Neuromodulation Devices: What They Are and How They Work

Neuromodulation devices are tools that deliver targeted energy to the nervous system to change how nerve cells fire, aiming to relieve symptoms of conditions ranging from chronic pain to Parkinson’s disease to epilepsy. They span a wide spectrum, from small non-invasive gadgets held against the scalp to surgically implanted electrodes threaded deep into the brain. The techniques fall into two broad camps: non-invasive methods like transcranial magnetic stimulation (TMS), transcranial direct current stimulation (tDCS), and transcranial ultrasound stimulation (TUS), and surgical methods like deep brain stimulation (DBS), spinal cord stimulation (SCS), and vagus nerve stimulation (VNS).¹ What unites all of them is a single principle: altering the electrical and chemical activity of neurons without destroying tissue.

How Non-Invasive Brain Stimulation Works

The non-invasive devices reach the brain through the skull, each using a different form of energy. TMS uses rapidly changing magnetic fields to induce small electrical currents in brain tissue. A coil placed against the scalp generates brief, high-intensity magnetic pulses that pass through bone and activate neurons in the cortex below. When these pulses are delivered repeatedly, a technique called repetitive TMS (rTMS), the effects outlast the stimulation session and can shift the brain’s baseline excitability up or down depending on the frequency and pattern of pulses.² That lasting change is what makes rTMS useful as a treatment rather than just a lab tool.

Transcranial direct current stimulation (tDCS) works differently. Instead of inducing pulses, it sends a weak, steady electrical current between two sponge electrodes placed on the scalp. The current is too faint to fire neurons outright, but it nudges their resting electrical state in one direction or another. Under the electrode where current flows in, neurons become slightly more likely to fire; under the electrode where it flows out, they become slightly less likely.³ Over time, tDCS also appears to strengthen the connections between neurons, improve local blood flow, and enhance communication between distant brain regions.⁴ The device itself is simple and portable, which has made it popular in research labs and, increasingly, in rehabilitation settings.

Transcranial ultrasound stimulation (TUS) is newer and works through mechanical pressure rather than electricity or magnetism. Focused sound waves pass through the skull and create tiny pressure fluctuations in brain tissue. These pressure changes activate a class of proteins on neuron membranes called mechanosensitive ion channels, which open in response to physical force and alter the neuron’s electrical state.⁵ One key player is a channel called Piezo1. Mouse experiments have shown that knocking out Piezo1 in the motor cortex significantly reduces the brain’s response to ultrasound, confirming that the channel is a major mediator of how ultrasound affects neurons.⁶ The advantage of ultrasound is spatial precision: it can be focused on small, deep brain structures without surgery, something the other non-invasive methods struggle with. Recent work has demonstrated that an ultrasound chip can selectively stimulate a tiny subregion of the mouse hippocampus and induce lasting changes in synaptic strength at that spot.⁷

Deep Brain Stimulation and Abnormal Oscillations

Deep brain stimulation is the most established surgical neuromodulation technique. A neurosurgeon threads thin electrodes into specific brain structures, typically the subthalamic nucleus (STN) or the globus pallidus for Parkinson’s disease, and connects them to a pulse generator implanted near the collarbone. The generator delivers continuous electrical pulses, and the patient’s symptoms can be adjusted after surgery by reprogramming the device. Early pacemakers and DBS devices paved the way for the sophisticated implantable technologies now in use.⁸

In Parkinson’s disease, the core problem DBS addresses is abnormal synchronized oscillations in the brain’s motor circuits. Neurons in the basal ganglia start firing in lockstep at a particular rhythm in the beta frequency band, and this pathological synchrony is closely linked to symptoms like rigidity and slowness of movement. DBS disrupts that rhythmic firing. Computational modeling has shown that excitatory stimulation can quench these beta oscillations and restore more normal signaling between the STN and the motor cortex.⁹ Getting the dose right matters enormously: too little stimulation leaves the abnormal oscillations untouched, while too much can suppress not only the pathological rhythm but also the brain’s ability to relay normal motor commands, effectively shutting down the circuit in a different way. Modeling work has identified a narrow therapeutic window where beta suppression and normal relay function coexist.¹⁰

DBS has also been applied to epilepsy through a related strategy called responsive neurostimulation (RNS). Rather than delivering continuous pulses, the RNS device monitors the brain’s electrical activity in real time and fires stimulation only when it detects the beginning of a seizure. This closed-loop approach has been successful overall, though individual outcomes vary widely. A multicenter study found that a specific pattern in high-frequency brain activity recorded before implantation could predict which patients would respond well to RNS, potentially helping clinicians select the right candidates.¹¹

Spinal Cord Stimulation and the Evolution of Waveforms

Spinal cord stimulation grew out of gate control theory, the idea proposed in 1965 that activating certain nerve fibers in the spinal cord can close a “gate” to pain signals traveling to the brain.¹² Traditional SCS works by delivering low-frequency electrical pulses to the dorsal columns of the spinal cord through implanted leads. Patients typically feel a tingling sensation, called paresthesia, in the area where they would otherwise feel pain. For decades, this was the standard approach for chronic pain that did not respond to medication.

Newer waveform designs have changed the landscape. High-frequency stimulation at 10,000 Hz and burst stimulation, which delivers packets of pulses in rapid clusters, both provide pain relief without any paresthesia. The SENZA trial showed that paresthesia-free high-frequency SCS was superior to traditional low-frequency stimulation for chronic low back pain with leg pain, and the SUNBURST crossover trial found that patients preferred burst stimulation over conventional SCS for similar reasons.¹³ Whether these two paresthesia-free approaches work through identical mechanisms has been a matter of debate. Both were hypothesized to modulate the brain’s medial pain pathway, which is associated with the emotional and suffering component of pain rather than the raw sensory signal.

A recent study using brain imaging during SCS provided some clarity. Both burst and 10 kHz stimulation activated the same region in the pregenual anterior cingulate cortex, a hub for the brain’s descending pain-inhibition system, and the level of activation there correlated with how much pain was suppressed. But burst stimulation also engaged additional areas, including the dorsal anterior cingulate cortex and the insula, and activity in those regions correlated specifically with reductions in back pain.¹⁴ This may explain why burst stimulation showed a larger reduction in back pain from baseline, though the direct head-to-head comparison between burst and 10 kHz did not reach statistical significance. The mechanisms are overlapping but not identical: burst appears to recruit an extra layer of pain-processing circuitry. At the cellular level, burst SCS does not depend on the same inhibitory signaling that traditional SCS uses, while high-frequency SCS appears to block the ability of large nerve fibers to generate signals, increasing pain suppression as the frequency goes up.¹⁵

Peripheral Nerve Stimulation

While SCS targets the spinal cord centrally, peripheral nerve stimulation (PNS) places electrodes directly near the specific nerve causing pain. PNS has emerged as a minimally invasive option for neuropathic pain, often implanted percutaneously using imaging guidance.¹⁶ Its availability has grown as newer, smaller devices and better implantation techniques have expanded clinician expertise.¹⁷

Some of the strongest evidence for PNS comes from postamputation pain, a notoriously difficult condition. A randomized, placebo-controlled trial found that about 58% of patients receiving active PNS achieved at least a 50% reduction in average pain during the first four weeks, compared with only 14% of those receiving a sham treatment. After eight weeks of therapy, roughly two-thirds of PNS patients maintained that level of pain relief, and 80% reported meaningful reductions in how much pain interfered with daily life.¹⁸ Some of these devices are designed to be temporary, with leads placed for a defined treatment period and then removed, while others are permanently implanted depending on the clinical situation.

The Vagus Nerve and the Inflammatory Reflex

Vagus nerve stimulation was originally approved for epilepsy and later for treatment-resistant depression, but its reach has expanded into territory few would expect from a neuromodulation device: the immune system. The vagus nerve plays a central role in regulating immune function through what is called the inflammatory reflex. Signals traveling down the nerve trigger the release of acetylcholine, which in turn dials down the production of pro-inflammatory molecules.¹⁹

This is not just a laboratory observation. In a study of patients with rheumatoid arthritis, an implanted vagus nerve stimulator delivered up to four sessions daily and significantly reduced production of the inflammatory protein TNF for up to 84 days.²⁰ Animal experiments have helped clarify the mechanism: even very brief vagus nerve stimulation, as short as a fraction of a second, can significantly reduce systemic TNF release, and the effect lasts more than 24 hours. The suppression depends on a specific receptor subunit called α7nAChR, because mice lacking that receptor show no anti-inflammatory benefit from stimulation.²¹ The idea that an electrical device can modulate immune inflammation opens a path toward treating autoimmune conditions with circuitry instead of, or alongside, drugs.

Closed-Loop and Wearable Systems

Most neuromodulation devices in clinical use today are “open-loop,” meaning they deliver stimulation according to a preset program regardless of what the brain or nerve is doing at any given moment. Closed-loop systems, by contrast, continuously monitor neural activity and adjust stimulation in real time based on what they detect. The epilepsy RNS device discussed earlier is one example already in clinical use, but the engineering challenge is far from solved.

Recording neural signals while simultaneously delivering electrical pulses is technically difficult because the stimulation creates large electrical artifacts that can drown out the very brain signals the device is trying to read. Specialized amplifier designs have been developed to tolerate these artifacts and maintain clean recordings during active stimulation, a critical requirement for any closed-loop DBS system to function reliably.²² Beyond the electronics, researchers have also developed a wearable closed-loop platform capable of recording activity from individual neurons and broader electrical fields in freely moving people, rather than requiring patients to sit still in a hospital. This system opens the door to studying how the brain behaves during natural, everyday activities and to designing stimulation that adapts to those real-world conditions.²³

A related hardware problem is what happens at the boundary between an implanted electrode and brain tissue over time. The body treats the electrode as a foreign object and forms scar tissue around it, which degrades signal quality and raises stimulation thresholds. Research has shown that the density mismatch between the electrode material and surrounding brain tissue is a key driver of this scarring. Probes made from materials that more closely match the density of brain tissue produce significantly less scar formation.²⁴ Softer, lighter electrode designs are an active area of development for improving the longevity of implanted devices.

Side Effects and the Question of Identity

Non-invasive devices carry relatively modest risks: TMS can cause headaches and, rarely, seizures; tDCS can cause skin irritation under the electrodes. The safety profile of implanted devices is more complex. DBS electrodes are placed with high precision, but even well-placed leads can spread current to adjacent brain structures and produce unintended effects.²⁵ Strategies like using narrower stimulation fields and bipolar electrode configurations can help minimize these spillover effects.²⁶

But some effects of DBS are harder to classify as simple side effects. Patients undergoing DBS for conditions like Tourette’s syndrome or treatment-resistant depression have reported shifts in personality, mood, and even their core values. In one documented case, a patient with severe Tourette’s reported feeling like a more “normal” version of herself after DBS, but her family raised concerns about observed changes in her political and religious views. Reviews of open-label depression studies have cited similar concerns, with patients questioning their own sense of self-identity after treatment.²⁷ These reports raise a question that is genuinely difficult to answer: if a device changes your personality, is the result still “you”? Ethicists have been debating this for over a decade, though the discussion is complicated by the lack of clear, agreed-upon definitions of what “personal identity” even means in this context.²⁸ It is worth noting that the diseases themselves change personality and behavior too. The question is not whether identity shifts occur, but how to weigh them against the often-devastating symptoms the device is treating.

Cost and the Path to Access

Implantable neuromodulation devices are expensive upfront. A spinal cord stimulator, including the device, surgical placement, and programming, can cost tens of thousands of dollars. The economic argument in their favor rests on whether they reduce longer-term spending on medication, repeated procedures, and emergency care. A cost-effectiveness analysis using data from the DISTINCT randomized trial and real-world insurance claims found that SCS for chronic low back pain could reach cost-effectiveness in roughly two years when compared with conventional medical management, depending on which cost estimates were used.²⁹ That break-even period assumes the device continues working and the patient’s healthcare use drops, both of which can vary. Insurance coverage for neuromodulation therapies remains inconsistent, and access is heavily influenced by geography, physician expertise, and whether a patient’s specific diagnosis falls within approved indications.

Sensory Stimulation and Emerging Frontiers

Not all neuromodulation requires electrodes or even focused energy beams. One of the more surprising recent developments is the use of 40 Hz sensory stimulation, flickering light and pulsing sound at 40 cycles per second, to entrain brain rhythms associated with cognition. Preclinical work in mouse models of Alzheimer’s disease showed that this approach could reduce the brain’s burden of amyloid-beta plaques by roughly 37% to 53%, inhibit the abnormal modification of tau protein, and improve learning and memory. Early human trials in patients with mild cognitive impairment and early Alzheimer’s have found the stimulation to be safe, well tolerated, and associated with a slowing trend in cognitive decline, though the human evidence is still limited.³⁰ The idea that a non-invasive sensory experience could clear toxic proteins from the brain would have seemed implausible a decade ago, and it remains to be seen whether larger, longer trials will confirm the early promise.

Further out on the horizon is optogenetics, a technique that uses genetically engineered light-sensitive proteins inserted into specific populations of neurons. Shining light of a particular wavelength onto these neurons can excite or silence them with a precision that electrical stimulation cannot match, because it can target one cell type while leaving its neighbors alone.³¹ Optogenetics has transformed animal neuroscience research over the past two decades, but clinical translation in humans remains in its earliest stages because it requires gene therapy to deliver the light-sensitive proteins and implanted light sources to activate them. The gap between the lab and the clinic is still wide, yet the technology illustrates where the field is heading: toward devices that can intervene in specific cell populations within specific circuits, with a degree of selectivity that today’s electrical devices simply cannot achieve.