Skip to main content

In 2006, Liss Murphy was in thrall to what she calls a “sepsis of the soul” — an intractable and debilitating depression. She had hardly spoken in two years. She felt almost nothing; she was barely eating; she wanted to die.

No conventional treatments had helped. So when she heard that doctors at Massachusetts General Hospital, in Boston, had developed an experimental cure for severe depression that involved permanently implanting electrodes in the brain, she didn’t hesitate. The procedure seemed like no big deal. “I never read the consent form,” she says. “I just didn’t care.” This was her last shot, she thought. She half-hoped they would make a fatal mistake during the operation.

A few months later, on June 6th, 2006, Murphy lay in an operating room in the neurosurgery wing of Mass General. She looked hardly alive — her body emaciated from eating almost nothing, her skull shaved in preparation for the surgery. A shiny, donut-shaped CT scanner surrounded her head. The doctors began their work by drillingtwo dime-sized openings into her skull. Then they gingerly lowered tiny electrodes, about the width of the graphite in a pencil, into a region of her cerebral cortex known as the internal capsule. Once the electrodes were in place, the doctors asked her to interact with a computer simulation, with the holes in her skull still open. Before the surgery, they had used the CT scanner and a computerized navigation system (a kind of GPS for the brain surgeon) to map her brain and determine the precise spots where they would implant the electrodes.

Together with an electrical pulse generator — a boxy rectangle, like a small external hard drive — sewn into Murphy’s chest cavity, the electrode would stimulate the region of her brain that the doctors believed to be responsible for her depression. The device, known as a deep-brain stimulator (DBS), is meant to regulate neural activity and bring the brain’s patterns back to normalcy. A wire from the pulse generator snakes up to the electrode, carrying electricity, which the electrode then transmits to the brain.

The doctors installed the electrodes and turned them on.

For Murphy, the moment was astonishing. A warmth surged through her. Everything felt lighter, clearer. But then those sensations stopped. The doctors had cut the current so that they could finish wiring the circuit, close her cranium back up, and insert the permanent pulse generator into her chest.

After the surgery, Murphy spent a few days in recovery, and then the doctors sent her home. She would need to heal for three weeks, they told her, before they could turn her device back on. Back at home, returned to the gray world of her depression, Murphy remembered that warm, light, clear feeling. I wish that could be forever, she thought.


Deep-brain stimulation, in its modern form, was first used in 1987. A decade later, the Food and Drug Administration approved the technique to treat a neurological disorder known as essential tremor, as well as tremor in Parkinson’s disease. Soon afterward — in 2002, 2003, and 2009, respectively — came approvals for other Parkinson’s symptoms, dystonia, and obsessive-compulsive disorder (OCD). The FDA, however, has yet to approve the technique for depression.

Attempts to make deep-brain stimulation work for depression gained momentum in 2005, after Helen Mayberg, a neurologist at Emory University, in Atlanta, Georgia, targeted a region called Brodmann area 25, which behaves abnormally in treatment-resistant depression. In a limited trial, Mayberg stimulated that region of the brain in six participants. Four experienced a remission of their depression.

Mayberg remembers the first patient, a nurse who saw the treatment as her last — if unlikely — shot at getting better, and thought that her participation might help the science. But when the surgeons stimulated Brodmann area 25 in her brain, the void she had felt for so long suddenly disappeared — an absence of an absence. Mayberg has seen similar sudden responses over and over again. The doctors at Mass General have seen it too. One of their patients started telling jokes; another laughed for the first time in a decade — hints that these patients might eventually be able to re-engage with their families, find employment, go on vacation, and even have some fun. “Their brain no longer holds them hostage,” Mayberg says.

After Murphy had healed at home, the doctors turned her device on again. At first, not much changed, unlike her original experience at the hospital. She remained depressed enough that she didn’t even feel disappointed. But after a few months, she began to feel differences. All treatments for depression, including medications and psychotherapy, seem to work the same way: They can take months to get going, although the science of why is still in question.

Once Murphy’s device began to work, there were little things she was able to do again. Like talking. Her voice, so unused to being used, sounded soft, like it came from a small animal. She started walking her dog. “I just kind of came back around,” she says.

This happened for a patient named John too. (John’s name has been changed, due to the sensitive nature of his illness.) At 56, he’d had OCD and depression since he was a teenager, and he’d been hospitalized seven or eight times. He’d tried medicine, electroconvulsive therapy, and even a vagal nerve implant, a device that stimulates one of 12 pairs of cranial nerves. Sometimes the treatments worked for a while, but he always ended up where he started. Then, in 2006, he read about the deep-brain stimulation team at Mass General and realized they were only an hour away. At the time, the team was running an OCD trial, for which John didn’t qualify because of his concurrent depression. But he elected to get the implant anyway. His insurance wouldn’t cover the procedure, which cost about $200,000, but his employer, a large manufacturing company, very generously fronted him the money.


It was worth the investment. As soon as the doctors turned his implant on, he felt different, cured — normal. The suddenness of the shift, he says, was bizarre. Gone, he says, were “the ball and chain” that he’d been “dragging around” for decades. “I only wish it had come 30 or 40 years ago,” he continues. “I would have had a different life.” His wife says it may have saved them. “John’s DBS has given the both of us our lives back,” she says.

These success stories are touching. But deep-brain stimulation doesn’t always work for depression. On a large scale, in fact, it has been so unsuccessful that at least two trials have been discontinued, including the 2013 Brodmann Area 25 Deep Brain Neuromodulation trial, overseen by St. Jude Medical. A mid-study analysis reportedly revealed that the trial had a maximum 17.2 percent chance of succeeding. Nonetheless, new research projects are underway, some funded by the Obama administration’s BRAIN Initiative, which has invested millions in research designed to provide a real-time understanding of how the brain works in sickness and in health. Agencies such as the National Institutes of Health and the Defense Advanced Research Projects Agency have used some of this money to fund deep-brain-stimulation projects.

DARPA, in particular, has taken a special interest in the procedure. That’s because the military has a vested interest in the treatment of mood disorders. Soldiers who come back from combat often suffer from post-traumatic stress disorder or depression. When BRAIN money became available, scientists at the agency wondered how they could use it to help those veterans. And by the fall of 2013 they had a plan: DARPA announced a program to develop “an implanted, closed-loop diagnostic and therapeutic system for treating, and possibly even curing, neuropsychological illness.” They called it Systems-Based Neurotechnology for Emerging Therapies (SUBNETS), and they put out a call for proposals to develop it. The desired device would monitor the brain’s activity in real-time and then automatically and electrically adjust it in response, rather than constantly stimulating a single spot, as Murphy’s and John’s devices do.

The call for proposals soon caught the attention of the three researchers at Mass General: the psychiatrist Darin Dougherty, the neurosurgeon Emad Eskandar, and the neuroengineer Alik Widge. This was the team that, in 2006, had designed and surgically implanted Liss Murphy’s device. They realized they were perfect candidates for the DARPA money. Already, they had the neurological expertise, the surgical experience, and the technological know-how to develop what DARPA had in mind — a next-generation version of their original device. Not only that, they were just across the Charles River from the Draper Laboratory, a high-tech engineering and research laboratory that they knew could engineer and build it.

DARPA call in hand, the Mass General team walked over to Draper’s headquarters and proposed working together on the problem. Draper agreed to a collaboration, and the two groups got to work on a formal proposal, which they turned in at the end of the year. The following March, they received good news: DARPA liked their approach and was awarding them a $30 million grant.


After winning the grant, Eskandar, Dougherty, and Widge rounded up the most normal people they could find — like, really, really normal. They show no signs of any mental or mood disorders. “We’ve run them through every psychological interview we can,” Widge says. “We’ve run them through every symptom interview we can. They say: ‘Nope, I’m doing great. I feel good.’”

The team found 36 of these people and put them through another battery of tests, to understand their responses to a variety of situations: How rigid or flexible is normal thinking? How dynamic or flat is a normal person’s emotional range? And what’s a normal balance between approaching and avoidant thinking? Theoretically, a perfectly mentally healthy person should fall about in the middle of all three domains. Somewhat tautologically, the team used these 36 normal people’s responses to define and codify what “normal” looks like, in terms of brain activity, in each domain.

To find out what the dysfunction looks like, on the other hand, they turned to epilepsy patients, who often have associated mood disorders. These patients spend a lot of time bored in their hospital beds, stuck watching daytime television, so when a hospital epilepsy expert walks in and says, “Would you like to play a gambling game against the computer?” (adding “and have your brain activity measured at the same time” soon after), they often say yes.

The doctors repeated the tests done on the 36 “ultranormals” to see how the epileptic people’s responses, and the map of the brain activity behind them, differed. How does “overly emotional” look different from “appropriate emotional response” in a brain scan? Which regions are more active, or dulled? Right now, the team is in the midst of this data-gathering and number-crunching. They need to figure out how to nudge the brain from over- or under-responsive to typically responsive in all three domains. With the new implant, if someone is two standard deviations too emotionally dynamic, for example, the doctors should be able to map the problem, and the device will then stimulate the brain back into the normal range. Preliminary results, published in Experimental Neurology in January of 2017, show proof of the concept that the doctors can identify and modulate brain networks that deal in emotions. This initial research also points specifically to an emotional linkage between the prefrontal cortex, the cingulate cortex, the insula, and the amygdala.

Dougherty, Eskandar, and Widge are convinced that, with this sort of knowledge, they can create an implant that will work for most people with psychiatric illnesses that resist other forms of treatment. The key, they think, will be creating a smart, self-adjusting system — exactly what DARPA called for. Their dream system would have to record brain activity all the time. It would sense when and how a symptom begins to rear its head inside yours. It would zap all the right places at the right times to combat disordered thinking, and in real-time would return your brain to its normal state.

But to make such a device, of course, you’d need to know the neuro-signatures of a number of symptoms — and scientists are still working on that. Major depressive disorder, for example, manifests itself in an unpredictable variety of ways. Most disorders don’t have a fixed set of symptoms. “You have increased sleep or decreased sleep, increased appetite or decreased appetite,” Widge says. “You might be feeling sad and crying all the time, or you might be emotionally completely flat, and the world just feels gray and nothing.” These different states involve different brain activity and different misfiring circuits, each of which causes different symptoms. Perhaps deep-brain stimulation hasn’t yet been regularly successful in treating depression because it only stimulates one spot — like Brodmann area 25 — and isn’t dynamic or holistic enough to account for multiple neural circuits and their real-time interactions.

There are bigger questions underlying this research, of course: Who decides what that management looks like, and what’s “normal”? And do we want to get into the business of stimulating the brain to restore, achieve, or enhance whatever normal is? These are similar to the ethical quandaries psychiatric prescribers and patients have to confront, but in permanently implanted form. Like, “How much of my feelings and behavior are me, and how much are my implant?”

Widge remembers one patient at Mass General, for example, who at a certain stimulation level would start laughing at everything Widge said. Her husband pointed it out.

“Well, he’s a funny guy,” the wife responded.

“A minute ago you said he was completely boring,” the husband said.

“Well, he changed what he was saying.”

Widge hadn’t changed anything, and he isn’t that funny. The setting was too high. He had made the woman euphoric, and she had immediately rationalized that euphoria as a reasonable response to changing events in the outside world.

In a 2001 study looking into electrical stimulation to calm symptoms of Parkinson’s disease, researchers noted that too much stimulation “induced funny associations, leading to infectious laughter or hilarity.” If you can make someone laugh, presumably you can also make them angry or violent — or just about anything else. Similar treatments have had such side effects as hypomania, depression with transient suicidality, worsened mood, and manic episodes with psychotic symptoms.

Now imagine if those alterations were purposeful, rather than merely being byproducts. Given that DARPA is a major funder of deep-brain-stimulation research, it’s not hard to take this a step further and worry that the military, among other organizations, could abuse this power. If a device can sense sadness and suppress it, replacing it with euphoria, it surely someday will be able to do the same for other states of mind. The military might put implants in soldiers in order to push them away from normal, so that, during combat, say, they become fearless and extraordinarily aggressive.

Governmental agencies in the United States have a history of questionable mind-control experiments, like giving LSD to unwitting citizens without their knowledge; they lobotomized more than 1,400 World War II veterans with psychiatric illness. But Jonathan Moreno, a bioethicist at the University of Pennsylvania, argues that, on a practical level, we are a very long way from having to worry about deep-brain stimulation in this way. Moreno has written a book, Mind Wars: Brain Science and the Military in the 21st Century, about the influence of the military on brain science and vice versa. Brain surgery, he points out, is brain surgery: It’s expensive and complicated and can’t be done just anywhere. “You always have to ask yourself, ‘What are my goals, and how can I accomplish my goals in the safest and most effective way?’” he says. “To jump to a device that is in somebody’s head that has to work the right way all the time and has to be maintained in the field — you’re just not going to do that.” Conventional training is a much better option.

From the beginning of its involvement in deep-brain stimulation, DARPA has involved an independent “ethical, legal, and social implications panel” to give feedback on the work. According to the U.S. Department of Defense, the panel includes the “academic community, medical ethicists, and clinical and research scientists.”

But scientists have a history of not quite thinking through the implications of their work. Those who first investigated fission didn’t necessarily intend for an atom bomb to drop on Japan. The researcher who developed the growth stimulator later used in Agent Orange didn’t intend to cause birth defects in Vietnam. Scientists can see — and are perhaps more motivated to see — the good in their work, but not always the potential bad, especially in the future, when it might be cheaper and easier to use. “Everything you learn about the body, you’re learning for many different purposes,” Moreno says. “Neuroscience is to a very great extent dual-use, or many-uses. There are people worried that what we learn to cure people could also hurt people.”

Still, the Mass General team believes the benefits of deep-brain stimulation outweigh the potential for misuse. “A better question,” Eskandar says, “is whether it would be unethical to withhold it from people who are deeply suffering.”

They do acknowledge, though, that their work has great power, and that as doctors they have great responsibility. “DARPA aside, we know there’s a history of psychiatric neurosurgery that’s not great — that’s actually bad,” Eskandar says. “It’s definitely very important for us not to replicate any of the mistakes that have happened in the past — that is, treat people against their will, or not document what we’re doing. So we’re very, very mindful of that.”

Widge, Dougherty, and collaborators published a 2016 paper in the journal Brain-Computer Interfaces that explores ethical concerns. It featured results from interviews with DBS patients, investigating questions like whether the devices allowed them to feel like their “authentic selves,” and how the devices affect their relationships. While some participants felt that a self-adjusting device could allow their true personhood to emerge, others worried about the device’s ability to create artificial emotions. Still, most were willing to trade a possibility like “emotional blunting” for feeling better. A significant concern, though, was that the treatment “might exacerbate what respondents currently experienced as a shifting of blame to them or the functioning of their device, with family or others pointing to the device as the reason for experiencing an undesirable emotion.” The researchers suggest better education for both the participants and their caregivers and family members.

In the lead-up to a DBS surgery at Mass General, doctors work hard to make sure patients are a good fit for the surgery and understand what they’re getting into. Every potential patient comes with a big pile of paper. One member of the Psychiatric Neurosurgery Committee reviews the pile and presents to the rest of the committee, which discusses how likely the surgery is to work for this particular person. “We’re either going to give thumbs-up or thumbs-down,” Dougherty says, although usually they first ask for more information. After that, the candidate comes onsite and meets with psychiatrists, neurologists, and neurosurgeons. “There’s quite a bit of vetting going on,” Dougherty says.

Of referred patients, fewer than half make it to the surgery stage. Those who do have to give informed consent — where the informing and consenting process can take hours — on video, in a room with an “independent consent monitor” along with the doctors. “They have to understand what’s being proposed and agree to it, so we’re not forcing surgery on someone who doesn’t want it or doesn’t understand it,” Eskandar says.

Creating super-normal humans or too-happy humans or anything other than “normal” humans is not the doctors’ goal. “We don’t want to make people euphoric or manic,” Eskandar says. “Happy things should make you happy. Sad things should make people sad. We want to put you back in the middle, where you can have a sort of normal range of emotions.”

But, Moreno says, the knowledge they and DARPA gain of how the brain works in real-time, to create different states of being, will exist in the world, independent of their use of it. “The long-term question is, ‘Will somebody learn something about the brain that will be of use later?’” he says. “And the answer is, ‘Of course.’”


When you walk into Draper Laboratory, the receptionist asks for your ID and checks to make sure you’ve been granted visitor clearance. The halls are quiet, with people typing at desks, 3-D printing their own parts, and doing who-knows-what behind big industrial doors. After an elevator ride upstairs and a walk down a long hallway, you arrive at the SUBNETS area, where, in addition to working on the DBS implant, researchers are developing the technologies that deliver sensation to amputees with prosthetic limbs. One hand, cut off at the wrist, faces the ceiling palm up. Another reaches toward a computer screen. A full-body “patient model” (known outside the lab as a dummy), wearing a modest gown, stares at the whole scene. The engineers keep it around and have rigged it up to an implant as a reminder that they are working on a device for humans, and not just a device.

On a nearby countertop, in visitor-conscious displays, early prototypes of the DARPA-funded deep-brain stimulator sit alongside the current model. At the project’s beginning, the real-time processing of neural signals and the subsequent adjustment required a computer-sized hub — which doesn’t work for humans trying to live their lives. The team has shrunk the latest version, which includes the battery as well as the hardware and software for self-adjustment, to something the size and shape of a small wallet. From here, five flexible cables snake out toward satellite nodes. Each satellite, says Draper program manager Philip Parks, “offers bidirectional communications,” recording the brain’s activity and delivering electrical stimulation as needed.

At the satellites’ ends sit five electrodes. They will eventually live in five different spots in the brain, whose locations the doctors customize to each individual. Versions predating the DARPA project (of the sort that the Mass General team used on Liss Murphy) had just one such stimulating node, meant to zap one region of the brain constantly. But the new satellites will monitor the brain’s activity, sending status updates back to the hub, which will then instruct them about when, where, and how much to zap to bring a patient’s brain back into precise alignment. In Murphy, the hub rests in the chest. The final version will be small enough to live inside the skull.

But we’re not there yet. So far, only a prototype of the implant exists, and clinical trials haven’t begun. This year, the doctors hope to start testing a preliminary version of the device outside the human body, temporarily connecting it to electrodes already in patients’ brains to make sure it records signals correctly.

Not long ago, Liss Murphy and several other DBS implantees gathered at Mass General to give researchers their thoughts about what should come next for deep-brain stimulation. When Murphy walked into the room, she looked around at everyone in the group. All of them, she knew, had touched the void and come back, and they all now had the same device buzzing in their brain.

The researchers passed around pictures of the newer model of the device that was inside their heads. The difference was clear. Seeing the new device was like holding a new iPad in your hand and knowing that you have to go home to a 10-year-old Dell. “I feel so outdated,” Murphy says, reflecting on that difference. “It’s so funny.”

But she doesn’t want the fancier version. Her old one works just fine. “I love my settings,” she says. “It sounds weird to say, but I do.”