For more than a decade, researchers at UC San Francisco (UCSF) have been working to advance the use of deep brain stimulation (DBS) as a treatment for neurological and psychiatric conditions. Supported by funding from the National Institutes of Health (NIH), particularly through the BRAIN Initiative, UCSF physician-scientists have developed personalized DBS approaches that respond to individual brain activity.
Deep brain stimulation involves implanting electrodes in the brain to deliver targeted electrical currents. This method has long been used to treat movement disorders such as Parkinson’s disease. However, traditional DBS systems provided continuous stimulation, which often did not adapt well to patients’ changing symptoms and proved less effective for other conditions.
Professors Philip Starr, MD, PhD, and Edward Chang, MD, both from UCSF Neurological Surgery, have led efforts to develop adaptive DBS techniques. These new methods only activate when abnormal brain activity associated with symptoms is detected. “Tailoring these treatments to the person’s neural signature is really the key that allows DBS to be effective across many conditions,” said Chang.
One area where this approach has shown promise is Parkinson’s disease. Shawn Connolly was diagnosed with Parkinson’s at age 39 and found limited relief with earlier technologies. In 2021, he participated in a clinical trial testing an adaptive DBS system developed by Starr and Simon Little, MBBS, PhD. The device used algorithms to detect early signs of symptoms and deliver stimulation only when needed. In February 2024, the Food and Drug Administration approved two similar adaptive DBS algorithms for people with Parkinson’s disease—one based on Little’s work—marking a significant step forward for treatment options.
“It’s definitely changed my life,” Connolly said in 2024. “I can just go through the whole day feeling good.”
The research team continues refining these technologies by developing ways to measure changes related to walking in Parkinson’s patients—a disruptive symptom that could benefit from further improvements in DBS systems.
UCSF also has a history of exploring DBS for chronic pain management. In 1972, Professors John E. Adams, MD, and Yoshio Hosobuchi, MD were among the first in the U.S. to test continuous DBS for persistent pain but found results were often temporary as patients’ brains adapted over time. More recently in 2023, Prasad Shirvalkar, MD, PhD identified biomarkers associated with pain using implanted electrodes and artificial intelligence models capable of predicting pain episodes based on brain signals. This enabled clinical trials of personalized DBS systems designed to activate only when pain markers are present.
DBS technology is being studied as a potential intervention for severe depression as well. Sarah—a participant in one of Chang’s clinical trials—experienced significant improvement after receiving a personalized system tailored using advanced brain mapping techniques at UCSF.
“I was at the end of the line,” she recalled about her experience before joining the study in 2021.” I was severely depressed… It was not a life worth living.” After her procedure: “In the early few months, the lessening of depression was so abrupt… But it has lasted… Those thoughts still come up but it’s just … poof …the cycle stops.”
UCSF Psychiatry Professor Andrew D. Krystal leads ongoing federally funded trials aimed at expanding access if successful.
At present NIH funding supports UCSF as one of about twelve hospitals offering continuous DBS care for individuals with severe obsessive-compulsive disorder (OCD). Drs. Andrew Moses Lee and Krystal are conducting studies aimed at identifying OCD-related biomarkers within brain activity—research that may lead toward more individualized therapies using adaptive stimulation systems.
Chang believes that adapting these treatments based on each patient’s unique neural patterns could eventually expand their application beyond current uses—to conditions such as addiction or Alzheimer’s disease—pending further research.
