Woman sitting in front of laptop with head in hands to illustrate depression
Credit: Jamie Grill/Getty Images

By analyzing the brain activity of patients undergoing deep brain stimulation (DBS) researchers from Georgia Institute of Technology, the Icahn School of Medicine at Mount Sinai, and Emory University School of Medicine have identified a unique brain activity pattern that could serve as a measurable indicator of treatment-resistant depression recovery.

In 2021, nearly three million Americans were estimated to be suffering from treatment-resistant depression (TRD)—a subset of major depressive disorder. DBS, a therapy involving the implantation of electrodes in the brain to stimulate neural activity, has shown promise in addressing this condition. Previously used primarily for movement disorders like Parkinson’s disease, DBS for depression remains experimental.

Reporting in Nature, scientists have now used DBS data together with artificial intelligence to directly monitor antidepressant effects throughout TRD treatment, providing doctors with a readout of disease status at any given time. According to the researchers, the method was even able to distinguish between typical day-to-day mood fluctuations and the possibility of an impending relapse of a depressive episode.

“Understanding and treating disorders of the brain are some of our most pressing grand challenges, but the complexity of the problem means it’s beyond the scope of any one discipline to solve,” said Christopher Rozell, PhD, professor of electrical and computer engineering at Georgia Tech and co-senior author of the paper.

“This research demonstrates the immense power of interdisciplinary collaboration. By bringing together expertise in engineering, neuroscience, and clinical care, we achieved a significant advance toward translating this much-needed therapy into practice, as well as an increased fundamental understanding that can help guide the development of future therapies.”

The study involved ten patients with severe TRD, all undergoing DBS at Emory University. Through six months of data analysis, the team identified a common biomarker that altered as each patient progressed towards recovery. After half a year of DBS therapy, an astounding 90 percent of subjects experienced significant improvements in their depression symptoms, with 70 percent no longer meeting depression criteria.

The high response levels observed within this study’s participant group empowered the researchers to create algorithms referred to as “explainable artificial intelligence.” These algorithms offer human insight into the decision-making mechanisms of AI systems, proving invaluable in helping the team recognize and understand the different neural patterns that distinguished a “depressed” brain from a “recovered” one.

“The use of explainable AI allowed us to identify complex and usable patterns of brain activity that correspond to a depression recovery despite the complex differences in a patient’s recovery,” explained Sankar Alagapan, PhD, a Georgia Tech research scientist and lead author of the study. ”This approach enabled us to track the brain’s recovery in a way that was interpretable by the clinical team, making a major advance in the potential for these methods to pioneer new therapies in psychiatry.”

Additionally, the study reaffirms a long-held observation by psychiatrists: as patients’ brains heal and their depression lessens, their facial expressions undergo noticeable changes. The AI tools developed by the team effectively identified these patterns in facial expressions, proving more reliable than current clinical rating scales.

Encouraged by these initial results, the team is in the process of verifying their findings in another cohort of patients at Mount Sinai. They aim to translate these discoveries into a commercially available version of the dual stimulation/sensing DBS system, opening up new avenues for treating treatment-resistant depression and offering hope to those who have long struggled with this debilitating condition.

Also of Interest