There was a scene in the Avatar movie where the protagonist, Jack, establishes a connection with creatures on the planet Pandora with the help of a neural interface encased in hair. He can remotely feel the breathing of the creatures and also share his thoughts with them.
This futuristic-looking image is becoming a reality with the help of brain-computer interface technology.
A few days ago, an American R&D team claimed to have developed a new neural implant – a high-density transparent graphene array. It is a thin, flexible, elongated, ultra-thin polymer film containing tiny electrodes. By simply pressing it against the surface of the brain, it can infer neural activity at the surface as well as deeper levels. This opens up new possibilities for brain-computer interface technology.
The biggest innovation in this study is that the new implant is completely transparent. Researchers can fire a laser directly through the film and rely on two-photon microscopy to map key neuronal components up to 0.25 millimeters below the surface of the brain layer in real-time, which is extremely convenient for researchers and developers.
In addition, the completely transparent implant also provides a clearer field of view for the microscope, reducing the shadows caused by previous graphene arrays and making interference-free imaging experiments possible.
Brain-computer interface technology has been in development for a long time, but invasive interfaces, which can be harmful to the human body, have always been a safety risk issue.
Since the human brain is very soft, invasive electrodes can easily cause irreversible damage to the brain. To use non-invasive electrodes to obtain signals from the deeper layers of the brain, rigid electrodes need to be implanted into the cerebral cortex tissue.
However, inserting such a steel needle-type implant into the brain is highly susceptible to human rejection. The foreign body encapsulates the electrode, causing the quality of the EEG signal to decay over time. Once the electrode fails, it also needs to be taken back out of the brain tissue, which in turn leaves a gaping hole in the cerebral cortex, triggering inflammation, scarring, and more.
The new neural implant developed by the team, on the other hand, extends the lifespan of the brain-computer interface to some extent. Duygu Kuzum, author of the paper, said that even though the implant sits on the surface of the brain, it is designed to go beyond the limits of physical perception as it can infer neural activity from deeper layers.
The inclusion of machine learning models was also a major innovation of the study. The developers used transparent graphene electrodes to capture surface neural activity at a higher resolution, then used the dataset to train a machine learning model, which in turn allowed the model to predict deeper neural activity from the captured surface signals.
Remarkably, the study also no longer restricts the duration and behavior of the experiment. Typically, scientists need to immobilize subjects under a microscope for up to two hours of observation, but this study no longer requires that. Tests can be performed on freely moving animals for long periods in succession, for hours or days, or even months.
According to Mehrdad Ramezani, the paper’s first author, this could give scientists a more complete picture of neural activity in dynamic and realistic scenarios.
“Electrodes are the first step in brain-computer interface research, and the development of the material is somewhat novel.” An industry stakeholder believes that it can record two signals simultaneously with only minimal invasion, building a correspondence between the surface and deep layers.
Of course, the study still has some uncertainties. The quality of the detected signals is one of the core criteria for judging brain-computer interface research. “However, the conclusion of the study does not mention a leap forward in the precision of detecting electrophysiological signals, which is still at the same level as the current stage of research.” The above source said.
In addition, the thesis research wants to realize the industrial landing, but also depends on the next research results.
“Thesis research from the landing is still a certain cycle, which is determined by the attributes of scientific research and industrial landing attributes. The scientific research attribute requires more to come up with something new, and the industrial landing needs the maturity of the application.” A researcher in a related field said.
At present, the medical scene is one of the largest application markets for brain-computer interfaces, and the outside world expects the technology to help treat neurological diseases such as Alzheimer’s disease and Parkinson’s disease.
The specific way, that is, brain-computer interface through the brain to read and stimulate, restore brain function or communication between the affected areas, and thus reduce mobility or control the progress of disease development.
Dai Xiaochuan, founder of Heze Technology, believes that this will be a brand new method of e-therapy. “The physiological function of the brain is closely related to the propagation of electrical signals in the neuronal network, and when there is an abnormality in electrophysiological activity or damage to the neural network, we can utilize brain-computer interfaces to go for targeted regulation and compensation, thus making it possible to be used to treat a wide range of brain diseases.”
This vision of curing rare diseases has naturally attracted capital and market attention. The global market for brain-computer interfaces reached $2.13 billion in 2022 and is expected to reach $9.44 billion by 2032, according to research firm Precedence Research.
According to Dr. Jia Yan, director of China Renaissance’s medical and life science and technology division, brain-computer interfaces have reaped a high premium in the capital market as a track with a lot of imagination. However, due to its long cycle and high risk, market-oriented investors, whose main goal is financial return, will be more conservative in their investments.
It is worth noting that with the continuous maturation of brain-computer interface technology, there are two technology routes, invasive and non-invasive, and the cycle of realizing industrial landing is also different.
The above researchers said they believe that it will take at least ten years to realize the industrial landing of invasive, and non-invasive, such as sleep devices and other products, has now found a suitable landing scene.
Due to the invasive type, there is a certain safety risk, not only need to pass the rigorous and lengthy clinical trial stage, but also need to obtain the official agency’s medical certification. “There is no one good or bad in the two tracks, but the invasive technology cycle has not arrived, and if you want to launch a product in two or three years, you can only do non-invasive.” The researcher emphasized.
In the commercial world, brain-computer interfaces to realize the landing, data privacy, division of responsibility, and other ethical and moral concerns are unavoidable. Before these issues are completely resolved, brain-computer interfaces still have a long way to go.
