Researchers at the University of Michigan have greatly reduced the power requirements of nerve junctions by adapting them to a subset of brain waves, while improving their accuracy ̵1; the discovery could lead to long-term brain implants that can treat neurological diseases and enable mind-controlled prosthetics. and machinery.
The team, led by Cynthia Chestek, an associate professor in biomedical engineering and undergraduate studies at the Institute of Robotics, estimated a 90% reduction in energy consumption at nerve junctions using their method.
“Right now, interpreting brain signals according to someone’s intentions requires computers as high as humans and a lot of electricity – road car batteries,” said Samuel Nason, the first author and doctor of the study. candidate in the Chestek Cortical Nerve Prosthetics Laboratory. “Reducing the amount of electricity to some extent will, in the long run, enable the use of brain-machine interfaces at home.”
Neurons, our brain cells that transmit information and actions throughout the body, are noisy transmitters. Computers and electrodes used to collect neural data listen to a radio station located between the stations. In the midst of a brainstorm, they have to decipher the real content. Complicated by these tasks, the brain is the focal point of this data, which increases power and processing beyond the limits of safe implantable devices.
At present, scientists can use transcutaneous electrodes or route wires through the skin to the brain to predict complex behaviors, such as grabbing a hand object from neuronal activity. This can be accomplished by using 100 electrodes that capture 20,000 signals per second and allow feats such as re-enabling the paralyzed arm, or allowing the prosthetic person to feel how heavy or soft the object is. This method is not only impractical in a laboratory environment, but also poses a risk of infection.
Some wireless implants, developed using highly efficient, application-specific integrated circuits, can achieve almost the same efficiency as transcutaneous systems. These chips can collect and transmit about 16,000 signals per second. However, they have not yet achieved consistent performance, and their customization is a barrier to obtaining approvals as safe implants compared to industrially manufactured chips.
“It’s a big leap forward,” Chestek said. “To receive the high-speed signals currently required for brain-computer interfaces wirelessly would be completely impossible given the power supplies of existing pacemaker-style devices.”
To reduce the need for energy and data, researchers are compressing brain signals. Focusing on spikes in nerve activity that cross a certain power threshold, called the threshold transition rate, or TCR, means that less data needs to be processed before it can be predicted whether neurons will shoot. However, to determine when a threshold has been exceeded, the TCR needs to hear the entire focal point of neuronal activity, and the threshold itself can change not only from one brain to another on different days, but also within the same brain. This requires adjusting the threshold and requires additional hardware, battery and time.
After compressing the data in another way, Chestek’s lab called for a specific feature of the neuronal data: increasing bandwidth. SBP is an integrated range of frequencies from multiple neurons from 300 to 1,000 Hz. By listening only to this frequency range and not paying attention to others, taking data from straw rather than a hose, the team found a very accurate prediction of behavior when energy needs are significantly lower.
Compared to transcutaneous systems, the team found that the SBP technique is just as accurate at one-tenth of both signals and 2,000 compared to 20,000 signals per second. Compared to other methods such as using threshold transition speed, the team approach not only requires much less raw data, but is also more accurate in predicting neuronal firing even in the case of noise and does not require setting a threshold.
The SBP command method solves another problem that limits the useful life of an implant. Over time, the interface electrodes fail to read signals between noise. Because this technique works just as well when the signal is half of what is needed from other methods, such as threshold switching, implants can be left in place and used for longer.
While new brain-machine interfaces can be developed to take advantage of the team approach, their work also opens up new possibilities for many existing devices, reducing the technical requirements for turning neurons into intentions.
“It turns out many devices don’t sell themselves for long,” Nason said. “These existing circuits, using the same bandwidth and power, are now adaptable to the entire area of brain-machine interfaces.”
The study “Low-power neuronal jump activity band, dominated by local individual units, improves the functioning of brain-machine interfaces.” Natural biomedical engineering.
Can a possible prosthesis one day be controlled by human thought?
The low-power neuronal jump activity band, dominated by local units, improves the functioning of brain-machine interfaces, Natural biomedical engineering (2020). DOI: 10.1038 / s41551-020-0591-0, www.nature.com/articles/s41551-020-0591-0
Submitted by the University of Michigan
Citation: Ultra-low power brain implants have a significant signal of gray matter noise (2020, July 27), received in 2020. July 29 From https://medicalxpress.com/news/2020-07-ultra-low-power-brain-implants-reaningful .html
This document is protected by copyright. Without fair dealing for the purposes of private study or research, no part may be reproduced without written permission. Content is provided for informational purposes only.