Electrical Stimulation publications

Neuromodulation is widely employed to enhance neural activity and restore impaired circuitry, yet current methods suffer from invasiveness, limited depth, or poor spatial resolution. Temporal interference stimulation (TIS) has emerged as a non-invasive approach to modulate deep brain regions via interference of high-frequency electric fields, achieving spatial selectivity. Despite its potential, the physiological effects remain unexplored. Here, we present a customized liquid metal-based neural interface to record electrophysiological responses to TIS across diverse neural tissues, from brain organoids to mouse brains. The softness of the interface minimizes invasiveness while ensuring stable access to deep neural regions. Electrophysiological analyses reveal TIS-induced dynamics in neurodevelopment and functional connectivity. Furthermore, we investigated latent neural dynamics, presenting population-level neural activity, to compare responses across models of differing complexity and maturity. By combining soft neural interfaces with multi-scale analyses, this study uncovers how neuromodulation shapes neural dynamics and supports the development of future therapeutic strategies.

Current disease-sensing devices primarily focus on distinguishing between healthy and diseased states, effective for diagnosis but limited in guiding optimal intervention timing for prevention. We developed a tripartite framework identifying pre-disease state in depression, a reversible phase preceding irreversible onset. Using complex systems theory, we analyzed early-warning signals emerging as biological systems approach critical transitions. Continuous monitoring of nine multimodal biomarkers—spanning electrophysiological, behavioral, and biological—enabled classification into normal, pre-disease, and disease states by quantitatively defining critical points. An artificial intelligence agent classified disease states with 95.2% accuracy using multimodal data, enabled by ultrasoft neural probes for stable, low-damage recordings. Therapeutic validation with a skin-attachable wireless vagus nerve stimulator integrating soft three-dimensional electrodes demonstrated superior efficacy during pre-disease states. Subjects treated during pre-disease showed faster recovery and greater therapeutic responses, while those treated after disease onset failed to achieve full recovery. This framework provides evidence-based rationale for early intervention.

A hallmark of intelligence is proficiency in solving new problems, including those that substantially differ from previously seen problems. Problem solving in turn depends on the goal-directed generation of novel ideas and behaviours 1 , which has been proposed to involve internal representations of discrete units (or symbols) that can be recombined into numerous possible composite representations 1–7 . Although this view has been influential in cognitive-level explanations of behaviour, definitive evidence for a neuronal substrate of symbols has remained elusive. Here we identify a neural population that encodes action symbols—recombinable representations of discrete units of motor behaviour—in a specific area of the frontal cortex. In macaque monkeys performing a drawing-like task, we found behavioural evidence that action elements (strokes) exhibit three crucial features that indicate an underlying symbolic representation: (1) invariance over low-level motor parameters; (2) categorical structure, which reflects discrete action types; and (3) recombination into novel sequences. Based on simultaneous neural recordings across eight regions of the motor, premotor and prefrontal cortex, we identified population activity specifically in the ventral premotor cortex that encodes planned actions in a manner that also reflects invariance, categorical structure and recombination. These findings reveal a neural representation of action symbols localized to the ventral premotor cortex and a putative neural substrate for symbolic operations.

Cochlear implants (CIs) have successfully restored hearing in more than one million patients with severe to profound hearing loss worldwide. Although CIs effectively restore speech perception in quiet environments, sound localization remains challenging for bilateral CI users, particularly their ability to utilize interaural time differences (ITDs). The majority of clinical CI processors use a coding strategy that encodes ITD information only in the envelope of electrical pulse trains rather than their pulse timing, which may contribute to the poorer spatial hearing perception of CI users. We recently demonstrated in a behavioural study on early deafened, bilaterally CI‐implanted rats that pulse timing ITDs completely dominate ITD perception, whereas sensitivity to envelope ITDs is almost negligible in comparison. Building on this, we here investigated the neurophysiological sensitivity of the inferior colliculus (IC) to envelope and pulse timing ITDs at two different pulse rates (900 and 4500 pulses/s) and three different stimulation modulations (5, 20 and 100 Hz) in CI rats with different hearing experiences. Our results indicate that IC neurons exhibit far greater sensitivity to pulse timing ITD than envelope ITD independent of pulse rate, modulation rate or hearing experience. These findings suggest that to improve binaural hearing outcomes in bilateral CI users, clinical stimulation strategies should provide informative pulse timing ITDs. image Key points Current bilateral cochlear implants mostly provide time cues of arriving sound in the envelope of stimulation rather than in the timing of pulses. A behavioural study on cochlear implanted rats has shown that pulse timing dominates binaural hearing. Here we examined the neural sensitivity to independent changes in time cues presented on the envelope or pulse timing, respectively, in rats with bilateral cochlear implants and different hearing experiences. Auditory midbrain neurons were consistently more sensitive to changes in pulse timing rather than envelope, irrespective of stimulus pulse rate, modulation rate or an animal's hearing experience.

Magnetogenetic deep brain stimulation (MG‐DBS) represents a wireless neuromodulation that has demonstrated long‐lasting behavioral benefits in Parkinson's disease models. However, the circuit‐level mechanisms underlying these therapeutic effects have remained uncharacterized due to limitations of conventional neural interfaces. We present a bio‐integrable soft neural interface featuring ultrasoft liquid‐metal probes with bioresorbable stiffeners and customizable interconnects directly printed onto cranial surfaces to match individual skull anatomy and nanoparticle injection sites. This platform enables stable multi‐regional recordings from deep brain structures without chronic tissue damage. We systematically investigate MG‐DBS therapeutic mechanisms in a Parkinson's disease mouse model. Circuit‐level analysis reveals that MG‐DBS modulates pathological beta‐band oscillations and inter‐regional synchrony across the cortico‐basal ganglia‐thalamic circuit. Direct comparison with conventional electrical DBS demonstrates that MG‐DBS effects persisted approximately fifteen‐fold longer after stimulation cessation. Our electrophysiological recordings elucidate the mechanistic basis for this sustained therapeutic effect, providing unprecedented insights into magnetogenetic neuromodulation dynamics.

Neural interfaces for monitoring and modulating spinal nerve activity are increasingly being designed to be flexible and stretchable to enhance their biomechanical compatibility and integration. However, excessive flexibility introduces practical limitations such as difficulty in insertion into narrow spinal spaces and long-term electrical instability, hindering real-world applications. In this study, we developed a spinal nerve interface by incorporating a liquid-metal conductor and dynamic stiffness-based variable-compliance structure, which can address the challenges of current flexible neural interface technologies. During insertion, the dynamic stiffness enhancer minimizes unintended buckling and ensures minimally invasive implantation into the intended target. The proximity of the proposed device to the spinal cord increases as it flexes automatically and rapidly in a humid environment. The liquid-metal conductor maintained stable electrical properties in freely moving rats, ensuring reliable and sustained functionality. This study lays the foundation for practical, fully implantable spinal bioelectronics designed with a focus on ease of implantation and long-term functionality.

Stimulation of sensory vagal pathways is typically delivered via invasive, cervical vagus nerve stimulation (cVNS) or noninvasive, trans-auricular nerve stimulation (taNS). While both methods are investigated therapeutically, their effects on brain physiology remain poorly understood, hindering mechanistic insights and stimulus optimization. In 6 awake nonhuman primates, we recorded cortical vagal-evoked potentials (VEPs) from subdural electrodes placed in prefrontal, sensorimotor and parietal cortical areas, in response to cVNS or taNS. Across 478 different taNS and cVNS protocols, we varied stimulation side, intensity, frequency, pulse count, and pulse width and assessed independent effects on amplitude and latency of early (EC; 30-100 ms), intermediate (IC; 101-200 ms) and late components (LC; 201-500 ms) of VEPs. Fixed and random effects of stimulation parameters and subjects, respectively, on VEPs were assessed using a linear mixed-effects model. Overall, cVNS elicits more robust VEPs than taNS, with larger EC, IC and LC amplitudes, in both hemispheres. cVNS-elicited ECs and LCs are largest in PFC and PC areas, whereas ICs are largest in SM areas. On the other hand, taNS generally does not elicit area-specific responses. cVNS-elicited ECs have slower latency than ta-NS elicited ECs. Higher stimulation frequencies and intensities and a longer pulse width elicit larger ECs and ICs for cVNS, and to some extent for taNS. Both short and long cVNS trains elicit stronger ECs, and long trains elicit slower ICs. Earlobe stimulation elicits VEPs that partially overlap with those from taNS. In conclusion, cVNS and taNS elicit cortical VEPs in a manner consistent with distinct engagement of ascending vagal pathways, with both similarities and differences in the effects of stimulation parameters on evoked responses.

Selective activation of small versus large diameter nerve fibers represents a broad challenge in the electrical modulation of neural activity for clinical applications. For example, off-target activation of motor nerve fibers was implicated in several failed vagus nerve stimulation (VNS) clinical trials by effectively limiting maximum applicable current amplitudes, thereby preventing robust activation of smaller diameter fibers responsible for therapeutic effects. Kilohertz high-frequency (HF) block is one of several strategies proposed to overcome this gap, but has yet to be thoroughly investigated in large, multifascicular, mixed nerves comparable in scale to humans. Here, we evaluated effects of HF block on the neural, muscular, and cardiac responses evoked by traditional low-frequency (LF) VNS in a large animal (swine) model. We found that HF selectively blocks neural fibers, and their associated physiological responses, in a size dependent manner with a high degree of specificity and establish initial thresholds for achieving block in large multifascicular nerves. Critically, we demonstrate that appropriately titrating HF amplitude can eliminate off-target LF-evoked motor responses while preserving activation of smaller diameter fibers targeted by clinical VNS. Beyond VNS, these data establish HF as a potentially viable strategy in any neuromodulation therapies where off-target activation of large diameter fibers limits therapeutic efficacy.

Primate societies rely on the production and interpretation of social signals, in particular those displayed by the face. Facial movements are controlled, according to the dominant neuropsychological schema, by two separate circuits, one originating in medial frontal cortex controlling emotional expressions, and a second one originating in lateral motor and premotor areas controlling voluntary facial movements. Despite this functional dichotomy, cortical anatomy suggests that medial and lateral areas are directly connected and may thus operate as a single network. Here, we test these contrasting hypotheses through structural and functional MRI (fMRI) guided electrical stimulation and simultaneous multichannel recordings from key facial motor areas in the macaque monkey brain. These areas include medial facial motor area M3 (located in the anterior cingulate cortex); two lateral face-related motor areas: M1 (primary motor) and PMv (ventrolateral premotor); and S1 (primary somatosensory cortex). Cortical responses evoked by intracortical stimulation revealed that medial and lateral areas can exert significant functional impact on each other. Simultaneous recordings of local field potentials in all facial motor areas further confirm that during facial expressions, medial and lateral facial motor areas significantly interact, primarily in the alpha and beta frequency ranges, whereas during voluntary chewing, coupling occurs at lower frequencies. These functional interactions varied across facial movement types. Thus, at the cortical level, the control of facial movements is not mediated through independent (medial/lateral) functional streams, but results from an interacting sensorimotor network.

The hypothalamus is critical for regulating behaviors essential for survival and locomotion, but how it integrates internal needs and transmits locomotion commands to the spinal cord (SC) remains unclear. We found that glutamatergic neurons in lateral hypothalamic area (LHA) are essential for regulating motivated locomotor activity. Using single-neuron projectome analysis, trans-synaptic tracing, and optogenetic manipulation, we showed that LHA facilitates motivated locomotion during food seeking via pontine reticular nucleus, oral part (PnO) projection neurons, rather than direct SC projections or indirect stress signaling via medial septum and diagonal band. Activating PnO-SC projection neurons also initiated locomotion. Importantly, LHA-PnO projection neurons were crucial for regulating locomotor recovery following mouse spinal cord injury (SCI). Motor cortex signals gated LHA deep brain stimulation treatment markedly promoted long-term restoration of hindlimb motor functions after severe SCI. Thus, we have identified a hypothalamic-pontine-spinal pathway and the stimulation paradigm for potential therapeutic intervention after SCI.

Objective. Infrared neural stimulation (INS) has emerged as a potent neuromodulation technology, offering safe and focal stimulation with superior spatial recruitment profiles compared to conventional electrical methods. However, the neural dynamics induced by INS stimulation remain poorly understood. Elucidating these dynamics will help develop new INS stimulation paradigms and advance its clinical application. Approach. In this study, we assessed the local network dynamics of INS entrainment in the auditory thalamocortical circuit using the chronically implanted rat model. Our approach focused on measuring INS energy-based local field potential (LFP) recruitment induced by focal thalamocortical stimulation. We further characterized linear and nonlinear oscillatory LFP activity in response to single-pulse and periodic INS and performed spectral decomposition to uncover specific LFP band entrainment to INS. Finally, we examined spike-field transformations across the thalamocortical synapse using spike-LFP coherence coupling measures. Main results. We found that INS significantly increases LFP amplitude as a log-linear function of INS energy per pulse, primarily entraining to LFP β and γ bands with synchrony extending to 200 Hz in some cases. A subset of neurons demonstrated nonlinear, chaotic oscillations linked to information transfer across cortical circuits. Finally, we utilized spike-field coherences to correlate spike coupling to LFP frequency band activity and suggest an energy-dependent model of network activation resulting from INS stimulation. Significance. We show that INS reliably drives robust network activity and can potently modulate cortical field potentials across a wide range of frequencies in a stimulus parameter-dependent manner. Based on these results, we propose design principles for developing full coverage, all-optical thalamocortical auditory neuroprostheses.

As part of a deep brain stimulation (DBS) procedure, direct electrophysiological recordings from thalamic and basal ganglia circuits were obtained from a pediatric participant with secondary dystonia to examine the neural mechanisms of motor timing. The participant performed a two-phase repetitive timing task that began with synchronization to a series of externally presented auditory cues, followed by a continuation phase in which the participant maintained the movement rhythm without external guidance. Simultaneous recordings from stereoelectroencephalography (sEEG) leads implanted in the globus pallidus internus (GPi), subthalamic nucleus (STN), ventral oralis (Vo), centromedian nucleus (CM), and thalamic nuclei receiving projections from the cerebellum (ventral intermediate nucleus (VIM) and centrolateral nucleus (CL)) revealed behaviorally relevant neural dynamics. Local field potentials (LFPs) from recorded nuclei exhibited strong, time-locked responses during the synchronization phase, with reduced amplitudes during the continuation phase. Time-frequency analysis showed consistent power increases around the corresponding mean of the inter-response intervals (IRIs) in both behavioral output and neural signals, suggesting frequency-specific entrainment across circuits. These findings suggest of cerebellar-thalamic and basal ganglia contributions to temporal coordination.

Purpose: We aim to develop a robust method to improve the estimation accuracy of motor-evoked potential (MEP) recruitment curves (RCs), including motor threshold, in small-sample settings which typically involve fewer than 40 stimuli. Methods: We present a hierarchical Bayesian (HB) method to model MEP size as a rectified-logistic function of stimulation intensity. This method is designed to account for small samples, handle outliers without discarding data, quantify estimation uncertainty, and simulate synthetic data that closely matches real observations, useful for optimizing experimental design. We validate its performance on transcranial magnetic stimulation (TMS), epidural spinal cord stimulation (SCS), and synthetic TMS datasets, and provide an open-source library for Python, called hbMEP, for diverse applications. Results: The rectified-logistic outperformed sigmoidal functions in predictive accuracy on TMS and SCS datasets, as demonstrated through cross-validation. A mixture extension of the HB model improved robustness to outliers by further increasing its predictive accuracy. The HB model reduced threshold estimation error by up to 70% on sparse synthetic TMS data compared to non-hierarchical models. Bayesian estimation with the HB model reduced the required number of participants by at least 23% to detect a shift in threshold with 80% power, compared to frequentist testing. Empirical results on human SCS data further validated its applicability to real data. Conclusion: By improving accuracy on sparse data, our method minimizes the number of stimuli needed to probe each individual's neuromuscular parameters across multiple muscles simultaneously, thereby reducing session duration and the risk of inadvertent neuromodulation. Our approach provides a more statistically powerful and conclusive framework for inferring changes in threshold, and therefore corticospinal excitability. The hbMEP library streamlines and unifies the analysis of RCs across stimulation modalities and experimental paradigms.

Sympathetic hyperactivity is a common feature observed in both primary and secondary hypertension, and neuromodulation-based therapies targeting this overactivity are under active investigation. The intermediolateral nucleus (IML) in the spinal cord plays a central role in sympathetic regulation, and its direct stimulation induces a transient, frequency-dependent reduction in blood pressure. However, this effect gradually diminishes as the baroreflex restores blood pressure to baseline. Therefore, this study aims to develop a closed-loop (CL) stimulation system that adjusts parameters in real-time based on neural activity recorded from the IML and to enhance and prolong the transient antihypertensive effect. This strategy was evaluated in normotensive and angiotensin II-induced hypertensive rat models and compared to open-loop (OL) stimulation of equal intensity. In both models, CL stimulation produced a more sustained blood pressure reduction than that of OL stimulation. These findings suggest that CL spinal cord stimulation offers a more effective and durable therapeutic option for treating hypertension.

Target engagement determines therapeutic efficacy. In spinal cord stimulation (SCS), efficacy depends on the attributes of the electrodes used to deliver stimulation, particularly their position, interelectrode distance, size, and current orientation. We tested the effects of these electrode parameters using cervical stimulation to elicit forelimb muscle responses in the rat. We implanted devices over the C6 dorsal root entry zone (DREZ) in eight rats and measured responses in six forelimb muscles. We used custom designed electrode arrays to systematically vary position, distance, and size with linear arrays, and current orientation with circular arrays. Stimulation consisted of biphasic and pseudomonophasic pulse shapes with bipolar or distant return, and a high-definition montage with four returns surrounding a central contact to increase local current density. Efficacy was calculated from the motor thresholds of recruitment curves, and parameters were compared using linear mixed models. For position, stimulation over the DREZ was most effective, reducing thresholds by 25.9% ( p = 0.0002) relative to the midline; efficacy decreased as the electrode was positioned medial or lateral to the DREZ. For distance, placing the electrodes farther apart significantly improved efficacy ( p = 0.0022). For size, large electrodes increased SCS efficacy, reducing thresholds by 21.5% ( p = 0.0026) relative to small electrodes. After accounting for position and distance, current orientation did not affect SCS efficacy. Lastly, the high-definition montage decreased efficacy, increasing thresholds by 16.7% ( p = 0.003) relative to monopolar stimulation with a distant return. In conclusion, electrode position, distance, and size had the greatest effect on SCS efficacy, with the optimal parameters combining large electrodes at the DREZ position with a distant return. These results have the potential to lower the current needed to engage cervical spinal circuitry and to inform the design of SCS systems to improve dexterity in people with neurological injury or disease.

Introduction Deep brain stimulation (DBS) is a neuromodulation method for treatment of various neurological disorders. Research on DBS has often focused on local inhibition or excitation effects, at the site of stimulation. However, it is well-known that DBS can lead to robust evoked potentials (EP) not only at the stimulation site, representing the local effect, but also in distant brain regions, representing the effects on distant targets. While the significance of these EPs for therapeutic outcomes is not known, it appears that the electrical effects of DBS have a partial modulatory impact on downstream targets. Nonetheless, it partly remains unclear through what mechanism DBS pulses travel to the distant targets or what portion of the pulses travel along the normal pathways from the stimulation site. The possible scenarios include orthodromic or antidromic pathways, accessory pathways, normally inhibited pathways, and direct electromagnetic activation of distant sites. We hypothesize that the pathways that transmit DBS pulses include the pathways that transmit intrinsic neural signals. Methods To test this, we performed a transfer function analysis on deep brain recordings from children with dystonia, during DBS-off condition and compared its impulse response with the transmission of signals from electrical stimulation during DBS-on condition. We compared impulse responses derived from intrinsic neural signals during voluntary movement (DBS-off) to evoked potentials (EPs) recorded during electrical stimulation (DBS-on), focusing on directional transmission (orthodromic vs. antidromic). Results DBS EPs were more accurately predicted by impulse responses corresponding to direct axonal activation rather than somatic relay. Significant correlations between intrinsic signal transfer functions and EPs, particularly in orthodromic directions ( p -value < 0.01) from pallidum to thalamus and subthalamic nucleus, support our hypothesis that DBS travels along physiological pathways. Discussion These results suggest that DBS engages existing motor pathways to reach distant targets, offering mechanistic insight into its network effects. This supports future approaches that could tailor treatment plans based on individual connectivity maps to improve clinical efficacy of DBS.

Restoring upper limb function is critical in individuals with central paralysis, and hand control is a priority in patients with neurological impairments. Functional electrical stimulation with implantable electrodes targeting the peripheral nervous system has the potential to selectively recruit hand muscles and generate multiple functional hand movements. However, the implantation of electrodes in the forearm or elbow areas requires multiple incisions for surgery, and elbow joint movements cannot be performed. In this study, we designed and implanted two epineural cuffs on the median and radial nerves in the distal brachial plexus of a single Japanese macaque ( Macaca fuscata ) monkey. The cuffs were successfully placed via an axillary approach using a single incision. Electrical stimuli were applied to innervate the contraction patterns of the hand, forearm, and triceps muscles relevant to the median and radial nerves. The evoked potentials of the target muscles electrically stimulated the distal brachial plexus to reliably and selectively innervate the upper limb muscles at the functional group level. Our results demonstrated that the distal brachial plexus can be a useful stimulation site for upper limb muscle contraction and that the axillary approach enables electrode placement to peripheral nerves required for upper limb control.

Closed-loop neural control is a powerful tool for both the scientific exploration of neural function and for mitigating deficiencies found in open-loop deep brain stimulation (DBS). Here, we present a protocol for artificial intelligence-guided neural control in rats using deep reinforcement learning (RL) and infrared neural stimulation (INS). We describe steps for integrating RL closed-loop control into neuroscience and neuromodulation studies. We then detail procedures for using and deploying infrared INS in chronic DBS applications. For complete details on the use and execution of this protocol, please refer to Coventry et al.1 and Coventry and Bartlett.2.

In this paper we aim to examine the correlation between diffusion tensor imaging parameters of anatomical connectivity and characteristics of signal transmission obtained from patient-specific transfer function models. Here, we focused on elucidating the correlation between structural and functional neural connectivity within a cohort of pediatric patients diagnosed with dystonia. Diffusion tractography images were obtained from 12 patients with dystonia prior to the deep brain stimulation surgery. For each patient, we processed the imaging data to estimate anatomical measures including fractional anisotropy, axial diffusivity, number of fibre tracts per unit area, and fibre tract length. After the implantation of temporary depth leads for each patient as part of their treatment plan, intracranial signals were recorded. Transfer function models of local field potential recordings and the corresponding measures of functional connectivity were computed for each patient. Linear mixed effect analysis was then employed to determine the relationship between transfer function measures and diffusion tractography parameters. Our results illustrate a positive correlation between fractional anisotropy, AD, and intrinsic neural transmission measures, representing amplification and spread of intrinsic neural oscillations, obtained from the transfer functions models. However, no significant correlation was found between the functional connectivity and number of fibre tracts or fibre lengths. Our findings suggest that white matter integrity, as measured by fractional anisotropy and AD, can potentially reflect the amplification and spread of intrinsic brain signals throughout the network. This study underscores the significant relationship between structural and functional connectivity, offering valuable insights into propagation of neural activity in the brain network and potential implications for optimizing non-invasive treatments and planning for neurological disorders.

Micro-electrocorticography (µECoG) electrodes have emerged to balance the trade-off between invasiveness and signal quality in brain recordings. However, its large-scale applicability is still hindered by a lack of comparative studies assessing the relationship between ECoG and traditional recording methods such as penetrating electrodes. This study aimed to compare somatosensory evoked potentials (SEPs) through the lenses of a µECoG and an intracortical microelectrode array (MEA). The electrodes were implanted in the pig’s primary somatosensory cortex, while SEPs were generated by applying electrical stimulation to the ulnar nerve. The SEP amplitude, signal-to-noise ratio (SNR), power spectral density (PSD), and correlation structure were analysed. Overall, SEPs resulting from MEA recordings had higher amplitudes and contained significantly more spectral power, especially at higher frequencies. However, the SNRs were similar between the interfaces. These results demonstrate the feasibility of using µECoG to decode SEPs with wide-range applications in physiology monitoring and brain–computer interfaces.