Jan 2 2019
Engineers at the University of California, Berkeley (UC Berkeley) have developed a novel neurostimulator that can simultaneously listen to and trigger electric current in the brain, possibly delivering optimized treatments to patients suffering from diseases like Parkinson’s and epilepsy.
In a proposed device, two of the new chips would be embedded in a chassis located outside the head. Each chip could monitor electrical activity from 64 electrodes located into the brain while simultaneously delivering electrical stimulation to prevent unwanted seizures or tremors. (Image credit: Rikky Muller, UC Berkeley)
Called WAND, the novel device functions similar to a “pacemaker for the brain,” tracking the electrical activity of the brain and providing electrical stimulation when it perceives that something is wrong. Such kind of devices can be highly effective at inhibiting debilitating seizures or tremors in patients suffering from many different neurological conditions. However, the electrical signatures preceding a tremor or seizure can be very mild, and the strength and frequency of electrical stimulation needed to prevent them are also equally sensitive. Doctors can take years to make minor adjustments before the devices offer optimal treatment.
WAND, short for wireless artifact-free neuromodulation device, is not only wireless but also autonomous; this means that as soon as it learns to identify the signs of seizure or tremor, it can modify the stimulation parameters on its own to inhibit the unnecessary movements. Moreover, since the WAND is a closed-loop device—that is, it can trigger and record at the same time—it is capable of modifying these parameters in real time.
The process of finding the right therapy for a patient is extremely costly and can take years. Significant reduction in both cost and duration can potentially lead to greatly improved outcomes and accessibility. We want to enable the device to figure out what is the best way to stimulate for a given patient to give the best outcomes. And you can only do that by listening and recording the neural signatures.
Rikky Muller, Assistant Professor, Electrical Engineering and Computer Sciences, University of California, Berkeley.
When compared to eight channels in other closed-loop systems, WAND has the potential to record electrical activity over 128 channels, or alternatively from 128 points in the brain. In order to demonstrate the device, the researchers utilized the WAND to detect and delay particular movements of the arms in rhesus macaques. The device has been detailed in a study published in Nature Biomedical Engineering on December 31st, 2018.
Ripples in a pond
Triggering and recording the brain’s electrical signals simultaneously is more like attempting to view tiny ripples in a pond and at the same time splashing one’s feet—the huge pulses of electricity supplied by the stimulation overwhelm the electrical signals from the brain.
At present, deep brain stimulators either cease recording while sending the electrical stimulation, or they record at a different brain region from where the stimulation is applied—fundamentally determining the tiny ripples at a varied point in the pond from the splashing.
“In order to deliver closed-loop stimulation-based therapies, which is a big goal for people treating Parkinson’s and epilepsy and a variety of neurological disorders, it is very important to both perform neural recordings and stimulation simultaneously, which currently no single commercial device does,” stated Samantha Santacruz, former UC Berkeley postdoctoral associate and now an assistant professor at the University of Texas in Austin.
At Cortera Neurotechnologies, Inc., scientists, headed by Rikky Muller, developed the WAND custom integrated circuits that are capable of recording the entire signal from the mild brain waves as well as the robust electrical pulses. It is this chip design that enables the WAND to deduct the signal from the electrical pulses, leading to a clean signal from the brain waves.
Present-day devices are adjusted in such a way that they can record only signals from the smaller brain waves and are also overwhelmed by the large stimulation pulses. As a result, this type of signal reconstruction is not possible.
“Because we can actually stimulate and record in the same brain region, we know exactly what is happening when we are providing a therapy,” said Muller.
In association with the laboratory of electrical engineering and computer science professor Jan Rabaey, the researchers constructed a platform device that has wireless and closed-loop computational capabilities, which, in turn, can be programmed for use in a wide range of clinical and research applications.
Santacruz, serving as a postdoc at UC Berkeley, and Jose Carmena, electrical engineering and computer science professor, headed experiments in which subjects were taught to utilize a joystick to shift a cursor to a particular location. Following a training period, the WAND device was able to detect the neural signatures that emerged as the subjects prepared to conduct the motion, and subsequently delivered electrical stimulation that delayed the movement.
While delaying reaction time is something that has been demonstrated before, this is, to our knowledge, the first time that it has been demonstrated in a closed-loop system based on a neurological recording only. In the future we aim to incorporate learning into our closed-loop platform to build intelligent devices that can figure out how to best treat you, and remove the doctor from having to constantly intervene in this process.
Rikky Muller, Assistant Professor, Electrical Engineering and Computer Sciences, University of California, Berkeley.
Benjamin C. Johnson and Andy Zhou of UC Berkeley join Santacruz as co-lead authors on the paper. Other contributing authors include George Alexandrov, Ali Moin, and Fred L. Burghardt of UC Berkeley. The study was partly supported by the Defense Advanced Research Projects Agency (W911NF-14-2-0043) and the National Science Foundation Graduate Research Fellowship Program (Grant No. 1106400). Authors Benjamin C. Johnson, Rikky Muller, Jan M. Rabaey, and Jose M. Carmena have financial interests in Cortera Neurotechnologies, Inc., which has filed a patent application on the built-in circuit applied in this study.