- BCI Award 2019 Winner: Closed-loop BCI for the treatment of neuropsychiatric disorders
BCI Award 2019 Winner: Closed-loop BCI for the treatment of neuropsychiatric disorders
Maryam, Yuxiao, and Omid, you submitted your BCI research “Developing a closed-loop brain-computer interface for treatment of neuropsychiatric disorders using electrical brain stimulation” to the BCI Award 2019 and won 3rd place. Could you briefly describe what this project was about?
Maryam: Neuropsychiatric disorders such as major depression are a leading cause of disability worldwide. Currently, about 20-30% of major depression patients are not responsive to any available treatments. This is about 5 million people in the US alone, who could greatly benefit from a novel treatment. Motivated by the view that neuropsychiatric disorders are signs of abnormal brain network activity, we aim to provide a new treatment in the form of a novel closed-loop BCI that aims to normalize these brain activity patterns using electrical stimulation.
To do so, we take a principled engineering approach to designing this brain-computer interface in our lab. Our aim is to design a BCI that can decide in real time how to stimulate the brain guided by novel decoders that track a patient’s mood symptoms from their brain activity. The goal of the BCI is to alleviate mood symptoms by applying the stimulation at the right time, with the right amount, and in a manner that is precisely tailored to the patient’s needs.
What technologies did you use?
Yuxiao: We take a principled engineering approach to the problem of developing a closed-loop BCI for neuropsychiatric disorders. We use tools from machine learning and control theory to develop novel mathematical models that explain how a patient’s mood is represented in their brain activity and how stimulation should be applied to normalize the abnormal brain activity patterns. For example, to decode mood from brain activity, in collaboration with Edward Chang’s team at UCSF, we recorded multi-site intracranial electrocorticography (ECoG) signals from seven epilepsy patients and concurrently measured their mood using a validated self-report questionnaire. Then, we developed a novel model that can match each patient’s ECoG signal with the mood report.
Based on this model, we built a decoder that automatically estimates real-time mood variations from the patient’s brain activity. Our goal is to develop a controller that optimally adjusts the amount of electrical stimulation in real-time to alleviate mood symptoms, thus realizing a closed-loop brain stimulation system.
How would therapy for the treatment of neuropsychiatric disorders look like using your invention?
Omid: We envision that in the future, our BCI can provide a precisely-tailored alternative therapy for treatment-resistant major depression. The BCI will tailor the delivery of electrical stimulation to each individual patient’s need by monitoring their symptoms in real time based on their brain activity. The goal will be to alleviate symptoms by applying electrical stimulation only when needed, and only with the minimal optimal amount needed.
From the patient’s point of view, the treatment will be similar to how standard deep brain stimulation (DBS) systems are currently implanted to treat Parkinson’s disease for thousands of patients each year. But, of course, the algorithmic technology within the implant will need to enable closed-loop stimulation and address the distinct challenges for neuropsychiatric disorders. The vision is that after implantation, the patient will be able to go back to their normal life while the device will keep providing the right amount of therapy. Of course, we still have to do much more to realize this vision in the future. We hope that the progress we have made so far will help facilitate such a BCI.
Do you work together with other institutions?
Maryam: Yes, the mood decoding work was a close collaboration with Edward Chang’s team at UCSF. We believe collaborations are key to success in this truly interdisciplinary domain of BCI design.
How long will it take to have your technology available?
Omid: We are only at the early stages of moving towards making closed-loop stimulation treatments for neuropsychiatric disorders clinically feasible. So far, we have provided the first demonstration that mood symptoms can be decoded from brain activity. Our next step is to develop models that can decide how to change the stimulation to normalize brain activity patterns underlying the symptoms, and to test these models in animal and human experiments. Finally, we will need to build a real-time optimal controller that can deliver the stimulation at the right time and with the right amount. Once we develop the technology and test it in animal models and human experiments, we need to test them in carefully designed clinical trials to assess efficacy and longevity for a larger group of patients and to validate safety.
How was it to be under the winners of the BCI Award 2019?
Yuxiao: This is a great recognition of our work toward developing novel BCI technologies that can provide alternative new therapies for millions of patients with treatment-resistant neuropsychiatric disorders. We sincerely thank the jury for selecting our work as one of the winners.
Developing a closed-loop brain-computer interface for treatment of neuropsychiatric disorders using electrical brain stimulation
Yuxiao Yang1, Omid G. Sani1, Morgan B. Lee2,3,4, Heather E. Dawes2,3,4, Edward F. Chang2,3,4, Maryam M. Shanechi1,5
1 Ming Hsieh Department of Electrical Engineering, Viterbi School of Engineering, University of Southern California, USA.
2 Department of Neurological Surgery, University of California, USA.
3 Weill Institute for Neuroscience, University of California, San Francisco, USA.
4 Kavli Institute for Fundamental Neuroscience, University of California, San Francisco, USA.
5 Neuroscience Graduate Program, University of Southern California, USA.