Out Of Options: The Drugs Don’t Work. Therapy Doesn’t Work. What’s Left?

You may know someone like Drew. For no reason at all, a shadow can fall across her world, and stay there for months at a time.

Under the shadow, she struggles — to get out of bed, to return messages from friends, to meet work deadlines. Some days, she struggles to keep going at all.

Drew has tried talk therapy and has cycled through too many medicines to keep track, but nothing has given her relief from these periods of major depression. And each time treatment fails, she loses a little more hope.

Now imagine Drew has a smartwatch that monitors her health and symptoms in real time, allowing her doctor to personalize her treatment and thus increasing the likelihood of success. Or that a tiny brain-computer interface is available that senses depression’s symptoms and swiftly restores balance, delivering the precise amount of electrical stimulation to just the right part of the brain.

This is the vision of Maryam Shanechi, the Alexander A. Sawchuk Chair and professor of electrical and computer engineering, computer science and biomedical engineering at the USC Viterbi School of Engineering. Since her doctoral studies, she has used her engineering expertise to decode signals from the most sophisticated information processing system on the planet — the human brain. Now, she is building on her discoveries with medical researchers from the Keck School of Medicine of USC in the hopes of creating game-changing new therapies for millions of people dealing with neuropsychiatric disorders such as major depression.

“Mental health conditions are a huge burden on society,” said Shanechi, who is the founding director of the USC Center for Neurotechnology. “They’re one of the main causes of disability worldwide. And a lot of patients do not respond to our standard therapies.”

As a world expert on neural engineering, mood disorders and brain-computer interfaces, which allow the brain to interact with and control external devices, Shanechi recognizes the power of collaboration and was eager to partner with medical experts from the Keck School of Medicine.

“I feel like the marriage — bringing together the power of engineering, computer science, psychiatry, neurosurgery and neuroscience — is important for making progress on these complex problems involving the brain and the mind,” she said.

The team is now working on one of the Center for Neurotechnology’s signature programs — a quest to objectively characterize major depression symptoms using multiple measurable signals. Their research combines inputs from electrical activity in the brain, physiological responses and behavioral indicators. Success could yield a real-time readout of how depression treatment is working, enabling clinicians to personalize therapy and expedite lifesaving help.

“I feel like the marriage — bringing together the power of engineering, computer science, psychiatry, neurosurgery and neuroscience — is important for making progress on these complex problems involving the brain and the mind.”

Maryam Shanechi

The unmet need is substantial, as treatment-resistant depression affects approximately 30% of those diagnosed with major depressive disorder.

“Ultimately, what we want to do is get to the most effective treatment as fast as possible, because every day with severe, treatment-refractory depression is a risk for suicide,” said collaborator Steven Siegel, the Franz Alexander Chair in Psychiatry at the Keck School of Medicine of USC.

Teamwork is a necessity

The project has attracted support from the National Institutes of Health’s BRAIN Initiative, with a recent $5 million grant awarded to Shanechi as principal investigator, alongside co-investigators including Siegel and neurosurgeon Charles Liu, director of the USC Neurorestoration Center.

“The BRAIN Initiative supports interdisciplinary approaches that a lot of traditional grant mechanisms do not,” said Shanechi, who, in addition to USC Viterbi, also holds an appointment in the USC Neuroscience Graduate Program. “With huge challenges such as treating mental illness, no single discipline can tackle it on its own. It can only happen when we work together.”

“It comes down to the quality of faculty that USC Viterbi has recruited in this biomedical direction. They’ve had success in attracting outstanding people like Maryam. And what we’re going after is simply not possible unless the resources in engineering and medicine are deployed simultaneously.”

Charles Liu

The research team features USC leaders and rising stars who specialize in measuring electrical brain activity, developing innovative therapies for depression, performing brain implant surgeries, creating computer and artificial intelligence (AI) models of brain activity and mental processes, designing implantable and wearable technologies, and implementing conversational AI agents. The team of Trojan investigators includes:

• Adam Frank, assistant professor of clinical psychiatry and the behavioral sciences at the Keck School of Medicine
• Jon Gratch, research professor of computer science and psychology at USC Viterbi
• Yasser Khan, assistant professor of electrical and computer engineering at USC Viterbi
• Brian Lee, associate professor of clinical medicine at the Keck School of Medicine
• Mohammad Soleymani, research associate professor of computer science at USC Viterbi

They are joined by bioethicists from the University of Washington who are examining the broader societal implications of this project.

According to Liu, the research excellence accruing at USC Viterbi makes collaborations such as this one especially inviting.

“It comes down to the quality of faculty that USC Viterbi has recruited in this biomedical direction,” he said. “They’ve had success in attracting outstanding people like Maryam. And what we’re going after is simply not possible unless the resources in engineering and medicine are deployed simultaneously.”

The power of AI to make meaning of abundant data

Contributions from machine learning, in which algorithms are trained to detect patterns from a flood of data, are essential to the team’s work. After all, the researchers are contending with the immense complexity of the human brain.

“One person could spend their whole life trying to understand a brain cell, or a synapse, or a channel,” said Siegel, who chairs the Department of Psychiatry and the Behavioral Sciences at the medical school. “But there are millions of channels in millions of synapses in tens of millions of neurons, all connected to each other in dynamic loops.”

In addition to brain processes, the researchers are also investigating physiology and behavior, meaning that harnessing the analytic power of AI is a must.

“These are complex mathematical and data analytic problems that we have to solve,” Shanechi said. “Developing machine-learning and AI tools will allow us to disentangle the symptom-relevant brain patterns from everything else. Innovating on the AI front will hopefully move us closer to realizing this dream of accurately tracking symptom states so that we can really tailor therapy to people’s needs.”

Shanechi has already made inroads with breakthrough research that developed AI algorithms to decode mood variations from the brain activity of human participants. But she stresses that the goal of the collaboration is not to read or regulate fleeting emotions, but to detect treatable symptoms of chronic major depression.

“Emotion is a natural part of life,” she said. “We want to detect and regulate symptom states. Once you have a readout of those symptom states, then you can actually personalize various types of therapies.”

Piggybacking on advances in epilepsy treatment

The project is benefitting from advances in addressing another disorder rooted in the brain — epilepsy. Patients with treatment-resistant epilepsy can now manage their condition with the help of an intracranial brain-computer interface, which sits inside the skull without penetrating into the brain. This device, approved by the U.S. Food and Drug Administration, uses electrodes to detect brain activity and deliver stimulation to reduce and prevent seizures, in what’s known as a closed-loop system.

Liu, the neurosurgeon, played a key role in the development of this device and was the first to implant it in a patient. The implantation is less invasive than most brain surgeries, with many patients discharged from the hospital within days of the procedure.

The team will collect data from volunteers who already have received implanted intracranial electrodes for standard monitoring of treatment-resistant epilepsy, which can co-occur with depression. During these participants’ stay in the epilepsy monitoring unit, brain activity is already being continuously recorded, and the researchers will use the data to study how they can decode depression symptom states from brain activity. The ultimate goal is to develop solutions for treatment-resistant depression using brain-computer interfaces.

“There is a unique opportunity in the care of our patients, where their brain recordings are available at no additional risk to them,” said Liu, a USC professor of clinical neurological surgery, surgery, psychiatry and the behavioral sciences, and biomedical engineering. “This is a win-win situation because the collaboration is aimed at solving problems that some of our epilepsy patients have. If we can solve depression as a comorbidity for epilepsy patients, where the brain-computer interface is a treatment, then we can generate the safety and efficacy data that could go toward helping those without epilepsy.”

He and his colleagues at the USC Neurorestoration Center are working to apply their research to provide practical benefits for people with neurological and psychiatric disorders. They aim to bridge the gap between laboratory research and patient treatment, and ultimately seek regulatory approval and commercialization.

“We think that neuroprosthetics are a very exciting new direction for finding the solution to mood disorders,” Liu said. “The pathway moving forward is one that we’ve actually had quite a bit of input on creating.”

How breaking the code translates to fighting depression

In the future, there could be an implant that detects symptoms of depression and automatically delivers electrical stimulation to counteract them, similar to the closed-loop system for epilepsy. But that is just one of many possibilities outlined by the researchers.

For instance, if certain physiological measures reliably predict depression symptoms, monitoring could take place on a smartwatch or other wearable device. And even intracranial electrodes that detect symptoms without a therapeutic function could provide a substantial boon by enabling doctors to adjust treatment based on real-time information about their patients.

“Most people who present as treatment-resistant simply haven’t had the breadth of treatment to meet their needs. Our vision of personalized therapy for mental health takes the wealth of options that we have, and tries to match it with an individual in a way that’s data-driven.”

Steven Siegel

Tracking the course of major depression one brain at a time is well-suited to tackling the disease precisely because a key clinical challenge is the significant variation in how depression manifests among individuals, as with other neuropsychiatric illnesses.

“There are a lot of differences from one patient to the other,” Shanechi said. “We can’t bunch them all together. We have to come up with treatment plans that work for each individual on their own.”

Siegel, the psychiatrist, points to effective treatments that are currently available, but not widely accessible. Beyond the familiar options of medicine and talk therapy, there’s evidence that stimulation of the brain with electricity or with magnetism could help close the gap for people with treatment-resistant depression. However, in practice, the dictates of insurance providers often keep patients on ineffective protocols for too long or create obstacles in escalating to newer modes of therapy.

“Most people who present as treatment-resistant simply haven’t had the breadth of treatment to meet their needs,” Siegel said. “Our vision of personalized therapy for mental health takes the wealth of options that we have, and tries to match it with an individual in a way that’s data-driven.”

The objective signals for depression that the team is seeking could guide mental health providers in matching treatment to outcome. For Shanechi, that concern for the lives — and quality of life — of people with neuropsychiatric disorders is the major motivating factor for the BRAIN Initiative-funded collaboration, as well as the reason why she founded the USC Center for Neurotechnology.

“When all is said and done, we’re trying to help millions of patients who really have no other options,” she said.