Neurostimulation is a therapeutic approach that involves the targeted delivery of electrical impulses to modulate neural activity within specific circuits of the central or peripheral nervous system, aiming to restore or improve function in various neurological and psychiatric conditions.[1] This technique encompasses both invasive methods, such as implantable devices, and noninvasive approaches, like magnetic or transcutaneous stimulation, to activate or inhibit neural pathways.[2]The origins of neurostimulation trace back to the mid-20th century, with early applications emerging in the 1960s through spinal cord stimulation for chronic pain management, based on the gate control theory of pain transmission.[3][4] Significant advancements occurred in the 1990s, including the U.S. Food and Drug Administration's approval of deep brain stimulation in 1997 for essential tremor, expanding its use to other movement disorders.[3] Over the decades, technological innovations, such as closed-loop systems that adapt stimulation based on real-time physiological feedback, have enhanced efficacy and personalization, reducing side effects and improving outcomes in clinical settings.[3]Key types of neurostimulation include deep brain stimulation (DBS), which involves surgically implanted electrodes in subcortical regions to treat conditions like Parkinson's disease;[1]vagus nerve stimulation (VNS), where a device stimulates the vagus nerve to manage epilepsy and depression; and spinal cord stimulation (SCS), used primarily for intractable pain by interrupting pain signal transmission.[5] Noninvasive variants, such as repetitive transcranial magnetic stimulation (rTMS) and transcranial direct current stimulation (tDCS), apply external fields or currents to the scalp to influence cortical activity without surgery.[2] Emerging closed-loop neurostimulation integrates sensing capabilities to dynamically adjust therapy, as seen in responsive neurostimulation systems for epilepsy that detect and interrupt seizure activity.[3]Neurostimulation has broad applications across medical fields, providing relief for chronic neuropathic pain through peripheral nerve stimulation, which targets specific dermatomes to block pain signals via non-nociceptive pathways.[2] In epilepsy, devices like responsive cortical stimulation have demonstrated median seizure reductions of around 75% in long-term studies (as of 2025) in drug-resistant cases.[6] For movement disorders, DBS effectively alleviates symptoms of Parkinson's disease and essential tremor by modulating basal ganglia circuits.[1] Additionally, it serves as an adjunct therapy for psychiatric disorders, including treatment-resistant depression via VNS or rTMS, offering alternatives when medications fail.[5] Ongoing research continues to refine these therapies, focusing on precision targeting and minimally invasive techniques to broaden accessibility.[1]
Fundamentals
Definition and Principles
Neurostimulation refers to the purposeful modulation of nervous system activity through the application of targeted electrical, magnetic, or other external or implanted stimuli to alter neural function, primarily for therapeutic purposes in treating neurological, psychiatric, or sensory disorders.[3] This technique aims to influence abnormal neural pathways by delivering controlled impulses that can either excite or inhibit neuronal activity, thereby restoring or enhancing normal physiological responses disrupted by disease processes.[7]At its core, neurostimulation operates on fundamental electrophysiological principles involving the manipulation of neuronal membrane potentials to induce excitation or inhibition. Excitation occurs when stimuli cause depolarization, reducing the membrane potential below the threshold to trigger an action potential—a rapid influx of ions leading to signal propagation along the axon—while inhibition results from hyperpolarization, which increases the threshold for firing and dampens synaptic transmission.[8] These effects leverage the all-or-nothing nature of action potentials and the integration of excitatory and inhibitory postsynaptic potentials at synapses to modulate network-level activity. A key aspect of effective stimulation is the strength-duration relationship, which describes the minimal stimulus intensity required to activate a neuron as a function of pulse duration; this is classically modeled by Lapicque's formula:I = I_r \left(1 + \frac{\tau}{t}\right)where I is the threshold current, I_r is the rheobase (the minimal current for infinite duration), \tau is the chronaxie (the pulse duration at twice the rheobase), and t is the stimulus duration. This hyperbolic relationship guides parameter selection in neurostimulation to achieve reliable neural activation while minimizing energy use and tissue damage.[9]Neurostimulation employs various types of stimuli to achieve these effects, broadly categorized into electrical currents that directly depolarize membranes via electrodes, magnetic fields that induce currents through electromagnetic induction, ultrasound waves that mechanically modulate ion channels, and optical methods that target light-sensitive proteins in optogenetics.[10] Therapeutically, these approaches pursue goals such as pain relief by interrupting nociceptive signals, movement restoration in disorders like Parkinson's disease through targeted basal ganglia modulation, sensory replacement as in cochlear implants for hearing loss, and mood regulation in conditions like depression via limbic system stimulation.[5]
Physiological Mechanisms
Neurostimulation exerts its effects primarily at the cellular level by modulating neuronal membrane potentials through the activation of voltage-gated ion channels. Electrical or other stimuli depolarize the neuronal membrane, leading to an influx of sodium ions (Na⁺) and subsequent efflux of potassium ions (K⁺) during action potential generation, particularly in axons and dendrites which have lower activation thresholds compared to neuronal somata. This direct activation propagates action potentials along neural fibers, altering excitability in targeted regions.[11][12]At the synaptic level, neurostimulation induces long-term potentiation (LTP) or long-term depression (LTD), depending on stimulation parameters, which strengthen or weaken synaptic efficacy over time. High-frequency stimulation, for instance, can trigger LTP by enhancing postsynaptic responses through increased expression of AMPA and NMDA receptors, while low-frequency patterns promote LTD via internalization of these receptors. Additionally, stimulation modulates neurotransmitter release, including excitatory dopamine and inhibitory GABA, thereby influencing synaptic transmission and network balance without necessarily evoking full action potentials. These effects align with Hebbian learning principles, where correlated pre- and postsynaptic activity drives synaptic plasticity to reinforce functional connections.[13][14][12][15]On a network scale, neurostimulation entrains oscillatory rhythms, such as theta (4-8 Hz) or gamma (30-100 Hz) bands, by synchronizing neuronal firing across brain circuits, which stabilizes pathological activity or enhances information processing. This entrainment arises from periodic forcing that aligns endogenous oscillations, promoting plasticity through sustained changes in circuit dynamics. Spatial targeting ensures specificity, minimizing off-target activation by focusing stimuli on precise neural populations via electrode placement or waveform design. Frequency dependence further refines outcomes: low frequencies (e.g., 1-10 Hz) typically facilitate excitation and LTP, whereas high frequencies (e.g., >100 Hz) often induce inhibition through depolarization block or synaptic depletion.[16][17][18]Key concepts include orthodromic conduction, where action potentials propagate in the normal physiological direction from dendrite to axon terminal, and antidromic conduction, which reverses this flow toward the soma, potentially activating upstream circuits. Post-stimulation after-effects, such as prolonged excitability changes lasting minutes to hours, result from lingering ion channel modifications and second-messenger cascades, contributing to therapeutic durability. These mechanisms collectively enable neurostimulation to reshape neural activity with high precision.[19][20]
Invasive Neurostimulation
Deep Brain Stimulation
Deep brain stimulation (DBS) is an invasive neuromodulation technique that delivers controlled electrical impulses to specific subcortical brain regions to alleviate symptoms of movement disorders and certain neuropsychiatric conditions. The therapy involves surgically implanting electrodes in targeted neural structures, which are connected to an implantable pulse generator that modulates abnormal neural activity. First approved by the U.S. Food and Drug Administration (FDA) in 1997 for treating essential tremor and Parkinsonian tremor via thalamic stimulation, DBS expanded to advanced Parkinson's disease in 2002, with over 250,000 procedures performed worldwide by 2025.[21][22][23]The DBS procedure typically occurs in two stages under stereotactic guidance to ensure precise electrode placement. Thin electrodes are implanted into deep brain structures such as the basal ganglia (including the subthalamic nucleus for Parkinson's disease or the globus pallidus interna for dystonia) or the thalamus (ventral intermediate nucleus for essential tremor), often while the patient is awake to allow real-time symptom assessment. These leads are tunneled under the skin to a pulse generator, a battery-powered device similar to a pacemaker, implanted subcutaneously in the chest or abdomen. Postoperatively, the system is programmed noninvasively to optimize therapeutic effects.[24][25][21]Stimulation parameters are adjustable and tailored to the patient's condition, commonly including pulse widths of 60-200 μs, frequencies of 130 Hz for motor symptom control, and amplitudes ranging from 1-5 V to balance efficacy and side effects. Primary applications include Parkinson's disease, where DBS targets the subthalamic nucleus to reduce tremor, rigidity, and bradykinesia; essential tremor via thalamic stimulation; and epilepsy through anterior thalamic nucleus targeting to decrease seizure frequency. For neuropsychiatric uses, the nucleus accumbens serves as a key target in obsessive-compulsive disorder to disrupt dysfunctional circuits, while emerging applications explore ventral capsule/ventral striatum sites for treatment-resistant depression and addiction.[26][24][27]Clinical outcomes demonstrate substantial benefits, particularly in advanced Parkinson's disease, with 50-70% improvement in motor symptoms such as tremor and dyskinesia, alongside reduced reliance on dopaminergic medications. Long-term follow-up shows sustained symptom relief for 5-10 years or more in many patients, though disease progression may occur. Rechargeable pulse generator models offer battery life of 3-5 years under typical usage before requiring replacement, with advancements extending this to over 15 years in some systems.[28][29][30]
Spinal Cord Stimulation
Spinal cord stimulation (SCS) is an invasive neuromodulation technique that delivers electrical impulses to the spinal cord via implanted electrodes to alleviate chronic pain, primarily by interrupting pain signal transmission.[31] The first SCS implant was performed in 1967 by C. Norman Shealy, who placed an electrode array on the dorsal columns of a patient with intractable pain, marking the clinical inception of the therapy based on emerging theories of pain modulation.[32] Over decades, SCS has evolved into a standard treatment for refractory neuropathic pain, with systems now incorporating advanced waveforms and patient-specific programming.[33]The procedure typically involves a trial phase followed by permanent implantation. During the trial, percutaneous leads—thin, flexible wires with multiple electrodes—are inserted into the epidural space of the thoracic or lumbar spine under fluoroscopic guidance, targeting the dorsal columns to cover the painful dermatomes.[34] If successful (often defined as at least 50% pain relief), surgical leads are placed via laminotomy, connected to an implantable pulse generator (IPG) in the abdominal or gluteal region, which is programmed externally to optimize stimulation.[35]Electrode arrays, such as paddle or cylindrical types, are positioned 2-4 mm off the midline at levels like T8-T10 for lower body pain, ensuring precise coverage without direct spinal cord penetration.[31]The primary mechanism of SCS draws from the gate control theory of pain, proposed by Melzack and Wall in 1965, wherein activation of large-diameter Aβ fibers in the dorsal columns inhibits nociceptive signals from small-diameter Aδ and C fibers in the spinothalamic tract at the dorsal horn.[33] This segmental gating reduces pain perception without altering ascending pathways. High-frequency paradigms, such as 10 kHz stimulation, provide paresthesia-free relief by suppressing dorsal horn neuronal hyperexcitability and modulating neuroglial interactions, distinct from traditional low-frequency effects.[36]SCS is FDA-approved for conditions including failed back surgery syndrome (FBSS), complex regional pain syndrome (CRPS), and refractory ischemic limb pain, where conservative treatments have failed.[34] In FBSS, it targets persistent low back and leg pain post-laminectomy, while for CRPS, it addresses neuropathic burning in affected limbs; ischemic applications include angina or peripheral vascular disease pain.[37] Off-label uses extend to spasticity in spinal cord injury, where stimulation facilitates motor recovery and reduces hypertonia.[38]Stimulation parameters are customizable, with frequencies ranging from 40 Hz (traditional tonic) to 10,000 Hz (high-frequency), pulse widths of 30-500 μs, and amplitudes adjusted to patient tolerance.[39] Burst stimulation delivers groups of high-frequency pulses at 40 Hz base rate, mimicking natural neuronal firing for enhanced analgesia without sensory side effects.[40] Adaptive closed-loop systems, incorporating sensors for posture or activity, dynamically adjust parameters to maintain efficacy, improving outcomes in mobile patients.[41]Clinical outcomes demonstrate substantial pain reduction in refractory cases, with meta-analyses reporting 50-60% of patients achieving at least 50% relief in overall pain intensity, alongside improvements in function and reduced opioid use.[42] For movement disorders like stiff-person syndrome, SCS has shown promise in alleviating rigidity and spasms, as evidenced by case reports of symptom control post-implantation.[43] A dorsal root ganglion (DRG) variant of SCS, targeting focal pain sites, received FDA approval in 2016 for CRPS and similar neuropathies, offering superior coverage for unilateral or distal symptoms compared to traditional epidural approaches.[44]
Intrathecal and Intraspinal Stimulation
Intrathecal and intraspinal electrical stimulation involve the delivery of electrical impulses directly into the subarachnoid space or spinal cord parenchyma to modulate neural activity in spinal pathologies, distinct from epidural approaches by enabling deeper tissue penetration. Procedures typically utilize catheter-based systems for intrathecal electrode placement or surgical implantation of microelectrode arrays into the spinal cord for intraspinal stimulation, often integrated with implantable pulse generators. While intrathecal drug delivery (e.g., morphine for pain) originated in the 1970s, electrical neuromodulation in these approaches emerged in research settings by the early 2000s, primarily for spinal cord injury (SCI) and chronic pain unresponsive to standard methods. These techniques remain largely experimental, with limited human clinical data as of 2025, and are not FDA-approved for routine use.[45][46][47]Primary targets include the anterior and posterior horns of the spinal cord to facilitate motor recovery in SCI by activating residual neural circuits, while intrathecal approaches address spasticity through targeted inhibition of hyperactive reflexes. Applications extend to scenarios unresponsive to standard spinal cord stimulation, such as severe neuropathic pain, where intraspinal microstimulation directly engages dorsal horn neurons to interrupt pain signaling pathways. Additionally, these methods support recovery in autonomic functions, building on principles of epidural stimulation but with greater precision for intramedullary structures.[47][48][49]Stimulation parameters emphasize low-intensity currents, typically ranging from 0.5 to 2 mA, to provide neuroprotection by reducing secondary injury cascades like inflammation and excitotoxicity following SCI, with intermittent protocols (e.g., 20-40 Hz frequencies) designed to foster synaptic plasticity and long-term circuit remodeling. Preliminary clinical outcomes from limited studies demonstrate potential efficacy in functional restoration, including improvements in bladder function and decreased reliance on opioids for chronic neuropathic pain management via sustained analgesia; however, large-scale human data are lacking. Ongoing research as of 2025 focuses on advancing these techniques for broader clinical translation.[38][50]
Non-Invasive Brain Stimulation
Transcranial Magnetic Stimulation
Transcranial magnetic stimulation (TMS) is a non-invasive neurostimulation technique that uses rapidly changing magnetic fields to induce electrical currents in targeted brain regions, primarily for diagnostic and therapeutic applications in neurology and psychiatry. Developed in 1985 by Anthony Barker and colleagues at the University of Sheffield, TMS allows focal stimulation of cortical areas without the need for surgery or direct skin contact, distinguishing it from invasive methods by its temporary and reversible effects.[51] The procedure involves placing a magnetic coil on the scalp over the region of interest, such as the dorsolateral prefrontal cortex for mood disorders, where brief pulses generate magnetic fields that penetrate the skull and induce neuronal depolarization. Repetitive TMS (rTMS) extends this to therapeutic protocols, delivering trains of pulses to modulate neural activity over multiple sessions, often tailored to the patient's motor threshold determined via initial single-pulse mapping.[52]The underlying mechanism relies on electromagnetic induction, governed by Faraday's law, which states that a time-varying magnetic field B induces an electric field E according to E = -\frac{\partial B}{\partial t}. This induced electric field depolarizes neurons in superficial cortical layers, typically up to 2-3 cm deep, altering cortical excitability through long-term potentiation or depression depending on stimulation patterns. High-frequency rTMS (above 5 Hz) generally increases excitability, while low-frequency (below 1 Hz) decreases it, enabling targeted modulation of dysfunctional circuits in conditions like depression.[52] Unlike direct electrical stimulation, magnetic fields pass unimpeded through skin and bone, minimizing discomfort and allowing precise, image-guided targeting with neuronavigation systems.[53]Clinically, TMS received FDA clearance in 2008 for treating major depressive disorder (MDD) in patients unresponsive to medications, marking its first major therapeutic approval. It is also approved for migraine prophylaxis, where single-pulse or low-frequency protocols reduce headache frequency by inhibiting cortical spreading depression. In stroke recovery, rTMS facilitates motor and language rehabilitation by enhancing ipsilesional excitability and suppressing contralesional hyperactivity, with protocols often applied post-acutely. Theta-burst stimulation (TBS), a patterned rTMS variant mimicking natural theta rhythms (bursts at 50 Hz repeated at 5 Hz), accelerates effects and shortens session times to 3-10 minutes while maintaining efficacy comparable to traditional rTMS. For auditory hallucinations in schizophrenia, low-frequency rTMS over temporoparietal regions disrupts aberrant language networks, providing symptom relief in treatment-resistant cases.[54][55][56][57][58]Treatment parameters vary by indication but commonly include frequencies of 1-20 Hz for rTMS, with intensities set at 80-120% of the resting motor threshold to balance efficacy and safety. Sessions deliver 1,000-3,000 pulses over 20-40 minutes, typically administered 5 days per week for 20-30 sessions across 4-6 weeks, though accelerated schedules with multiple daily sessions are emerging for faster response. In treatment-resistant depression, response rates range from 30-50%, with remission in 20-30% of patients, outperforming sham in randomized trials and offering sustained benefits for 6-12 months in responders. Adverse effects are mild, including transient headache or scalp discomfort in 20-30% of cases, with rare seizures under standard protocols. By 2025, TMS is utilized in thousands of clinics worldwide, reflecting its integration into routine clinical practice.[52][59]
Transcranial Electrical Stimulation
Transcranial electrical stimulation (tES) is a non-invasive neuromodulation technique that applies weak electrical currents to the scalp to modulate brain activity, primarily targeting cortical regions. It encompasses variants such as transcranial direct current stimulation (tDCS), which delivers a constant low-intensity current to shift neuronal excitability, and transcranial alternating current stimulation (tACS), which uses oscillating currents to influence brain rhythms. Originating from animal studies in the 1960s, tES has evolved into a tool for both research and clinical applications, with early demonstrations of cortical polarization effects in rats showing lasting changes in neuronal firing.[60][61]The procedure involves placing saline-soaked sponge electrodes on the scalp, typically in a 2x1 montage for conventional tES, where the anode and cathode positions determine the direction of current flow. In tDCS, anodal stimulation depolarizes neurons to increase excitability, while cathodal stimulation hyperpolarizes neurons to reduce it, inducing polarity-dependent shifts in cortical activity without triggering action potentials since currents remain subthreshold. For tACS, sinusoidal currents at specific frequencies, such as 10 Hz for alpha-band entrainment, synchronize neural oscillations by aligning with endogenous rhythms, potentially enhancing perceptual or cognitive processing. These methods rely on the skull's conductivity to deliver currents of 0.5–2 mA, with sessions lasting 20–30 minutes to achieve measurable neuroplastic effects.[62][63][64]Mechanistically, tDCS modulates resting membrane potentials, facilitating or inhibiting synaptic transmission and long-term potentiation depending on polarity, which can persist for minutes to hours post-stimulation. tACS, in contrast, promotes entrainment of brain oscillations, such as alpha-band activity (8–12 Hz) in parieto-occipital regions, by phase-locking neural firing to the external stimulus, thereby influencing network synchronization without net polarization. These effects are subthreshold, avoiding direct neuronal depolarization, and are shaped by factors like current density and electrode montage. High-definition tES (HD-tES) refines this by using multi-electrode arrays (e.g., 4x1 ring configurations) to achieve more focal targeting, reducing unwanted spread to adjacent areas compared to conventional setups.[63][64]Common parameters for tDCS include currents of 1–2 mA applied via electrodes of 25–35 cm², with anodal placement over target regions like the dorsolateral prefrontal cortex for excitatory effects and durations of 20–30 minutes to balance efficacy and safety. Montages vary by goal, such as bifrontal for mood regulation or unilateral for motor enhancement, with current intensity influencing the magnitude of excitability changes. tACS parameters mirror these but specify frequency (e.g., 10 Hz for alpha entrainment) and waveform to match oscillatory targets, ensuring minimal skin irritation through saline conduction.[65][65]Applications of tES span cognitive enhancement, where tDCS over prefrontal areas boosts working memory and learning in healthy individuals, and therapeutic uses like reducing anxiety symptoms through dorsolateral prefrontal modulation or improving aphasia recovery post-stroke by facilitating language network plasticity when paired with speech therapy. HD-tES enhances focal applications, such as targeting intraparietal sulcus for attention deficits, offering precision for individualized interventions. Outcomes include representative improvements of 10–20% in task performance, as seen in category learning paradigms where anodal tDCS accelerated accuracy gains by up to 20.6% relative to sham. While promising, tES for depression remains investigational, with ongoing FDA trials for home-based devices as of 2025, and consumer units are available but not cleared for medical treatment, emphasizing the need for supervised use.[66][67][68][69][70]
Transcranial Focused Ultrasound Stimulation
Transcranial focused ultrasound stimulation (tFUS) is a non-invasive neuromodulation technique that employs acoustic waves to target and modulate brain activity with high spatial precision, enabling deeper penetration than other non-invasive methods like transcranial magnetic or electrical stimulation. The procedure utilizes phased-array transducers to generate and focus ultrasound beams through the intact skull, allowing for focal stimulation of brain regions without surgical intervention. In low-intensity modes, tFUS is applied for neuromodulation to alter neural excitability, while higher intensities enable thermal ablation for therapeutic lesioning. This approach has gained prominence for its ability to achieve millimeter-scale resolution and access subcortical structures, distinguishing it within the spectrum of non-invasive brainstimulation techniques.[71]The primary mechanisms of tFUS in neuromodulation involve non-thermal mechanical effects, where pressure waves from the ultrasound interact with neuronal membranes to activate mechanosensitive ion channels, such as Piezo1 and mechanosensitive channels of large conductance (MscL), leading to changes in cellular excitability and synaptic transmission. Thermal effects are minimal in low-intensity applications, as intensities are kept below levels that cause significant heating, though higher intensities can gate heat-sensitive channels like TRPV1 above 42°C for ablation purposes. These mechanical interactions can produce both excitatory and inhibitory outcomes depending on parameters, influencing ion channels, neurotransmitter release, and neural network activity without causing tissue damage. Seminal studies have demonstrated these effects in both animal models and human trials, highlighting tFUS's potential for reversible modulation.[72][71]Key parameters for tFUS neuromodulation include frequencies typically ranging from 0.5 to 1 MHz to optimize skull penetration and focal depth, spatial-peak pulse-average intensities below 720 mW/cm² to ensure non-thermal effects, and pulse durations on the millisecond scale to control the duration of stimulation. Applications encompass both ablative and modulatory uses; for instance, high-intensity tFUS received FDA approval in 2016 for essential tremor treatment via thalamotomy, disrupting tremor-generating circuits in the ventral intermediate nucleus. In addition to essential tremor, the FDA approved high-intensity tFUS in 2021 for tremor-dominant Parkinson's disease and in 2025 for staged bilateral treatment in advanced Parkinson's disease.[73][74] Investigational applications include psychiatric disorders like depression and addiction, where low-intensity tFUS targets regions such as the prefrontal cortex, as well as Alzheimer's disease, where it may disrupt amyloid plaques or open the blood-brain barrier for drug delivery.[71]Clinical outcomes demonstrate efficacy, with single-session ablative treatments for essential tremor achieving approximately 47% reduction in hand tremor severity at 3 months, as measured by the Clinical Rating Scale for Tremor, with effects persisting for years in many patients. For neuromodulation, early human trials since 2013 have shown mood improvements and altered cortical excitability, with ongoing investigations into addictiontreatment revealing potential reductions in craving-related activity. Advancements as of 2025 include the development of portable multi-focus systems, enabling real-time targeting and enhanced accessibility for clinical and research use, such as high-pressure multi-mode devices for precise, non-invasive delivery. The first human neuromodulationtrial occurred in 2013, marking the transition from preclinical to clinical application.[75][76][77]
Peripheral and Autonomic Neurostimulation
Vagus Nerve Stimulation
Vagus nerve stimulation (VNS) is a neuromodulation technique that delivers electrical impulses to the vagus nerve, primarily targeting its afferent fibers to modulate autonomic and central nervous system functions. Initially explored in the 1930s through electrical stimulation studies in animal models that demonstrated anti-epileptic effects, VNS evolved into a clinical therapy for refractory epilepsy, receiving FDA approval in 1997 as an adjunctive treatment for reducing seizure frequency in adults and later in children.[78][79] As of 2025, over 130,000 patients worldwide have received VNS implants, reflecting its established role in neurology.[80]The invasive procedure involves surgical implantation of a pulse generator in the chest subcutaneous tissue, connected via a lead wire to helical electrodes wrapped around the left cervical vagus nerve in the neck. These helical electrodes, consisting of two active contacts (anode and cathode) and an anchoring tether, encircle the nerve to ensure stable stimulation without damaging its structure. The surgery typically lasts 1-2 hours under general anesthesia, with stimulation parameters programmed postoperatively via a wirelesswand. Non-invasive transcutaneous auricular VNS (taVNS) applies surface electrodes to the tragus of the ear, targeting the auricular branch of the vagus nerve, allowing outpatient use without surgery.[81][82][83]Mechanistically, VNS primarily activates the approximately 80% afferent sensory fibers of the vagus nerve, which project to the nucleus tractus solitarius (NTS) in the brainstem. From the NTS, signals propagate to interconnected brainstem nuclei, such as the locus coeruleus and raphe nuclei, and higher cortical networks, including the thalamus and limbic system, influencing neurotransmitter release (e.g., norepinephrine, serotonin) and autonomic balance. This pathway underpins VNS's therapeutic effects across neurological and inflammatory conditions.[84][85][86]Clinically, VNS is FDA-approved for drug-resistant epilepsy and treatment-resistant depression (approved 2005), where it serves as an adjunct to medications. In epilepsy, long-term use leads to ≥50% seizure reduction in about 50% of patients, with efficacy improving over years. For depression, it enhances mood via central projections, showing sustained response rates in refractory cases. Emerging applications include anti-inflammatory effects in rheumatoid arthritis, where VNS inhibits cytokine production via the cholinergic anti-inflammatory pathway, and stroke rehabilitation, promoting neuroplasticity and motor recovery through paired stimulation with physical therapy.[87][88][89]Stimulation parameters are tailored to balance efficacy and tolerability, typically including pulse widths of 250-500 μs, frequencies of 20-30 Hz, and duty cycles of 30 seconds on followed by 5 minutes off, with current intensities starting low (0.25-1.25 mA) and titrated upward. These settings optimize afferent activation while minimizing side effects like hoarseness or cough. In taVNS protocols for migraines, similar frequencies are used, yielding 20-30% reductions in attack frequency and intensity in responsive patients.[90][91][92]Overall outcomes highlight VNS's safety profile, with common adverse events (e.g., voice alteration, dyspnea) decreasing over time and no increased mortality risk. In epilepsy cohorts, responder rates (≥50% seizure reduction) reach 46-65% at 1-5 years, alongside quality-of-life improvements. For taVNS in migraines, clinical trials report significant headache day reductions (e.g., 2-4 days/month), positioning it as a non-pharmacological option. Ongoing research emphasizes personalized programming to enhance these benefits.[93][94][95]
Transcutaneous Electrical Nerve Stimulation
Transcutaneous electrical nerve stimulation (TENS) is a non-invasive technique that delivers low-voltage electrical currents through electrodes placed on the skin to stimulate peripheral nerves, primarily for pain management. Developed in the 1970s following the introduction of the gate control theory of pain by Melzack and Wall in 1965, TENS aims to modulate pain perception by interfering with nociceptive signals in the spinal cord.[96] The procedure involves applying self-adhesive surface electrodes directly over or near the affected nerves or painful area, with placement adjusted based on the target site; for example, supraorbital TENS for headaches positions electrodes across the forehead to target the supraorbital branch of the trigeminal nerve.[97] Sessions typically last 20-60 minutes, and devices are battery-powered for ease of use.[98]The primary mechanisms of TENS involve the activation of large-diameter A-beta sensory fibers, which excite inhibitory interneurons in the spinal cord dorsal horn, effectively "gating" the transmission of smaller-diameter A-delta and C-fiber pain signals to higher brain centers—a process rooted in the gate control theory.[99] High-frequency TENS primarily relies on this non-opioid segmental mechanism, while low-frequency TENS additionally promotes the release of endogenous opioids, such as beta-endorphins, from the central nervous system, contributing to broader analgesia.[100] These dual pathways allow TENS to provide both immediate sensory modulation and longer-lasting biochemical effects without systemic side effects associated with pharmacological agents.[101]TENS parameters are tailored to optimize therapeutic effects, with frequencies ranging from 2 to 150 Hz and pulse widths of 50-250 microseconds; intensity is adjusted to the patient's sensory or tolerance threshold, producing a tingling or comfortable muscle contraction without discomfort.[97] Conventional TENS uses high frequencies (50-150 Hz) at sensory intensities for acute pain relief, whereas acupuncture-like TENS employs low frequencies (2-10 Hz) at higher intensities to mimic needle stimulation and enhance opioid-mediated responses. These settings can be programmed on portable, user-friendly devices that enable home-based treatment, improving accessibility for chronic conditions.[102]Clinical applications of TENS include chronic neuropathic pain, such as painful diabetic neuropathy, where systematic reviews indicate moderate evidence of reduced pain intensity when applied at adequate intensities compared to placebo.[99] For osteoarthritis, particularly knee pain, TENS provides tentative evidence of symptom relief and improved function, though results vary across studies and are more pronounced when combined with exercise.00973-0/fulltext) Supraorbital TENS, exemplified by the Cefaly device cleared by the FDA in 2014 for migraine prevention in adults, targets trigeminal nerve branches to reduce headache frequency and severity.[103]Outcomes from TENS generally show 30-50% reductions in acute pain intensity immediately post-treatment, with sustained benefits in chronic cases depending on consistent use; for instance, meta-analyses report clinically meaningful relief (≥30% pain reduction) in various musculoskeletal and neuropathic conditions.[104] Portable TENS units facilitate daily self-administration, enhancing patient adherence and quality of life.
Sacral nerve stimulation (SNS), also known as sacral neuromodulation, involves the implantation of a device that delivers electrical impulses to the sacral nerves to modulate pelvic organ function, with initial clinical programs beginning in the early 1980s following preclinical work in the 1970s by researchers such as Tanagho and Schmidt demonstrating detrusor contractions in animal models.[105][106]Pudendal nerve stimulation represents a targeted variant, focusing on the pudendal nerve branches to enhance sphincter control and bladder coordination, often explored as an alternative or adjunct for refractory cases.[107] The U.S. Food and Drug Administration first approved SNS in 1997 for the treatment of urge urinary incontinence, marking a milestone in neuromodulation for pelvic disorders.[108]The procedure typically begins with a percutaneous nerve evaluation (PNE) phase, where a temporary lead is inserted through the S3 sacral foramen under local anesthesia and fluoroscopic guidance to assess efficacy over 1-2 weeks, allowing patients to evaluate symptom improvement before committing to permanent implantation.[109] If successful, a tined quadripolar lead is percutaneously placed to anchor it securely, with electrodes positioned such that contacts 2 and 3 straddle the foramen for optimal nerve contact, followed by connection to an implantable pulse generator in the gluteal region.[109][110] This staged approach minimizes risks and ensures reversibility, with the entire process being minimally invasive compared to earlier surgical techniques.[109]Mechanistically, SNS modulates sacral afferent pathways to inhibit detrusor overactivity by altering spinal reflex arcs and reducing C-fiber mediated bladder sensations, thereby promoting bladder storage without compromising voiding efficiency.[109][111] In pudendal nerve stimulation, electrical impulses facilitate relaxation of the external urethral sphincter while enhancing bladder contraction coordination, potentially via somatic afferent inhibition of micturition reflexes, which is particularly useful in dyssynergic conditions.[112][113]Primary applications of SNS include refractory overactive bladder (OAB) with urgency incontinence, non-obstructive urinary retention, fecal incontinence, and interstitial cystitis/bladder pain syndrome, where it restores normal pelvic nerve signaling to improve continence and reduce urgency episodes.[109][114] Pudendal nerve stimulation is particularly indicated for neurogenic bladder dysfunction, such as in spinal cord injury or tethered cord syndrome, where it aids in managing detrusor-sphincter dyssynergia and improving voiding efficiency.[107][115]Stimulation parameters are individualized but commonly involve frequencies of 10-20 Hz to optimize reflex modulation, amplitudes adjusted to the sensory threshold (typically 0.5-3 V), and pulse widths around 210 μs, often delivered in intermittent cycling modes (e.g., 1-5 minutes on/off) to conserve battery life and enhance long-term efficacy.[116][117] Higher frequencies (up to 50 Hz) may be used for specific neurogenic cases to further inhibit detrusor activity.[118]Clinical outcomes demonstrate substantial benefits, with approximately 70% of patients achieving at least 50% improvement in urge incontinence episodes and overall continence rates exceeding 60% at long-term follow-up, alongside significant enhancements in quality of life for fecal incontinence.[119][120] For pudendal variants, studies report up to 52% reduction in symptom scores for neurogenic lower urinary tract dysfunction.[121]As of 2025, innovations in minimally invasive percutaneous SNS include battery-free systems like the Neuspera device, which received FDA approval in June 2025 and uses wireless external charging to deliver ultra-miniaturized stimulation directly via percutaneous leads, with six-month pivotal trial data demonstrating efficacy comparable to established sacral neuromodulation therapies for urgency urinary incontinence and reduced surgical burden.[122][123]
Sensory Neurostimulation
Cochlear Implants
Cochlear implants are neurostimulation devices designed to restore hearing in individuals with profound sensorineural hearing loss by directly interfacing with the auditory system. The system consists of an external component, including a microphone that captures sound and a speech processor that converts it into digital signals, and an internal component comprising a receiver surgically implanted under the skin behind the ear and an electrode array threaded into the cochlea. The electrode array is typically inserted through a cochleostomy into the scala tympani of the cochlea, allowing precise placement to stimulate surviving auditory nerve fibers. This procedure, performed under general anesthesia, usually takes 2-3 hours and involves creating a small incision behind the ear to access the mastoid bone and cochlea.[124][125]The mechanism of cochlear implants involves bypassing the damaged or absent hair cells in the cochlea, which normally transduce mechanical sound vibrations into neural signals, by delivering direct electrical stimulation to the spiral ganglion neurons of the auditory nerve. The speech processor encodes acoustic information into electrical pulses that are transmitted wirelessly via a coil to the internal receiver, which then activates specific electrodes along the array to evoke patterned neural activity corresponding to sound frequencies. This electrical stimulation mimics the tonotopic organization of the cochlea, with basal electrodes representing higher frequencies and apical ones lower frequencies, thereby enabling perception of speech and environmental sounds. Stimulation parameters typically include biphasic pulses with phase durations of 10-400 μs, pulse rates up to 1,000 pulses per second (pps) per channel, and 12-24 electrodes to provide multi-channel input for spectral resolution.[126][127][128]Cochlear implants are primarily indicated for individuals with bilateral profound sensorineural hearing loss, where traditional hearing aids provide insufficient benefit, and are approved for pediatric use starting from 9 months of age in cases of severe-to-profound deafness. In adults, implantation follows confirmation of poor aided speech recognition, often below 50% on sentence tests, while children benefit from early intervention to capitalize on developmental windows. The first multi-channel cochlear implant received FDA approval in 1985, and by 2022, over 1 million devices had been implanted worldwide. Outcomes demonstrate substantial improvements, with approximately 80% of adult recipients achieving open-set sentence recognition in quiet environments post-implantation, enabling conversational speech understanding without visual cues. In children, auditory brain plasticity plays a critical role, as early implantation (before age 2) promotes reorganization of central auditory pathways, leading to near-normal language acquisition and speech production in many cases when combined with intensive rehabilitation.[129][130][131][132]
Visual Prostheses
Visual prostheses, also known as retinal prostheses or bionic eyes, are implantable neurostimulation devices designed to restore partial vision in individuals with severe visual impairment caused by retinal degenerative diseases. These systems work by electrically stimulating surviving inner retinal cells or the optic nerve, converting external visual information into patterns of perceived light known as phosphenes. Unlike natural vision, the restored perception is typically low-resolution and monochromatic, enabling basic detection of light, motion, and large objects rather than fine details.[133][134]The primary types of visual prostheses are epiretinal and subretinal implants. In epiretinal prostheses, such as the Argus II system, a microelectrode array is surgically attached to the inner surface of the retina over the macular region, secured by a retinal tack. An external wearable component includes a camera mounted on glasses that captures images, a video processing unit that converts them into electrical signals, and a transmitter coil that wirelessly delivers these signals to the implanted receiver. Subretinal prostheses, like the PRIMA system, are positioned beneath the retina in the subretinal space, often using photovoltaic arrays that directly convert incident light into electrical stimulation without requiring an external power source. Surgical implantation typically involves a vitrectomy and precise placement to avoid damage to the optic nerve or choroid, with recovery periods ranging from days to weeks.[133][134][135]These devices bypass damaged photoreceptors by directly stimulating inner retinal neurons, thereby preserving the natural signal processing pathways to the brain. Epiretinal implants primarily target retinal ganglion cells, generating action potentials that propagate along the optic nerve. Subretinal implants, in contrast, stimulate bipolar and remaining photoreceptor cells in the inner nuclear layer, potentially recruiting more preserved retinal circuitry for improved spatial resolution. In both cases, the stimulation evokes phosphenes that correspond to electrode activation patterns, effectively creating a pixelated visual field. Optic nerve stimulation variants, though less common in retinal prostheses, directly target axonal bundles for broader field coverage.[134][133]Visual prostheses are primarily applied to treat retinitis pigmentosa (RP) and age-related macular degeneration (AMD), conditions where photoreceptor loss leads to profound blindness while sparing inner retinal layers. The Argus II Retinal Prosthesis System, developed by Second Sight Medical Products, targets adults aged 25 or older with severe to profound RP and bare or no light perception in both eyes. Approved by the U.S. Food and Drug Administration (FDA) in 2013 as a humanitarian device, it was the first such implant to receive regulatory clearance, with over 350 units implanted worldwide before production ceased in 2019 due to company challenges. Following the company's bankruptcy in 2020, patients faced significant issues, including lack of technical support and repairs, leading to device obsolescence and vision loss for some recipients. While not FDA-approved for AMD, subretinal systems like PRIMA show promise for late-stage dry AMD by targeting central vision restoration.[136][133][137][138]Stimulation parameters are tuned to safely elicit phosphenes while minimizing tissue damage, typically using charge-balanced biphasic pulses delivered via platinumelectrodes. Representative settings include pulse frequencies of 2-20 Hz to match natural firing rates, currents ranging from 50-200 μA per electrode to thresholdperception without exceeding charge density limits (around 30 μC/cm²/phase), and pulse widths of 0.45-1 ms. Resolution is constrained by electrode count and spacing; the Argus II features 60 electrodes with 525 μm diameter and 200 μm spacing, yielding a visual field of approximately 20 degrees. Higher-density arrays in emerging prototypes aim for hundreds to thousands of electrodes to enhance acuity.[139][134][133]Clinical outcomes demonstrate modest but meaningful vision restoration, with patients achieving basic functional improvements over no-light-perception baselines. In Argus II trials, over 90% of subjects could detect light and motion, perform square localization tasks with 60-80% accuracy, and recognize large objects or letters at distances up to 1 meter when the device was active. Grating visual acuity reached approximately 1.8 logMAR (equivalent to 20/1260 Snellen), enabling orientation and mobility in controlled environments. Five-year follow-up data confirmed sustained benefits in visual tasks, though long-term device usage varied due to comfort and cognitive demands, with average satisfaction ratings around 6/10. Subretinal implants like PRIMA have reported acuities up to 20/460, and as of 2025, clinical trials have shown further improvements, with some patients achieving up to 20/42 equivalent acuity using digital enhancements for tasks like reading.[140][134]As of 2025, hybrid approaches combining visual prostheses with gene therapies are emerging to address limitations in resolution and longevity. These integrate optogenetic sensitization—using viral vectors to express light-sensitive proteins in retinal cells—with prosthetic stimulation, potentially amplifying phosphene quality in RP and AMD. Early preclinical and Phase I trials, such as those exploring adeno-associated virus (AAV)-based enhancements, indicate improved cellular responsiveness, paving the way for synergistic devices that could extend to broader degenerative conditions.[141][133]
Retinal and Optic Nerve Stimulation
Retinal and optic nerve stimulation represent targeted approaches within neurostimulation for restoring vision in conditions where the visual pathway is disrupted, particularly in degenerative diseases affecting the outer retina or nerve fibers. These techniques bypass damaged photoreceptors or ganglion cells by directly activating surviving inner retinal elements or axonal bundles, eliciting phosphenes—perceived points of light that form the basis of artificial vision. Unlike broader visual prostheses, retinal stimulation employs subretinal implants to interface closely with bipolar and amacrine cells, while optic nerve stimulation uses cuff electrodes to engage bundles of optic nerve fibers, enabling cortical-independent visual signaling.Subretinal photovoltaic arrays, such as those in the PRIMA system, consist of dense silicon photodiode arrays implanted beneath the retina, converting projected near-infrared light from external goggles into localized electrical currents for stimulation. This mechanism directly activates bipolar cells in the inner nuclear layer, generating retinal ganglion cell spikes that propagate to the brain without requiring batteries, as power and visual data are delivered optically via pulsed illumination at safe irradiances below 10 mW/mm². Applications focus on advanced retinitis pigmentosa (RP) and age-related macular degeneration (AMD), where inner retinal layers remain viable, allowing restoration of central vision for tasks like reading large print or object recognition. Clinical outcomes demonstrate meaningful improvements, with 80% of patients achieving enhanced visual acuity (up to 20/440) and contrast sensitivity after 12 months, enabling functional activities such as face detection in simulated environments. As of 2025, further trial data indicate additional gains, with some patients reaching 20/42 equivalent acuity using enhancements.Optic nerve stimulation employs multi-electrode cuff devices, typically self-sizing spiral cuffs with four contacts wrapped around the intraorbital optic nerve, to deliver electrical pulses that selectively activate axonal fiber bundles. The mechanism involves varying pulse durations (0.5-1 ms) and frequencies to produce patterned phosphene clusters, mimicking topographic organization and bypassing retinal degeneration while relying on intact central visual pathways. These cuffs, powered wirelessly via inductive coupling from an external unit, are applied in cases of optic nerve damage, including advanced glaucoma or traumatic optic neuropathy, where retinal elements are preserved but axonal transmission is impaired. Seminal trials began in the late 1990s, with the first human implants in 1998 demonstrating stable phosphene elicitation in blind patients, allowing shape and motion discrimination. Animal models have shown over 100 distinct phosphenes per session, correlating with improved spatial resolution. As of 2025, ongoing feasibility studies explore non-invasive optic nerve stimulation approaches for glaucoma, reporting preliminary enhancements in visual field sensitivity through targeted fiber recruitment at stimulation frequencies matched to the visual flicker fusion threshold of 50-60 Hz.[142]
Systemic and Functional Stimulation
Cardiac Electrotherapy Devices
Cardiac electrotherapy devices encompass implantable systems such as pacemakers and implantable cardioverter-defibrillators (ICDs) that deliver electrical impulses to regulate heart rhythm by stimulating cardiac conduction pathways. The first implantable pacemaker was surgically placed on October 8, 1958, in Sweden by surgeon Åke Senning and engineer Rune Elmqvist, marking the inception of modern cardiac pacing therapy.[143] Leadless pacemakers, which eliminate traditional wired leads to reduce complications, received initial U.S. Food and Drug Administration (FDA) approval in April 2016 for treating certain bradycardias.[144] These devices operate on principles of electrical stimulation to restore synchronized cardiac activity, adapting to patient needs through programmable outputs.Implantation procedures for these devices typically involve transvenous approaches, where leads are inserted via percutaneous venous access—often through the subclavian or cephalic vein—and advanced fluoroscopically to the right atrium and/or ventricle for pacing or sensing.[145] For subcutaneous ICDs, the generator is placed under the skin along the left mid-axillary line below the armpit, with a sensing and defibrillating electrode tunneled subcutaneously parallel to the sternum from the xiphoid process to the manubrium, avoiding intravascular placement to minimize infection risks.[146] These methods ensure precise electrode positioning while balancing procedural efficiency and patient safety.Mechanistically, pacemakers deliver low-energy pulses to depolarize myocardial tissue, with configurations designed to avoid unintended phrenic nerve stimulation—such as multipolar left ventricular pacing vectors that adjust electrode spacing or output to exceed pacing thresholds without diaphragmatic capture.[147] ICDs extend this by detecting tachyarrhythmias and delivering high-energy biphasic shocks, where current flows initially in one direction before reversing polarity mid-delivery, reducing defibrillation energy requirements by 20-40% compared to monophasic waveforms through more uniform transmembrane potential changes.[148]Applications include treating bradycardia via demand pacing to maintain adequate heart rates, preventing ventricular tachycardia or fibrillation in high-risk patients with ICDs, and cardiac resynchronization therapy (CRT) for heart failure by synchronizing ventricular contractions through biventricular pacing.[149] Device parameters are tailored post-implantation, with typical lower pacing rates set at 60 beats per minute (bpm) to mimic resting sinus rhythm and upper rates limited to 120 bpm during tracking to prevent excessive ventricular rates; sensing thresholds are programmed to detect intrinsic signals above 2-5 millivolts, while anti-tachycardia pacing delivers bursts of 8-15 pulses at 88% of the tachycardia cycle length to terminate reentrant arrhythmias non-invasively.[150][151]Clinical outcomes demonstrate substantial benefits, with primary prevention ICDs improving survival in high-risk heart failure patients by reducing sudden cardiac death, though the absolute benefit diminishes in those over 75 years due to comorbidities.[152] By 2025, remote monitoring via smartphone apps—such as Medtronic's MyCareLink or Abbott's myMerlinPulse—enables daily transmission of device data including arrhythmias and battery status, facilitating early intervention and reducing clinic visits by up to 50%.[153]
Functional Electrical Stimulation
Functional electrical stimulation (FES) is a neurostimulation technique that applies controlled electrical impulses to peripheral nerves and muscles to elicit functional movements in individuals with neuromuscular impairments, such as those resulting from stroke or spinal cord injury (SCI).[154] By artificially activating motor nerves, FES bypasses disrupted neural pathways to produce coordinated muscle contractions, thereby restoring or augmenting voluntary motor function during activities like walking or grasping.[154] This approach differs from basic electrical muscle stimulation by synchronizing impulses with the user's intended movements, often through sensorfeedback, to facilitate natural gait patterns or limb control.[155]The procedure for FES typically involves placing electrodes on the skin surface or implanting them near target nerves or muscles. Surface electrodes, such as self-adhesive pads, are commonly used for non-invasive applications and positioned over key muscle groups like the quadriceps or peroneal nerve; these deliver transcutaneous stimulation without surgery.[156] For more precise control, cuff electrodes can be wrapped around peripheral nerves, providing targeted activation while minimizing skin irritation.[154] In rehabilitation settings, FES is integrated into devices like stationary bikes, where electrodes stimulate leg muscles in rhythm with pedaling to promote cycling exercise and cardiovascular fitness in paralyzed limbs.[157]At its core, FES mechanisms rely on depolarizing motor axons to induce timed, sequential contractions in paralyzed muscles, mimicking physiological recruitment patterns. Electrical pulses propagate along the nerve to the neuromuscular junction, triggering muscle fiber activation and force generation without requiring central nervous system input.[154] This process enhances motor relearning by pairing stimulation with residual voluntary effort, potentially promoting neuroplasticity through repeated afferent feedback to the brain.[155] In cases of denervation or incomplete injury, FES also maintains muscle excitability by countering disuse-related changes.FES finds primary applications in restoring mobility for conditions like stroke-induced hemiplegia, where it aids upper and lower limb recovery by facilitating arm reaching or leg extension during therapy.[158] For SCI patients, it supports gait training through multi-channel systems that coordinate hip, knee, and ankle activation to simulate walking on treadmills or overground.[159] A key use is in drop foot orthoses, where peroneal nerve stimulation during the swing phase lifts the foot, improving clearance and reducing tripping risks in ambulatory individuals with foot weakness.[160]Stimulation parameters are tailored to optimize contraction while minimizing fatigue and discomfort. Biphasic pulses, which alternate positive and negative phases for charge balance and tissue safety, typically have durations of 200-500 μs and amplitudes adjusted to motor threshold.[161] Frequencies range from 20-50 Hz to achieve fused tetanic contractions for sustained force, with lower rates (10-20 Hz) used intermittently to reduce fatigue in SCI cases.[154] Advanced systems incorporate electromyography (EMG) feedback or inertial sensors to coordinate stimulation timing with gait cycles, ensuring precise synchronization.[162]Clinical outcomes demonstrate FES's efficacy in enhancing functional independence, with studies reporting 15-25% increases in walking speed for stroke and SCI patients after 4-12 weeks of gait training, alongside improved balance and endurance.[163] It also prevents muscle atrophy by preserving fiber cross-sectional area and strength, reducing secondary complications like joint contractures in non-weight-bearing limbs.[155] Long-term use in drop foot applications has shown sustained reductions in fall incidence and energy expenditure during ambulation.[160]FES originated in the 1960s, pioneered for paraplegic individuals to enable standing via quadriceps stimulation, marking early efforts to restore lower limb function post-SCI.[164] By 2025, AI-integrated FES systems have advanced the field, using machine learning for real-time gait prediction and adaptive pulse modulation to enhance precision in foot drop correction and hybrid robotic therapies.[165]
Technologies and Devices
Microelectrode and Implantable Technologies
Microelectrode technologies form the cornerstone of implantable neurostimulation systems, enabling precise electrical interfacing with neural tissues for applications such as deep brain stimulation (DBS) and spinal cord stimulation (SCS). These devices typically employ platinum-iridium alloys as electrode materials due to their high corrosion resistance, biocompatibility, and ability to sustain chronic implantation without significant degradation. Platinum-iridium electrodes facilitate safe charge delivery by supporting reversible electrochemical reactions, minimizing tissue damage during stimulation.To address limitations in traditional metallic electrodes, such as high impedance that reduces signal efficiency, carbon nanotubes (CNTs) have emerged as advanced coating materials, lowering impedance by up to 90% while enhancing charge transfer at the electrode-tissue interface. Flexible polymers like polyimide are widely used as substrates for these microelectrodes, providing mechanical compliance to match the softness of neural tissue and reducing inflammatory responses from mechanical mismatch. Polyimide-based designs allow for multi-channel arrays that conform to curved brain surfaces, improving long-term stability in cortical implants.Key designs include the Utah array, a silicon-based microelectrode shank with up to 100 penetrating electrodes for penetrating cortical recording and stimulation, and the Michigan probe, which features micromachined needles on a flexible base for targeted neural access. High-density arrays exceeding 1,000 channels, such as those developed for brain-machine interfaces, enable simultaneous stimulation and recording from large neural populations, supporting advanced prosthetic control. The first microelectrode arrays for neural interfacing were pioneered in the 1970s, with early silicon-based prototypes demonstrating feasibility for extracellular recordings.Innovations in implantable technologies include bioresorbable implants made from materials like magnesium or silk fibroin, which dissolve harmlessly in the body after delivering targeted stimulation, eliminating the need for surgical removal. Wireless power delivery via radiofrequency (RF) coils allows untethered operation of implants, reducing infection risks from percutaneous wires and enabling deeper tissue targeting. A notable recent advancement is neural dust sensors, miniaturized (sub-millimeter) piezoelectric devices for minimally invasive recording and stimulation, with prototypes achieving wireless ultrasound powering by 2025.Despite these advances, challenges persist, including tissue encapsulation where glial scarring forms a barrier around electrodes, increasing impedance and attenuating signals within months of implantation. Electrode impedance typically increases rapidly in the initial weeks post-implantation due to acute inflammatory responses, protein adsorption, and early glial scarring, often followed by stabilization or reduction over longer periods as the tissue response matures and stimulation protocols adapt; fibrosis contributes to long-term variability, with changes reported up to several hundred percent initially but averaging stabilization after months. A critical parameter in these systems is charge injection capacity, defined as Q = I \times t, where Q is the total charge, I is the current, and t is the pulse duration; this is typically limited to 10-50 nC per phase to prevent irreversible Faradaic reactions and neural damage.[166]
Non-Invasive Device Innovations
Non-invasive neurostimulation devices have evolved to prioritize portability and user independence, enabling treatments outside clinical environments through external, skin-applied technologies. These innovations emphasize ergonomic designs that integrate seamlessly into daily routines, reducing barriers to adoption for conditions like chronic pain, migraines, and cognitive deficits. By leveraging wireless connectivity and advanced materials, recent developments have enhanced efficacy and comfort while maintaining safety standards.[167]Key designs include wearable transcranial magnetic stimulation (TMS) helmets, which deliver repetitive magnetic pulses to targeted brain regions via helmet-mounted coils, allowing mobile sessions without fixed equipment. Home transcranial direct current stimulation (tDCS) headsets, such as lightweight, adjustable models with saline-soaked sponges or integrated electrodes, facilitate self-administered low-intensity current application for mood or focus enhancement. Ultrasound caps for transcranial focused ultrasound (tFUS) use wearable arrays of transducers to focus acoustic waves on deep neural structures, offering precise, non-thermal modulation without electromagnetic interference.[168][169][170][171][172][173][174]Innovative features extend to Bluetooth-enabled transcutaneous electrical nerve stimulation (TENS) units, which connect to smartphone apps for customizable pulse patterns and remote monitoring during pain management. AI-optimized transcutaneous auricular vagus nerve stimulation (taVNS) earpieces employ machine learning algorithms to adjust stimulation based on real-time physiological feedback, targeting anxiety or inflammation via ear clip electrodes. Portable devices like gammaCore provide handheld, non-invasive vagus nerve stimulation through the neck, delivering short electrical bursts to abort migraine attacks, with rechargeable models supporting multiple daily uses.[175][176][177][178]Material advancements focus on user comfort and longevity, with dry electrodes—composed of conductive polymers or nanocomposites—eliminating the need for conductive gels, thus preventing residue and improving wearability during extended sessions. Hydrogel interfaces, featuring adhesive, biocompatible layers, minimize skinirritation by providing flexible, moisture-retaining contact that conforms to body contours without causing allergic reactions. These materials support prolonged, residue-free application in both stationary and ambulatory settings.[179][180][181][182][183][184]Accessibility has improved through regulatory pathways and economic factors, with many over-the-counter devices classified as FDA Class II, requiring 510(k) clearance for moderate-risk neuromodulation but allowing direct consumer purchase without prescription. Costs have decreased to the $100-500 range for entry-level tDCS and TENS models, driven by mass production and direct-to-consumer sales, broadening availability for home use. The first consumer tDCS devices emerged in the 2010s, marking a shift from clinical-only tools to accessible wellness products. By 2025, augmented reality (AR)-integrated stimulation systems have appeared for cognitive training, overlaying visual cues with synchronized tDCS to enhance learning protocols. A core concept in these devices is closed-loop feedback, where integrated electroencephalography (EEG) or electromyography (EMG) sensors dynamically adjust stimulation parameters in response to neural or muscular signals, optimizing outcomes in real time.[185][186][170][171][187][176][188][189]
Limitations and Safety
Surgical and Procedural Risks
Surgical risks in invasive neurostimulation procedures, such as deep brain stimulation (DBS) and spinal cord stimulation (SCS), primarily encompass perioperative complications that can necessitate revisions or additional interventions. Infection rates for these implants typically range from 1% to 5%, often involving skin flora like Staphylococcus species, which may require hardware removal in severe cases.[190][191] Intracranial or epidural hemorrhage occurs in approximately 1-3% of DBS cases, posing risks of neurological deficits or fatality due to vascular injury during electrode insertion.[192] Electrode migration, a hardware-related issue, affects up to 10% of patients, leading to loss of therapeutic efficacy and revision rates as high as 10-20% in SCS implants.[193][194]Anesthesia approaches vary by procedure to balance patient comfort and surgical precision. In DBS, microelectrode recording is often performed under local anesthesia to allow intraoperative physiological mapping, minimizing risks like disorientation but potentially causing patient anxiety.[195] Conversely, SCS implantation frequently uses general anesthesia, which reduces motion artifacts but increases the potential for respiratory complications or delayed emergence.[196] Intraoperative issues, such as off-target stimulation, can induce transient adverse events including seizures in 0.5-2% of DBS cases or acute hypertension from unintended autonomic activation.[197]Patient selection is critical to mitigate procedural hazards, with absolute contraindications including active coagulopathy, which elevates bleeding risks, and uncontrolled infections.[198] Post-implantation MRI compatibility must be ensured, as non-conditional devices can cause lead heating or displacement, prohibiting scans unless under specific protocols.[199]Mitigation strategies have significantly lowered risks over time. Stereotactic navigation systems achieve targeting accuracy below 1 mm, reducing placement errors and associated hemorrhages.[200] Prophylactic antibiotics, administered preoperatively in over 90% of cases, decrease infection incidence by up to 50% through protocols targeting common pathogens.[201] Overall complication rates have declined by approximately 50% since the early 2000s, attributed to advanced imaging like intraoperative CT and MRI integration.[202] As of 2025, robotic-assisted implantation platforms enhance precision in DBS and SCS, reducing targeting errors and potentially lowering revision rates through improved accuracy and real-time adjustments.[203]
Adverse Effects and Long-Term Concerns
Neurostimulation therapies, while effective for various neurological conditions, are associated with several chronic neurological side effects. In deep brain stimulation (DBS), patients may experience dysarthria or mood swings in approximately 5-10% of cases, often linked to electrode placement and stimulation parameters affecting speech and emotional regulation pathways.[204] Similarly, spinal cordstimulation (SCS) can lead to tolerance buildup over time, necessitating frequent reprogramming to maintain efficacy as neural adaptation reduces pain relief.[205]For non-invasive methods, such as repetitive transcranial magnetic stimulation (rTMS), common side effects include headache (up to 30%) and scalp discomfort, with rare risks of seizure (0.1-0.2% in screened patients). Transcranial direct current stimulation (tDCS) typically causes mild transient effects like tingling or itching, with no serious long-term adverse events reported in clinical use.[206]Systemic adverse effects also arise with prolonged use. Vagus nerve stimulation (VNS) has been associated with weight gain in some patients, potentially due to altered metabolic signaling, alongside risks of infection recurrence at the implant site.[207] Hoarseness is a common but typically mild side effect in VNS, occurring during stimulation cycles.[208]Long-term concerns include device-related degradation and potential psychological dependencies. Electrode corrosion, particularly of platinum components, can occur with chronic electrical pulsing, leading to tissue reactions and reduced performance over years of implantation.[209] Battery depletion in implantable neurostimulators typically requires replacement every 5-10 years, involving additional surgical interventions.[210] Dependency risks emerge as patients may develop reliance on stimulation for daily functioning, complicating discontinuation.[211]Ethical considerations surround access disparities and off-label applications. High costs and limited availability exacerbate inequities, particularly in low-resource settings, restricting neurostimulation to privileged populations.[212] Off-label use for cognitive enhancement raises concerns about unintended consequences and equitable regulation.[213] The reversibility of neuroplasticity changes induced by chronic stimulation remains uncertain, as some adaptations may persist post-cessation.[214]Ongoing monitoring is essential to mitigate these issues, with annual follow-ups recommended to assess devicefunction, side effects, and efficacy adjustments.[215] Explant rates range from 5-15%, often driven by persistent adverse effects or loss of benefit.[216]Animal studies as of 2025 indicate gene expression changes from chronic stimulation, such as alterations in GABAergic pathways.[217]
History
Early Discoveries and Milestones
The earliest recorded applications of electrical stimulation for therapeutic purposes trace back to ancient civilizations, where the electric torpedo fish (Torpedo marmorata) was employed by Roman physicians to alleviate pain. In the 1st century AD, Scribonius Largus documented the use of these fish placed on patients' heads or affected areas to treat headaches and gout, leveraging the natural electric discharges to induce numbness and relief.[218] This rudimentary form of neurostimulation persisted into later eras, foreshadowing the scientific exploration of bioelectricity.In the late 18th century, Italian anatomist Luigi Galvani's experiments laid the groundwork for understanding inherent electrical activity in living tissues. Galvani demonstrated in 1791 that frog legs contracted when exposed to electrical sparks, attributing the phenomenon to "animal electricity" generated within the nerves and muscles, a discovery that confirmed the bioelectric basis of neural function and inspired subsequent advancements in electrophysiology.[219] Building on this, Galvani's nephew Giovanni Aldini extended the work to human applications in 1804, applying voltaic currents from batteries to the heads and limbs of executed criminals, eliciting facial expressions, limb movements, and respiratory-like actions, which hinted at the potential for electrical intervention in neurological disorders.[220] Mid-19th-century contributions from German physiologist Emil du Bois-Reymond further advanced the field; in the 1840s, he pioneered precise recordings of electrical currents in nerves and muscles using sensitive galvanometers, establishing electrophysiology as a discipline and demonstrating that nerve impulses were electrochemical in nature.[221]The late 19th and early 20th centuries marked pivotal milestones in cortical neurostimulation. In 1870, German neurologists Gustav Fritsch and Eduard Hitzig conducted groundbreaking experiments on dogs, showing that direct electrical stimulation of specific frontal lobe regions elicited contralateral muscle contractions, thereby mapping the motor cortex and overturning prior beliefs that the cerebral surface was insensible to electricity.[222] Canadian neurosurgeon Wilder Penfield expanded this in the 1930s through intraoperative stimulation during epilepsy surgeries, systematically mapping sensory and motor areas in awake patients to create the iconic cortical homunculus, which illustrated the somatotopic organization of the brain and informed safe resection techniques.[223] Concurrently, observations in the 1930s revealed the anticonvulsant potential of vagus nerve stimulation (VNS); studies by Bailey and Bremer in 1938 demonstrated that electrical activation of the vagus nerve in animals altered electrocortical activity, suppressing seizure-like patterns and suggesting modulatory effects on brain excitability.[224]By the mid-20th century, deeper brain interventions emerged as precursors to modern deep brain stimulation (DBS). In the 1940s, stereotactic thalamotomy procedures, involving lesioning or low-frequency stimulation of thalamic nuclei, were developed to treat movement disorders and psychiatric conditions, with early reports indicating symptomatic relief without permanent ablation.[225] In the late 1940s, stereotactic techniques enabled deeper brain interventions for movement disorders and psychiatric conditions. Chronic subcortical electrical stimulation emerged in the early 1950s, pioneered by researchers such as Robert Heath and José Delgado, who implanted electrodes in psychiatric patients to achieve reversible therapeutic effects as an alternative to lesioning procedures.[226] The decade closed with neurosurgeon C. Norman Shealy's 1967 implantation of the first percutaneousspinal cordstimulation (SCS) device, inspired by the gate control theory of pain, which delivered dorsal column stimulation to alleviate chronicintractable pain in a patient with terminal cancer, marking the advent of implantable neuromodulation for sensory modulation.[227]
Evolution of Clinical Applications
The clinical application of neurostimulation began to transition from experimental procedures to regulated therapies in the 1970s and 1980s, with spinal cord stimulation (SCS) emerging as a key modality for chronic pain management. The first human implantation of an SCS device occurred in 1967, marking the start of its therapeutic use, though widespread adoption followed the development of fully implantable systems in the early 1980s. By 1989, the U.S. Food and Drug Administration (FDA) approved SCS specifically for relieving chronic intractable pain of the trunk and limbs due to nerve damage, building on earlier investigational uses. Concurrently, vagus nerve stimulation (VNS) entered clinical trials for epilepsy in the late 1980s, with initial implants demonstrating feasibility as an adjunctive therapy for refractory seizures; it received FDA approval in 1997.The 1990s saw further maturation, particularly with deep brain stimulation (DBS) for movement disorders and the expansion of non-invasive techniques. In 1997, the FDA granted humanitarian device exemption approval for DBS targeting the ventral intermediate nucleus of the thalamus to treat essential tremor and parkinsonian tremor, paving the way for its application in Parkinson's disease, which received full approval in 2002 for advanced cases. Transcranial magnetic stimulation (TMS), introduced in the mid-1980s, gained traction in the 1990s for research into depression, with studies showing its potential to modulate cortical excitability non-invasively, though FDA clearance for treatment-resistant depression came later in 2008. VNS also gained FDA approval in 2005 for treatment-resistant depression.Entering the 2000s, neurostimulation diversified with the commercialization of transcranial direct current stimulation (tDCS) and the mainstream integration of cochlear implants. tDCS devices became commercially available around 2000, following demonstrations of its ability to induce lasting changes in brain excitability for cognitive and motor enhancement, with early applications in rehabilitation. Cochlear implants, initially approved by the FDA in 1984, achieved mainstream status in the 2000s as implantation rates surged, restoring hearing in thousands of profoundly deaf individuals annually by modulating auditory nerve activity.The 2010s brought advancements in stimulation parameters and novel targets, enhancing efficacy and accessibility. High-frequency SCS at 10 kHz received European CE mark approval in 2010 and FDA clearance in 2015 for chronic back and leg pain, offering paresthesia-free relief superior to traditional low-frequency methods in randomized trials. Transcutaneous auricular VNS (taVNS) emerged for anti-inflammatory effects, with clinical studies from the mid-2010s showing reduced cytokine levels in conditions like rheumatoid arthritis by non-invasively activating vagal pathways. Visual prostheses, such as the Argus II retinal implant, gained FDA approval in 2013 for retinitis pigmentosa, enabling phosphene-based vision restoration in blind patients through epiretinal stimulation.Regulatory frameworks evolved to support global dissemination, often with European CE marks preceding FDA approvals by several years, facilitating earlier access in the EU. By the 2020s, neurostimulation adoption expanded significantly in Asia, with countries like China and Japan approving advanced DBS and SCS systems, contributing to over 20% of global implants by 2023. By 2025, the field had marked over four decades since the introduction of the first fully implantable SCS system in 1981, underscoring its enduring impact. Recent integrations with artificial intelligence, such as adaptive DBS systems approved by the FDA in 2025, enable real-time personalization of stimulation parameters based on neural biomarkers, improving outcomes in Parkinson's disease.
Research Directions
Ongoing Clinical Trials
As of November 2025, numerous clinical trials continue to evaluate the efficacy and safety of neurostimulation techniques, focusing on pain management, neuropsychiatric disorders, sensory restoration, and neuromodulation for neurological conditions. The COVID-19 pandemic delayed many invasive trials in prior years, leading to increased emphasis on non-invasive methods. Primary endpoints often involve scales like the Visual Analog Scale (VAS) for pain and seizure frequency for epilepsy, with sample sizes typically 100–500 for statistical power.[228]In pain management, ongoing multicenter trials assess dorsal root ganglion (DRG) stimulation for refractory chronic lower limb neuropathic pain, comparing it to spinal cord stimulation (SCS) for long-term relief and function.[229] Recent completed studies, such as the 2019–2023 comparison of burst SCS to tonic SCS, have shown superior pain reduction with burst in chronic back and leg pain via VAS and preference metrics.[230]For neuropsychiatric applications, completed trials like ADvance II (NCT03622905) evaluated deep brain stimulation (DBS) of the fornix in mild Alzheimer's disease, reporting slowed cognitive decline in older patients over 12 months via double-blind assessment.[231] Planned trials, such as NCT06953388 starting in 2026, will test transcutaneous auricular vagus nerve stimulation (taVNS) for PTSD, building on preclinical data showing reduced anxiety and improved autonomic responses.[232]Sensory neurostimulation includes ongoing evaluations like NCT05626426 (Phase I), investigating optic nerve cuff electrodes for safety and visual enhancement in optic neuropathies using low-intensity stimulation. Planned studies, such as NCT07213505 (not yet recruiting as of October 2025), aim to assess AI-driven cochlear implants for improved speech recognition in noise.[233]In neuromodulation, trials like NCT06722339 evaluate transcranial focused ultrasound (tFUS) for obsessive-compulsive disorder (OCD), targeting circuits such as the ventral capsule/ventral striatum for symptom reduction.[234] For epilepsy, responsive neurostimulation trials using closed-loop DBS adapt to seizure onset, achieving 50–70% frequency reductions in drug-resistant cases, supported by systems like the RNS.[235]
Emerging Techniques and Innovations
Optogenetics uses light-sensitive ion channels like channelrhodopsins for precise, cell-type-specific neuron activation, offering millisecond control beyond traditional electrical methods. Adeno-associated viruses (AAV) deliver opsins for long-term expression in preclinical models. As of 2025, fully implantable wireless optogenetic platforms enable multisite stimulation, with human trials advancing in retinal diseases (e.g., phase 3 planning for vision restoration), while central nervous system applications remain preclinical or pending ethics approvals.[236]Nanotransducers, such as magnetic nanoparticles (e.g., iron oxide), enable non-invasive neural control by converting ultrasound or magnetic fields into mechanical/thermal stimuli for localized firing. Preclinical studies demonstrate deep brainneuromodulation for Parkinson's motor symptoms via neurondepolarization, with research focusing on biodegradable materials to reduce risks.[237]Acoustic photonic intellectual neurostimulation (APIN) combines sound and light to mimic natural inputs, promoting neuroplasticity in neurodegeneration and pain. Reported in 2024 case studies for dysmenorrhea, APIN achieved up to 80% symptom alleviation in adolescents via non-invasive sessions; early 2025 cohorts show cognitive improvements in small neurodegenerative groups.[238]Brain-computer interfaces (BCIs) like Neuralink feature high-channel implants with thousands of electrodes for recording and stimulation, restoring motor function in paralysis. As of September 2025, 12 patients with severe paralysis have received implants, enabling thought-based cursor control and digital interaction with sustained functionality over 2,000 device-days.[239]Hybrid gene therapy-neurostimulation enhances plasticity by pairing genetic modifications (e.g., neurotrophic factors via viral delivery) with electrical/optical inputs. Preclinical spinal cord injury models show 30–50% better behavioral outcomes than stimulation alone; stroke studies indicate neuroprotective motor improvements.[240]Closed-loop systems integrate machine learning (ML) to predict neural states from biomarkers like EEG, adjusting parameters in real-time for conditions like epilepsy or Parkinson's. Early human data show ML-enhanced DBS achieving up to 70% better symptom control versus open-loop, using reinforcement learning for personalization.[241]