Imagine a hand that will not stop shaking. It spills the coffee, smears the signature, and rattles the fork against the plate. Now imagine a doctor holding a small tablet near your chest, tapping a setting, and watching the shaking fade within seconds. That is what deep brain stimulation can do for many people with Parkinson’s disease and other tremors. A thin wire sits deep in the brain, a battery sits under the collarbone, and a steady stream of tiny electrical pulses calms circuits that have started to misfire. More than 100,000 people had been treated this way by 2014, according to the Lasker Foundation, and the number has kept growing. Here is how a brain pacemaker works, how it was found almost by accident, and what it can and cannot fix.
This article is for general information only and is not medical advice. If you or someone you care for is living with Parkinson’s disease or a tremor, talk with a neurologist about whether any treatment, including surgery, is right for you.
Three parts, all under the skin
A deep brain stimulation system has three pieces. The first is the lead, a thin insulated wire with a few small metal contacts near its tip. A surgeon slides it through a small hole in the skull, roughly the width of a fingertip, until the tip rests in a chosen spot deep in the brain. Many people get one lead on each side.
The second piece is the pulse generator, a small sealed case holding a battery and electronics. It sits under the skin of the upper chest, near the collarbone, much like a heart pacemaker. The third piece is an extension wire that runs under the skin of the scalp, behind the ear, and down the neck to connect the two. Nothing pokes out of the body. Doctors adjust the settings wirelessly, and patients usually get a handheld remote that can check the device or turn it on and off.
Brain wiring crosses over, so the right side of the brain controls the left side of the body, and the other way around. A lead on the right side mostly helps the left hand and leg.
A discovery made by accident
Surgeons were treating tremor long before anyone implanted a stimulator. In the 1950s, they learned that destroying a tiny patch of tissue in the right place could quiet a shaking limb. These operations were risky, though, and the damage was permanent. When the drug levodopa arrived in the late 1960s and gave many people with Parkinson’s their movement back, most surgeons put their tools away.
The modern story starts in Grenoble, France, in 1987. Neurosurgeon Alim-Louis Benabid was about to burn a small lesion in the thalamus, a relay station deep in the brain, to stop a patient’s tremor. As was routine, the patient was awake, and Benabid sent small electrical pulses through a probe first to make sure he had the right spot. At 30 to 50 pulses per second, the current only caused tingling or muscle twitches. But when he pushed it to about 100 pulses per second, the tremor suddenly stopped. When he turned the current off, the tremor came back.
That gave Benabid a way to get the effect of a lesion without destroying anything. His team began leaving stimulating electrodes in place, connected to small implanted pulse generators. Meanwhile, in the United States, neurologist Mahlon DeLong was studying monkeys with a Parkinson’s-like condition. He found that cells in a small structure called the subthalamic nucleus were firing far too much, and in 1990 his team reported that knocking out that structure eased the monkeys’ symptoms. In 1993, Benabid’s group tried high-frequency stimulation of the subthalamic nucleus in people with advanced Parkinson’s. It worked. In the U.S., the Food and Drug Administration approved deep brain stimulation for tremor in 1997 and for Parkinson’s disease in 2002. In 2014, Benabid and DeLong shared the Lasker-DeBakey Clinical Medical Research Award, one of medicine’s top prizes, for the work.
Why a steady buzz quiets a shaking hand
Parkinson’s disease slowly destroys brain cells that make dopamine, a chemical messenger the brain uses to start and smooth out movement. Without enough dopamine, a set of deep brain structures called the basal ganglia falls out of balance. Groups of cells there start firing too much and too much in lockstep, like a crowd that has fallen into a chant and drowns out normal conversation. That stuck rhythm seems to be part of what makes movement slow, stiff, and shaky.
Nobody fully understands why fast electrical pulses help, and scientists still debate the details. The leading idea is that steady, high-frequency stimulation breaks up the stuck rhythm, so the faulty signal can no longer hijack the circuit. In effect, it works much like the old lesion surgeries, except it can be adjusted, and it can be switched off.
The surgery, often done awake
Before surgery, doctors map the brain with MRI or CT scans. On the day of the operation, many patients wear a stereotactic head frame, a rigid metal frame that holds the head still and gives the surgeon a precise set of coordinates. Others are treated with frameless systems that do the same job.
According to Mayo Clinic, the electrodes are most often placed while the patient is awake. The scalp is numbed, but the brain itself has no pain sensors, so the probe does not hurt. Being awake lets the team test the effect in real time. They can ask the patient to move a hand, feel for stiffness in the wrist, or listen to the patient talk. Many teams also listen to the crackle of individual brain cells through a very fine recording electrode, since each deep structure has its own firing pattern. Some centers now place the leads under general anesthesia instead, using MRI during surgery to guide the wire. The pulse generator goes into the chest in a separate step, with the patient asleep.
Switching the system on is not the end. A few weeks after surgery, a clinician begins programming the device, choosing which contacts to use, how strong the pulses are, and how fast they come. Mayo Clinic says finding the best settings can take as long as four to six months of visits.
Where the wires go
Doctors pick the target based on the symptoms they most want to fix. These are the three most common:
| Target | Where it is | Mainly used for |
|---|---|---|
| Subthalamic nucleus (STN) | A small structure just below the thalamus | Parkinson’s slowness, stiffness, and tremor. It often lets people take less medication. |
| Globus pallidus internus (GPi) | Part of the basal ganglia, the brain’s movement control loop | Parkinson’s, especially the wriggling extra movements caused by medication, and dystonia |
| Ventral intermediate nucleus of the thalamus (VIM) | A relay station in the thalamus | Tremor, including essential tremor and tremor-heavy Parkinson’s |
What it can fix, and what it can’t
Deep brain stimulation is not a cure, and it does not seem to slow the disease. What it can do is smooth out the ups and downs of Parkinson’s. People on levodopa for many years often have “off” periods when a dose wears off and they freeze or stiffen. Many also get dyskinesias, the wriggling, dance-like movements that come with too much medication. Stimulation can shorten the off periods, calm those extra movements, and often reduce the medication needed.
A useful rule of thumb is that the surgery tends to help most with the symptoms that levodopa already helps. That is why doctors usually test a patient both on and off medication before agreeing to operate. Tremor is the main exception, since stimulation can sometimes control it better than pills.
Other problems usually respond poorly. Balance, freezing while walking, swallowing, speech, and memory often do not improve much, and speech can get worse, especially with subthalamic stimulation. Mood changes are possible too, including apathy or depression, so careful screening before and follow-up after surgery matter. People with dementia, severe depression, or frequent falls even at their best are often not good candidates.
Then there are the surgical risks. Mayo Clinic lists bleeding in the brain, stroke, infection, misplaced leads, and seizures among the possible complications. In large reviews, bleeding that causes symptoms happens in roughly 1 to 2 percent of patients, and hardware infections in a few percent. The risks are small, but they are real. Stimulation settings can also cause side effects such as tingling, tight facial muscles, slurred speech, or double vision. Those often ease when the device is reprogrammed. Batteries do not last forever, so people with non-rechargeable devices eventually need a minor outpatient surgery to swap the generator.
Beyond Parkinson’s
Essential tremor, a common condition that causes shaking hands, was among the first uses approved in the United States. The FDA has since allowed deep brain stimulation for dystonia, a condition that twists muscles into painful postures, for some cases of severe obsessive-compulsive disorder, and for hard-to-treat epilepsy. Researchers are also testing it for depression, chronic pain, Tourette syndrome, and more. Results for depression have been mixed so far.
Deep brain stimulation is one of several kinds of brain implants now in use. It sends signals into the brain. Others, like the chips Neuralink is testing, listen to brain activity and turn it into commands for a computer, as we covered in our short history of Neuralink. A cochlear implant works on the hearing nerve instead, as explained in our piece on how cochlear implants turn sound into hearing.
A pacemaker that listens
For most of its history, deep brain stimulation has worked like a sprinkler on a timer. The device delivers the same pulses all day, whether a person’s brain needs more or less at that moment. That is starting to change.
The leads that deliver stimulation can also pick up the brain’s own electrical activity. In Parkinson’s, a rhythm in the so-called beta range tends to rise when symptoms are worse and fall when medication kicks in. In February 2025, the FDA approved a feature from the medical device company Medtronic called BrainSense Adaptive deep brain stimulation. It tracks that signal on compatible Medtronic devices and automatically raises or lowers the stimulation in real time. The approval rested on a trial called ADAPT-PD, run at 10 centers in the United States, Canada, the Netherlands, and France, with people who were already using the standard version. Medtronic said it was the first adaptive system of its kind to win U.S. approval.
Adaptive stimulation will not work for everyone, and researchers are still learning which patients benefit most. But it marks a shift in the basic idea, from a device that simply pushes current into the brain to one that responds to what the brain is doing.
The bottom line
Deep brain stimulation grew out of a surprise in a French operating room, where a burst of fast pulses stopped a tremor cold. Nearly four decades later, it is a standard option for people with Parkinson’s whose medication no longer covers the whole day, and for many with essential tremor and dystonia. It will not cure the disease or fix every symptom, and it means brain surgery, programming visits, and a battery in the chest. But for the right person, a thin wire and a steady electrical hum can give back a steady hand.
Two books for going deeper
Ending Parkinson’s Disease: A Prescription for Action — Four Parkinson’s experts, including neurologist Michael S. Okun, a longtime deep brain stimulation specialist, explain what drives the disease, how it is treated today, and what it would take to prevent it. A clear, practical read for patients and families.
We Are Electric — Science journalist Sally Adee tells the 200-year story of the electricity that runs through our bodies, from twitching frog legs to today’s brain implants. A lively companion if this piece made you curious about why a few pulses can change how a brain behaves.