This and other early success stories prompted a recent large-scale trial with 90 people with depression. The bad news: on average, the study did not find any improvements after a year of treatment.
But to Dr. Edward Chang, a neuroscientist at University of California, San Francisco (UCSF) who is leading one of the projects, that’s not the end of the story.
Most DBS systems treat depression like the same thing, whether or not the person exhibits different symptoms. These systems pound the brain with constant electrical pulses. The stimulation protocol is set by a physician, rather than according to the patient’s actual brain state.
The key, according to Chang, is to design implants specifically to treat symptoms tailored to each individual—and to turn the system on only when needed.
Together with Dr. Omid G. Sani, an electrical engineer at the University of Southern California, the team developed an algorithm that translates brain waves to subjective feelings of mood.
His team worked with six patients with epilepsy, who already have electrodes implanted into their brains to hunt down the source of their seizures. Over a period of one to three weeks, the team tracked each patient’s brain activity in detail, while monitoring their mood through a standardized questionnaire.
By comparing the two types of information, the team developed an algorithm that extracted a small number of “neural predictors”—dynamic patterns of activity in neural networks—that could accurately predict a person’s ever-changing feelings.
A few “hotspots” emerged, including the limbic system, a previously identified hub for regulating mood and motivation.
“These dynamical mood biomarkers and mood decoding algorithms can provide insight into brain processes underlying mood regulation,” concludes Sani.
This is also the first step towards personalized stimulation therapies for depression based on feedback of the estimated mood, he adds.
According to Nature, Sani and Cheng have already developed a prototype system ready for human tests. Similar closed-loop systems have already been tested in a handful of people, but Cheng stresses that these preliminary results need to be further validated.
Across the Board
The second team, led by neurosurgeon Dr. Emad Eskandar at the Massachusetts General Hospital, takes a slightly different approach.
Dubbed “trans-diagnostics,” the approach looks at commonalities in different psychiatric disorders, rather than a specific mood disorder. The team then develops algorithms that capture and distill brain activity associated with these common traits—for example, increased anxiety, forgetfulness, or lack of empathy.
To Eskandar, brain waves are only a small part of the data. His team also hopes to record the activity from individual neurons, in a bid to identify those that go awry in mental illness.
It’s a highly ambitious goal. If successful, Eskandar would be able to hunt down disease symptoms from single neurons to neural circuits to brain-wide activity—a full, multi-layered atlas for physicians to consult for treatment options.