Brain-computer interfaces (BMIs) are revolutionizing the way people with severe disabilities interact with the world. these remarkable devices allow individuals who are unable to move or speak to communicate simply by thinking. Imagine a quadriplegic person effortlessly moving a cursor on a computer screen or controlling a robotic limb with the power of their mind – this is the promise of BMI technology.
How does it work? A tiny implant, containing an array of microelectrodes, is placed in the brain. These electrodes pick up the neural signals associated with specific thoughts or intentions. A computer then interprets these signals and translates them into commands that can control a computer cursor, a robotic arm, or other devices.
While BMIs hold immense potential, a significant challenge has been the degradation of the implanted microelectrode arrays over time.”Not only do we observe day-to-day variations, but over time the performance of brain–computer interfaces degrade for a variety of reasons,” explains Azita Emami, the Andrew and Peggy Cherng Professor of Electrical Engineering and Medical Engineering at Caltech. “There may be a small movement of the implant or its electrodes. The electrodes themselves may deteriorate or become encapsulated in brain tissue. Some people think that over time the neurons move away from the implant as they react to it as a foreign object in the brain. For whatever reason, the signals we receive become noisier.”
This signal degradation makes it increasingly challenging for BMIs to accurately interpret the user’s intent. Conventional methods,which relied on detecting strong neural spikes,become less effective as the signal weakens.
Now, Emami and her team have developed a groundbreaking solution using machine learning.”Where before we relied on counting neural spikes, we have now created a neural network that automatically extracts information from the entire neural signal, from all the little dips and picks and changes in the signal, and converts this into the intent of the patient,” says Benyamin Haghi, formerly a graduate student in the emami lab.
Emami adds, “Over time the BMI has been trained on both a signal that is neural activity and a signal that looks like noise, and is thus able to interpret the user’s intent.”
The results have been remarkable. One participant, JJ, who lost mobility due to a vehicle accident, was able to continue using his three-year-old implant thanks to this new algorithm. “When we first started working with him, his implant was three years old, and it had already degraded. We were thinking of removing the implant, but with our new algorithm, he’s happy with continuing to use the system he already has. JJ can move a cursor very precisely on a grid, just as he did when the implant was new. He can play video games and control a computer habitat that mimics driving,” Emami shares.
This breakthrough in BMI technology brings us closer to a future where individuals with severe disabilities can regain independence and control over their lives. Emami’s work is a testament to the power of innovation and collaboration in pushing the boundaries of what’s possible.
A groundbreaking new algorithm developed by researchers at the California Institute of Technology (Caltech) is revolutionizing the field of brain-machine interfaces (bmis). Called FENet, short for Feature Extraction Network, this innovative technology has the remarkable ability to decipher brain signals with unprecedented accuracy, paving the way for more intuitive and effective control of prosthetic limbs and other assistive devices.
What sets FENet apart is its ability to learn from the neural data of a single patient and then generalize that knowledge to other individuals. “This means that there is some basic type of information in the neural data that we are picking up,” explains Mohammad Emami, a postdoctoral scholar in the Andersen lab and lead author of the study.
“FENet has already extended our clinical study with JJ by two years,” says Richard Andersen, the James G. Boswell Professor of Neuroscience and director of the T&C Chen Brain-Machine Interface Center at Caltech. “BMI research is a perfect field for interdisciplinary research, in this case melding the disciplines of engineering, computer science, and neuroscience.”
Current bmis rely on a complex and cumbersome system of wired connections between the brain implant,a connector,and a microsystem that processes the raw data before sending it to a computer.Emami envisions a future where FENet enables the miniaturization of this system, leading to wearable or implantable devices that communicate wirelessly with computers.
“This is quite a cumbersome system,” Emami says. “Our goal now,after the creation of FENet to better interpret brain signals,is to miniaturize the system so that one day it can be a wearable or an implant that communicates wirelessly with the computer.”
The research, published in Nature Biomedical Engineering under the title “Enhanced control of a brain-machine interface by tetraplegic participants via neural-network-mediated feature extraction,” highlights a significant leap forward in BMI technology. The study’s co-authors include members from the Emami and Andersen labs, as well as researchers from UCLA Neurosurgery. Funding for the project was provided by the National Institutes of Health, Caltech’s S2I, the T&C Chen Brain-machine interface Center at Caltech, the Boswell Foundation, the Braun Foundation, and the Heritage Medical Research Institute.
## Brainwave Breakthrough: Machine Learning Gives New Life to Failing Brain Implants
**World Today news Exclusive Interview with Professor Azita Emami**
**World Today News:** Brain-computer interfaces (BMIs) hold incredible promise for people with severe disabilities, allowing them to regain control over their habitat through the power of thought. However, a major challenge has been the degradation of implanted microelectrode arrays over time, leading to declining performance.Can you tell us about the innovative solution your team has developed at Caltech?
**Professor Azita Emami:** Your right, signal degradation has been a persistent obstacle in the advancement of BMI technology.Customary methods relying on detecting strong neural spikes become less effective as the signal weakens over time.
Our team has turned to machine learning to overcome this challenge. We’ve developed a neural network algorithm that analyzes the complete neural signal – not just the spikes – extracting meaningful details from the subtle dips, peaks, and variations. This allows us to interpret the user’s intent even when the signal is weak or noisy.
**World Today News:** How dose this machine learning approach differ from previous methods?
**Professor Azita Emami:** Think of it like this:
previously, we were trying to decode a sentance by only focusing on the capitalized words. Now, we’re looking at the entire sentence, including the punctuation, the subtle changes in tone and grammar, to understand the full meaning.
Our algorithm has been trained on both strong and noisy signals, allowing it to effectively recognize patterns and accurately interpret the user’s intentions even as the signal degrades.
**World Today News:** What kind of impact has this had on users like JJ, who participated in your study?
**
Professor Azita Emami:** JJ’s case is a perfect example of the potential of this technology. His implant was three years old and significantly degraded when we started working with him. We were considering removing it, as it was no longer functioning effectively.
But thanks to our new algorithm, JJ can now continue using his existing implant. He can control a computer cursor with precision, play video games, and even operate a virtual driving simulator. It’s remarkable to see how this technology has given him back a sense of control and independence.
**World Today News:** This is truly groundbreaking! What does the future hold for this technology?
**Professor Azita Emami:** This breakthrough brings us closer to a future where BMIs can be used effectively for longer periods, empowering individuals with disabilities to live fuller and more independent lives. We are constantly refining our algorithms and exploring new applications for this technology, with the ultimate goal of making it widely accessible and transformative for those who need it most.
**World Today News:** Thank you, Professor Emami, for sharing your groundbreaking work with us. Your research offers a beacon of hope for individuals with disabilities, unlocking new possibilities and paving the way for a more inclusive future.
