Phase 6: Beyond the scalpExplainer16 min

Neuromodulation, mechanisms and evidence

Deep brain stimulation, spinal cord stimulation, vagus nerve stimulation, cochlear and retinal implants, focused ultrasound, TMS, and tDCS. What each does to tissue, what the evidence says, and the honest state of "we do not fully know why it works."

AssumesWhere oscillations come from Project D: Stimulation, studied properlySpineClinical / regulatory / human factorsComputational neuroscience

You are skimming: the title, the first figure, and the short version. Switch to Read in the header for the full page, or Deep to open every deep dive.

Neuromodulation is the largest commercial part of this field and the least discussed by students, who find recording more interesting than stimulating. It is also the part with the most patients helped: over a hundred thousand people with deep brain stimulators, a million with cochlear implants, many more with spinal cord stimulators. For most of these, the mechanism is only partly understood, which is a strange and important fact about medicine. Here is each modality: what it stimulates, what it treats, what the evidence looks like, and what we think it does.

Deep brain stimulation

Electrodes in a deep nucleus (the subthalamic nucleus or globus pallidus for Parkinson’s, the thalamus for tremor), pulsed continuously at around 130 Hz, 60 to 90 µs, 2 to 4 mA, from a pacemaker-like generator in the chest. Approved for Parkinson’s disease, essential tremor, dystonia, epilepsy, and obsessive-compulsive disorder; trials in depression and others. The effect on tremor is dramatic and immediate: turn it on and the shaking stops.

Mechanism: it was thought to silence the target, mimicking a lesion. It is now understood to do several things at once: drive axons near the electrode at the stimulation frequency, overwriting the pathological rhythm (the excessive beta synchrony of the parkinsonian basal ganglia) with a regular one the downstream circuits can ignore; alter the firing of the nucleus’s neurons; and, over time, induce plasticity. Adaptive DBS, which senses beta bursts and stimulates only then, is in clinical use and supports the rhythm-disruption account. The frontier is closed-loop control and steering current between segmented electrode contacts to shape the field.

Spinal cord stimulation

Electrodes in the epidural space over the dorsal columns, from a chest or abdominal generator. Long used for chronic pain (gate-control theory: stimulating large sensory fibres suppresses pain transmission; the evidence for benefit is real but modest and placebo effects are substantial, as sham-controlled trials show). Since about 2018, targeted SCS with patterns timed to intended movement has restored stepping in people with spinal cord injury previously thought complete, by amplifying residual signals across the lesion and engaging spinal circuits. This is the most exciting current application and it is early.

Vagus nerve stimulation

A cuff on the left vagus nerve in the neck. Approved for epilepsy and depression, and, paired with rehabilitation, for stroke recovery of arm function. Mechanism for epilepsy: modulation of brainstem and thalamic activity that raises seizure threshold, incompletely understood. For stroke: timed release of noradrenaline and acetylcholine that enhances plasticity during the movements being practised, which the plasticity explainer describes and which is well supported by animal work. Non-invasive ear-clip and neck versions exist with weaker evidence.

Cochlear and retinal implants

The sensory systems explainer covers them. The cochlear implant is the field’s clearest success: mechanism understood, evidence overwhelming, a million users. The retinal implant is its cautionary counterpart: mechanism understood, evidence for modest benefit, the leading company gone.

Responsive neurostimulation for epilepsy

Electrodes on or in the seizure focus, a generator in the skull that records continuously, detects the electrographic start of a seizure, and stimulates to abort it. Approved since 2013; seizure reduction improves over years of use, suggesting plasticity, and the device’s recordings have taught the field about seizure cycles over days and weeks that nobody knew existed. It is the first closed-loop brain implant in routine clinical use.

Transcranial magnetic stimulation

A coil on the scalp discharges a brief, strong magnetic pulse that induces current in the cortex beneath, depolarizing neurons. Single pulses probe motor cortex excitability (the motor evoked potential). Repetitive protocols over the left prefrontal cortex are approved for depression, with response rates around a third to a half in medication-resistant patients and an accelerated protocol (many sessions a day for a week) recently approved. Mechanism: induced plasticity in prefrontal-limbic circuits, incompletely understood. Requires a trained operator; seizure risk is low but real.

Transcranial direct current stimulation

A milliamp or two through sponge electrodes on the scalp, for twenty minutes. The current that reaches the cortex is a fraction of a volt per metre, enough to shift membrane potentials slightly and bias firing, not to drive it. The literature is large and inconsistent: modest effects on motor learning and some cognitive tasks in some studies, failures to replicate in others, large variability between people, and a strong placebo component. Not approved for any indication in the United States. It is the modality whose circuit is simplest and whose evidence is weakest, which is exactly why the safety charter treats it the way it does.

Focused ultrasound

Sound energy concentrated deep in the brain through the skull. At high intensity it ablates tissue and is approved for essential tremor (a non-invasive alternative to DBS, with a lesion instead of a stimulator). At low intensity it modulates neural activity reversibly, by mechanisms (mechanical effects on channels, possibly thermal) under active study. The only technique that can reach deep structures non-invasively with millimetre focus, which makes it the most watched experimental modality.

Predict before you look

Deep brain stimulation stops a Parkinson's tremor within seconds of being switched on. Does that mean its mechanism is understood?

No. DBS was adopted because it worked, visibly, in the operating room, and the mechanism has been argued about for thirty years since. The current synthesis is that it disrupts pathological synchrony by driving axons at a regular high frequency, changes local firing, and induces plasticity over time. Medicine is full of treatments whose effectiveness preceded their explanation; a student who can say “it works, and here is what we think and do not know about why” is more credible than one who claims certainty either way.

Reading the evidence

For each modality, ask: is there a sham-controlled trial, and what did it show? What is the effect size, and in whom? What is the mechanism claimed, and what is the evidence for it specifically (not just that the treatment works)? Cochlear implants and DBS for tremor pass every question. Spinal cord stimulation for pain passes with a modest effect and large placebo. tDCS fails the first question in most applications. Learning to run this checklist is the clinical spine’s core skill.

Stop here if you like. Eight modalities, what each does, what the evidence says, and where the mechanism is honestly unknown. That is neuromodulation.

Deep dive Why high-frequency stimulation can silence and drive at once 3 min

A neuron’s cell body and its axon respond differently to the same pulses. At 130 Hz, cell bodies near the electrode often stop firing in their own pattern (depolarization block, or inhibitory afferent activation), while axons passing by are driven to fire at the stimulation frequency. Downstream, the target receives a regular 130 Hz input instead of the pathological bursty beta pattern, and regular input is, in effect, informationally silent. So DBS both silences the nucleus’s own output pattern and replaces it with a regular one. Computational models (including work in the Dorval lab on this campus) showed this “informational lesion” account, and it reconciles the older silencing and driving hypotheses.

Recall

What does the current synthesis say deep brain stimulation does?

It drives axons near the electrode at a regular high frequency, overwriting pathological synchrony (an informational lesion), alters local firing, and induces plasticity over time; the effect was established clinically long before the mechanism was understood.

Recall

How does vagus nerve stimulation help stroke recovery?

Timed with rehabilitation movements, it triggers release of noradrenaline and acetylcholine that enhance plasticity in the circuits being trained.

Recall

What questions should you ask of any neuromodulation claim?

Is there a sham-controlled trial and what did it show; what is the effect size and in whom; what mechanism is claimed and what is the evidence for that mechanism specifically.

Explain it to your roommate

Explain what this page was about to your roommate in three sentences. No jargon they would not know.