus and bidirectionally connected with the cortical network. Our results showed that reduction of inhibitory input in the cerebellum orchestrated the cortical local field potential (LFP) dynamics to generate specific motor outputs of oscillations of the theta, alpha, and beta bands in the computational model as well as in mouse motor cortical neurons. The therapeutic potential of deep brain stimulation (DBS) was tested in the computational model by increasing the sensory input to restore cortical output. Ataxia mice showed normalization of the motor cortex LFP after cerebellum DBS. We provide a novel approach to computational modelling to investigate the effect of DBS by mimicking cerebellar ataxia involving degeneration of Purkinje cells. Simulated neural activity coincides with findings from neural recordings of ataxia mice. Our computational model could thus represent cerebellar pathologies and provide insight into how to improve disease symptoms by restoring neuronal electrophysiological properties using DBS.

Neuroanatomical mapping of the lumbosacral spinal cord in individuals with chronic spinal cord injury

With emerging applications of spinal cord electrical stimulation in restoring autonomic and motor function after spinal cord injury, understanding the neuroanatomical substrates of the human spinal cord after spinal cord injury using neuroimaging techniques can play a critical role in optimizing the outcomes of these stimulation-based interventions. In this study, we have introduced a neuroimaging acquisition and analysis protocol of the spinal cord in order to identify: (i) spinal cord levels at the lumbosacral enlargement using nerve root tracing; (ii) variability in the neuroanatomical characteristics of the spinal cord among individuals; (iii) location of the epidural stimulation paddle electrode and contacts with respect to the spinal cord levels at lumbosacral enlargement; and (iv) the links between the anatomical levels of stimulation and the corresponding neurophysiological motor responses. Twelve individuals with chronic, motor complete spinal cord injury implanted with a spinal cord epidural stimulator were included in the study (age: 34 ± 10.9 years, sex: 10 males, 2 females, time since injury: 8.2 ± 9.9 years, American Spinal Injury Association Impairment Scale: 6 A, 6 B). High-resolution MRI scans of the spinal cord were recorded pre-implant. An analysis of neuroanatomical substrates indicates that the length of the spinal column and spinal cord, location of the conus tip and the relationship between the spinal cord levels and vertebral levels, particularly at the lumbosacral enlargement, are variable across individuals. There is no statistically significant correlation between the length of the spinal column and the length of the spinal cord. The percentage of volumetric coverage of the lumbosacral spinal cord by the epidural stimulation paddle electrode ranges from 33.4 to 90.4% across participants. The location of the spinal cord levels with respect to the electrode contacts varies across individuals and impacts the recruitment patterns of neurophysiological responses. Finally, MRI-based spinal cord modelling can be used as a guide for the prediction and preplanning of optimum epidural stimulation paddle placement prior to the implant surgery to ensure maximizing functional outcomes. These findings highlight the crucial role that the neuroanatomical characteristics of the spinal cord specific to each individual play in achieving maximum functional benefits with spinal cord electrical stimulation.

A computational roadmap to electronic drugs

A growing number of complex neurostimulation strategies promise symptom relief and functional recovery for several neurological, psychiatric, and even multi-organ disorders. Although pharmacological interventions are currently the mainstay of treatment, neurostimulation offers a potentially effective and safe alternative, capable of providing rapid adjustment to short-term variation and long-term decline of physiological functions. However, rapid advances made by clinical studies have often preceded the fundamental understanding of mechanisms underlying the interactions between stimulation and the nervous system. In turn, therapy design and verification are largely driven by clinical-empirical evidence. Even with titanic efforts and budgets, it is infeasible to comprehensively explore the multi-dimensional optimization space of neurostimulation through empirical research alone, especially since anatomical structures and thus outcomes vary dramatically between patients. Instead, we believe that the future of neurostimulation strongly depends on personalizable computational tools, i.e. Digital Neuro Twins (DNTs) to efficiently identify effective and safe stimulation parameters. DNTs have the potential to accelerate scientific discovery and hypothesis-driven engineering, and aid as a critical regulatory and clinical decision support tool. We outline here how DNTs will pave the way toward effective, cost-, time-, and risk-limited electronic drugs with a broad application bandwidth.

Neuromodulation devices for heart failure

Autonomic imbalance with a sympathetic dominance is acknowledged to be a critical determinant of the pathophysiology of chronic heart failure with reduced ejection fraction (HFrEF), regardless of the etiology. Consequently, therapeutic interventions directly targeting the cardiac autonomic nervous system, generally referred to as neuromodulation strategies, have gained increasing interest and have been intensively studied at both the pre-clinical level and the clinical level. This review will focus on device-based neuromodulation in the setting of HFrEF. It will first provide some general principles about electrical neuromodulation and discuss specifically the complex issue of dose-response with this therapeutic approach. The paper will thereafter summarize the rationale, the pre-clinical and the clinical data, as well as the future prospectives of the three most studied form of device-based neuromodulation in HFrEF. These include cervical vagal nerve stimulation (cVNS), baroreflex activation therapy (BAT), and spinal cord stimulation (SCS). BAT has been approved by the Food and Drug Administration for use in patients with HfrEF, while the other two approaches are still considered investigational; VNS is currently being investigated in a large phase III Study.

An Open-source Computational Model of Neurostimulation of the Spinal Pudendo-Vesical Reflex for the Recovery of Bladder Control After Spinal Cord Injury

Spinal cord stimulation (SCS) could be used to restore control of the bladder after spinal cord injury, but substantial development is still required to tailor this technology for bladder function. Computational models could be utilized to accelerate these efforts enabling in-silico optimization of stimulation parameters. However, no model of the spinal pudendo-vesical reflex can simulate the effect of stimulation amplitude on neuron recruitment. This limitation hinders accurate prediction of bladder pressure changes for different stimulation configurations. Here., we implemented an open-source realistic spiking neural network model of the pudendo-vesical reflex enabling exploration of the impact of stimulation amplitude and frequency on bladder pressure changes. We used the o2S2 PARC platform to design a parallel implementation of the bladder reflex circuits with NEURON. Our model successfully reproduced and expanded previous studies., producing a decrease in bladder pressure at low stimulation frequency (10 Hz) and excitation at high stimulation frequency (≥33 Hz) in isovolumetric experiments. We then explored the effect of mixed nerve recruitment., simulating a common case of poorly selective spinal cord stimulation. We found that high recruitments of pudendal nerve axons are necessary to maintain this bi-modal behavior., regardless of stimulation specificity. Our framework is fully open-source and can be used to simulate any type of axon stimulations such as SCS and peripheral nerve stimulation.

Optimizing sensory fiber activation during cervical transcutaneous spinal stimulation using different electrode configurations: A computational analysis

BACKGROUND

Cervical transcutaneous spinal cord stimulation (tSCS) is a rehabilitation tool which has been used to promote upper-limb motor recovery after spinal cord injury. Importantly, optimizing sensory fiber activation at specific spinal segments could enable activity-dependent neuromodulation during rehabilitation.

METHODS

An anatomically realistic cervical tSCS computational model was used to analyze the activation of α-motor and Aα-sensory fibers at C7 and C8 spinal segments using nine cathode electrode configurations. Specifically, the cathode was simulated at three vertebral level positions: C6, C7, and T1; and in three sizes: 5.0 x 5.0, 3.5 x 3.5; and 2.5 x 2.5 cm² , while the anode was on the anterior neck. Finite element method was used to estimate the electric potential distribution along α-motor and Aα-sensory fibers, and computational models were applied to simulate the fiber membrane dynamics during tSCS. The minimum stimulation intensity necessary to activate the fibers (activation threshold) was estimated and compared across cathode configurations in an effort to optimize sensory fiber activation.

RESULTS

Our results showed that nerve fibers at both C7 and C8 spinal segments were recruited at lower stimulation intensities when the cathode was positioned over the C7 or T1 vertebra compared with the C6 position. Sensory fibers were activated at lower stimulation intensities using smaller electrodes, which could also affect the degree of nerve fiber activation across different positions. Importantly, Aα-sensory fibers were consistently recruited before α-motor fibers.

CONCLUSIONS

These results imply that cathode positioning could help optimize preferential activation of hand muscles during cervical tSCS.

Preferential activation of proprioceptive and cutaneous sensory fibers compared to motor fibers during cervical transcutaneous spinal cord stimulation: A computational study

OBJECTIVE

Cervical transcutaneous spinal cord stimulation (tSCS) is a promising technology that can support motor function recovery of upper-limbs after spinal cord injury. Its efficacy may depend on the ability to recruit sensory afferents, conveying excitatory inputs onto motoneurons. Therefore, understanding its physiological mechanisms is critical to accelerate its development towards clinical applications. In this study, we used an anatomically realistic cervical tSCS computational model to compare α-motor, Aα-sensory, and Aβ-sensory fiber activation thresholds and activation sites.

APPROACH

We developed a 3D geometry of the cervical body and tSCS electrodes with a cathode centred at the C7 spinous process and an anode placed over the anterior neck. The geometrical model was used to estimate the electric potential distributions along motor and sensory fiber trajectories at the C7 spinal level using a finite element method. We implemented dedicated motor and sensory fiber models to simulate the α-motor and Aα-sensory fibers using 12, 16, and 20 µm diameter fibers, and Aβ-sensory fibers using 6, 9, and 12 µm diameter fibers. We estimated nerve fiber activation thresholds and sites for a 2 ms monophasic stimulating pulse and compared them across the fiber groups.

MAIN RESULTS

Our results showed lower activation thresholds of Aα- and Aβ-sensory fibers compared with α-motor fibers, suggesting preferential sensory fiber activation. We also found no differences between activation thresholds of Aα-sensory and large Aβ-sensory fibers, implying their co-activation. The activation sites were located at the dorsal and ventral root levels.

SIGNIFICANCE

Using a realistic computational model, we demonstrated preferential activation of dorsal root Aα- and Aβ-sensory fibers compared with ventral root α-motor fibers during cervical tSCS. These findings suggest high proprioceptive and cutaneous contributions to neural activations during cervical tSCS, which inform the underlying mechanisms of upper-limb functional motor recovery.

International Regulatory Standards for the Qualitative Measurement of Deep Brain Stimulation in Clinical Research

Deep brain stimulation (DBS) has progressed to become a promising treatment modality for neurologic and psychiatric disorders like epilepsy and major depressive disorder due to its growing personalization. Despite evidence pointing to the benefits of DBS if tested on these personalized qualitative metrics, rather than randomized-control trial quantitative standards, the evaluation of these novel devices appears to be based on the latter. This study surveyed the presence of this trend in the national regulatory guidelines of the prominent DBS researching countries. It was found that two governing bodies, in the European Union and Australia, acknowledged the option for qualitative measures. These findings support further development of national regulatory guidelines, so the neuroscientific community developing these neurotechnologies can better understand the impact their treatments have on patients.

Neural Recruitment During Conventional, Burst, and 10-kHz Spinal Cord Stimulation for Pain

Spinal cord stimulation (SCS) is a popular neurostimulation therapy for severe chronic pain. To improve stimulation efficacy, multiple modes are now used clinically, including conventional, burst, and 10-kHz SCS. Clinical observations have produced speculation that these modes target different neural elements and/or work via distinct mechanisms of action. However, in humans, these hypotheses cannot be conclusively answered via experimental methods. Therefore, we utilized computational modeling to assess the response of primary afferents, interneurons, and projection neurons to conventional, burst, and 10-kHz SCS. We found that local cell thresholds were always higher than afferent thresholds, arguing against direct recruitment of these local cells. Furthermore, although we observed relative threshold differences between conventional, burst, and 10-kHz SCS, the recruitment order was the same. Finally, contrary to previous reports, axon collateralization produced complex changes in activation thresholds of primary afferents. These results motivate future work to contextualize clinical observations across SCS paradigms. PERSPECTIVE: This article presents the first computational modeling study to investigate neural recruitment during conventional, burst, and 10-kilohertz spinal cord stimulation for chronic pain within a single modeling framework. The results provide insight into these treatments' unknown mechanisms of action and offer context to interpreting clinical observations.

Dorsal Root Ganglion Stimulation for Chronic Pain: Hypothesized Mechanisms of Action

Dorsal root ganglion stimulation (DRGS) is a neuromodulation therapy for chronic pain that is refractory to conventional medical management. Currently, the mechanisms of action of DRGS-induced pain relief are unknown, precluding both our understanding of why DRGS fails to provide pain relief to some patients and the design of neurostimulation technologies that directly target these mechanisms to maximize pain relief in all patients. Due to the heterogeneity of sensory neurons in the dorsal root ganglion (DRG), the analgesic mechanisms could be attributed to the modulation of one or many cell types within the DRG and the numerous brain regions that process sensory information. Here, we summarize the leading hypotheses of the mechanisms of DRGS-induced analgesia, and propose areas of future study that will be vital to improving the clinical implementation of DRGS. PERSPECTIVE: This article synthesizes the evidence supporting the current hypotheses of the mechanisms of action of DRGS for chronic pain and suggests avenues for future interdisciplinary research which will be critical to fully elucidate the analgesic mechanisms of the therapy.

High-resolution computational modeling of the current flow in the outer ear during transcutaneous auricular Vagus Nerve Stimulation (taVNS)

BACKGROUND

Transcutaneous auricular Vagus Nerve Stimulation (taVNS) applies low-intensity electrical current to the ear with the intention of activating the auricular branch of the Vagus nerve. The sensitivity and selectivity of stimulation applied to the ear depends on current flow pattern produced by a given electrode montage (size and placement).

OBJECTIVE

We compare different electrodes designs for taVNS considering both the predicted peak electric fields (sensitivity) and their spatial distribution (selectivity).

METHODS

Based on optimized high-resolution (0.47 mm) T1 and T2 weighted MRI, we developed an anatomical model of the left ear and the surrounding head tissues including brain, CSF/meninges, skull, muscle, blood vessels, fat, cartilage, and skin. The ear was further segmented into 6 regions of interest (ROI) based on various nerve densities: cavum concha, cymba concha, crus of helix, tragus, antitragus, and earlobe. A range of taVNS electrode montages were reproduced spanning varied electrodes sizes and placements over the tragus, cymba concha, earlobe, cavum concha, and crus of helix. Electric field across the ear (from superficial skin to cartilage) for each montage at 1 mA or 2 mA taVNS, assuming an activation threshold of 6.15 V/m, 12.3 V/m or 24.6 V/m was predicted using a Finite element method (FEM). Finally, considering every ROI, we calculated the sensitivity and selectivity of each montage.

RESULTS

Current flow patterns through the ear were highly specific to the electrode montage. Electric field was maximal at the ear regions directly under the electrodes, and for a given total current, increases with decreasing electrode size. Depending on the applied current and nerves threshold, activation may also occur in the regions between multiple anterior surface electrodes. Each considered montage was selective for one or two regions of interest. For example, electrodes across the tragus restricted significant electric field to the tragus. Stimulation across the earlobe restricted significant electric field to the earlobe and the antitragus. Because of this relative selectivity, use of control ear montages in experimental studies, support testing of targeting. Relative targeting was robust across assumptions of activation threshold and tissue properties.

DISCUSSION

Computational models provide additional insight on how details in electrode shape and placement impact sensitivity (how much current is needed) and selectivity (spatial distribution), thereby supporting analysis of existing approaches and optimization of new devices. Our result suggest taVNS current patterns and relative target are robust across individuals, though (variance in) axon morphology was not represented.

Image-based biophysical modeling predicts cortical potentials evoked with subthalamic deep brain stimulation

BACKGROUND

Subthalamic deep brain stimulation (DBS) is an effective surgical treatment for Parkinson's disease and continues to advance technologically with an enormous parameter space. As such, in-silico DBS modeling systems have become common tools for research and development, but their underlying methods have yet to be standardized and validated.

OBJECTIVE

Evaluate the accuracy of patient-specific estimates of neural pathway activations in the subthalamic region against intracranial, cortical evoked potential (EP) recordings.

METHODS

Pathway activations were modeled in eleven patients using the latest advances in connectomic modeling of subthalamic DBS, focusing on the hyperdirect pathway (HDP) and corticospinal/bulbar tract (CSBT) for their relevance in human research studies. Correlations between pathway activations and respective EP amplitudes were quantified.

RESULTS

Good model performance required accurate lead localization and image fusions, as well as appropriate selection of fiber diameter in the biophysical model. While optimal model parameters varied across patients, good performance could be achieved using a global set of parameters that explained 60% and 73% of electrophysiologic activations of CSBT and HDP, respectively. Moreover, restricted models fit to only EP amplitudes of eight standard (monopolar and bipolar) electrode configurations were able to extrapolate variation in EP amplitudes across other directional electrode configurations and stimulation parameters, with no significant reduction in model performance across the cohort.

CONCLUSIONS

Our findings demonstrate that connectomic models of DBS with sufficient anatomical and electrical details can predict recruitment dynamics of white matter. These results will help to define connectomic modeling standards for preoperative surgical targeting and postoperative patient programming applications.