Neuralink Envisions $5,000 Brain-Computer Interface to Treat Most Neurological Disorders

Oct 31, 2024 07:31 CST Updated 07:31

Neuralink

Brain-Computer Interface System Developer

Amino Watch - Original Production by the Digital Therapeutics Group

Author | Wu Yue

“I believe that the vast majority of diseases or brain conditions can be addressed with the Neuralink device, with costs expected to drop to around $5,000, similar to the cost of laser eye surgery.”

This is Elon Musk’s latest statement on brain-computer interfaces at the 2024 Annual Congress of Neurological Surgeons.

Elon Musk has always been a controversial figure, with each of his innovations and even remarks sparking intense debate. Some revere him as a savior of the tech industry, while others criticize him as a reckless dreamer. Nevertheless, it is undeniable that his vision and execution have repeatedly pushed the boundaries of technology.

If Musk’s vision of affordable brain implants becomes a reality, it could revolutionize the brain-computer interface market and make this technology accessible to more people, especially those with disabilities.

Of course, helping paralyzed patients regain mobility is merely the initial objective; Musk’s ultimate goal is to achieve a high-bandwidth connection between the brain and computers, establishing a symbiotic relationship between “human intelligence and digital intelligence” to mitigate the civilizational risks posed by AI.

So, how far has Neuralink actually progressed?

/ 01 /

"Telepathy"

Brain-computer interfaces are a common scene in science fiction works.

Real-World Brain-Computer Interfaces: One end connects to your brain, extracting your thoughts from the dense and complex electrical signals of neurons, while the other end links to external devices such as computers or machinery. This technology directly converts thoughts into control signals and executes commands, bypassing the body entirely. It establishes direct information exchange between the central nervous system and external devices without relying on the peripheral nerves and muscular systems of the limbs.

In 2021, Elon Musk unveiled an experimental monkey that used mind control to type the sentence “I want to eat snacks” on a computer. This demonstration was powered by brain-computer interface technology.

Cochlear implants, which help individuals with disabilities regain hearing, represent the most successful and widely clinically applied brain-computer interface technology to date. By converting sound signals into electrical signals and transmitting them directly to the brain, this technology enables a large number of deaf individuals to restore their ability to hear and communicate.

Characters in movies such as Avatar or Iron Man, who flexibly control their limb movements through brain-computer interface technology, are not yet achievable at present.

In recent years, brain-computer interfaces (BCIs) have made significant progress following years of development by research teams both in China and abroad. Currently, BCI technologies are categorized into three major pathways: invasive, semi-invasive, and non-invasive. Among these, the invasive approach carries the highest risk factor.

Neuralink employs an invasive approach that requires craniotomy to implant a chip into the human cerebral cortex. The chip reads and analyzes your brainwaves, then transmits these signals out.

According to Neuralink’s official website, the “N1” implant is comparable in size to five stacked coins. Its construction comprises a biocompatible enclosure, a battery, chips and electronics, and 64 threads. The N1’s built-in miniature battery is equipped with a compact inductive charger, supporting external wireless charging. Furthermore, the 64 threads of the N1 are distributed with 1,024 electrodes capable of recording neural activity and are highly flexible.

It is worth noting that, according to Neuralink’s plan, the implantation surgery is not performed by human surgeons but is instead carried out by the “R1” robot. During the procedure, a section of the patient’s skull is removed and replaced with the brain–computer interface. Once the implantation is complete, the brain–computer interface device reads and analyzes brain activity, wirelessly transmitting the information to a nearby laptop or tablet.

In early 2022, the FDA rejected the company’s clinical application based on safety and other considerations. The FDA expressed concern that the device’s thin wires carrying electrodes might migrate to other regions of the brain; it also raised questions about whether the device could be removed without damaging brain tissue, as well as a series of other safety issues, including the potential for the device to overheat and cause brain tissue damage.

Previously, BrainGate, a company developing invasive brain-computer interfaces, encountered instances where electrodes failed within the brain. The cause was encapsulation of the electrodes by glial cells. Consequently, safety considerations mandated by the U.S. Food and Drug Administration (FDA) represent an unavoidable reality for the industry.

Now that Neuralink has successfully obtained approval for human clinical trials and completed its first patient implant, it may indicate that more evidence has been gathered regarding safety and other aspects.

Of course, it is also possible that Musk made concessions. Previously, the FDA believed that conducting slower, phased trials would be more suitable for Neuralink, meaning fewer initial human subjects would receive implants, with additional testing conducted several months later.

Elon Musk was dissatisfied with this recommendation, as it could delay the progress toward final FDA approval.

However, the amyotrophic lateral sclerosis (ALS) patient recruitment program currently posted on Neuralink’s official website indicates that Neuralink requires participants to be at least 22 years of age, have impaired function in all four limbs with no improvement for over one year, and have at least one family member or friend available to provide care. The entire trial will last six years, including an 18-month baseline study period.

Neuralink stated that it would perform 11 surgeries in 2024, 27 in 2025, and 79 in 2026, with the number projected to rise to 22,000 by 2030.

This also aligns with the FDA’s recommendation for a slower, phased approach.

Neuralink’s first brain-computer interface product is named “Telepathy.” Musk also shared his thoughts on the product’s prospects: “(In the future, humans) will only need to think to control a mobile phone or computer, and through them, control almost any device. Imagine if Stephen Hawking could communicate faster than a fast typist or an auctioneer; that is our goal.”

Compared to the so-called transformation of humans into cyborgs, this goal is clearly more pragmatic and represents a crucial step forward.

/ 02 /

# Ushering in a New Era of Treatment

Elon Musk has always been a controversial figure, with each of his innovations sparking intense debate. Some revere him as a savior of the tech industry, while others criticize him as a reckless dreamer. Nevertheless, it is undeniable that his vision and execution have repeatedly pushed the boundaries of technology.

As John Donoghue, a brain-computer interface expert at Brown University in the United States, stated, “Elon Musk has completed substantial preliminary work building on the research of many scientists, including the work our team has been conducting since the early 21st century.”

Since 1969, scientists have been striving to advance experiments on external interfaces with the nervous system, exploring the mysteries of the human brain. Brain-computer interfaces (BCIs) create communication pathways for information exchange between the human brain and external devices. What benefits can this bring us?

The applications are limited only by imagination. At the very least, in the field of medicine, it holds many promising and remarkable uses.

Many human diseases, such as epilepsy and Parkinson’s disease, as well as quadriplegia resulting from spinal cord injury, are caused by the brain’s inability to connect with the peripheral nervous system. Brain-computer interfaces hold promise for improving these conditions.

Current treatments, including medications and surgical techniques, offer insufficient therapeutic efficacy for patients with central nervous system damage caused by conditions such as stroke and amyotrophic lateral sclerosis (ALS). These patients often remain in a state of long-term paralysis or functional impairment, resulting in poor quality of life.

Brain-computer interface (BCI) technology has transcended the traditional scope of biological tissue repair, achieving functional substitution through human-machine interaction. This advancement holds revolutionary significance and has already yielded breakthroughs in specific case studies. At the opening ceremony of the 2014 FIFA World Cup in Brazil, a young man with high-level paraplegia, equipped with mechanical assistance and utilizing a brain-computer interface, kicked the first ball of the tournament. This demonstration showcased the most intuitive clinical value of BCIs to the world: providing assistive tools for motor impairments.

Companies, including Neuralink, are proving this point. Research on brain-computer interfaces in China is also proceeding with great vigor.

In 2020, a team from Zhejiang University enabled a person with quadriplegia to precisely control an external robotic arm and hand, allowing for the accurate execution of basic actions such as eating and shaking hands. In addition to advancements in motor-assistive therapy, brain-computer interface (BCI) technology has also made remarkable strides in the field of speech restoration.

In June 2023, Huashan Hospital affiliated to Fudan University led the release of a neural network model tailored for Chinese. This model can separately decode lexical tones and basic syllables from intracranial recordings and synthesize them into speech, thereby enabling patients with tone-language articulation disorders or aphasia to directly express their “inner voice” via brain-computer interface technology.

It can be said that the application of brain-computer interfaces (BCIs) originated in medicine. By utilizing BCI technology to acquire and analyze information from the aforementioned functional areas of the brain, a new era of treatment can be ushered in for numerous patients.

Although brain-computer interfaces (BCIs) still have a long road to traverse from clinical development to final market approval, with numerous challenges regarding the safety, stability, and reliability of devices and surgical procedures (invasive) needing resolution, there is an urgent and vast demand for their medical application. This is driven by the significant need among stroke-induced paralysis patients and individuals with Parkinson’s disease to restore function through BCI technology.

According to McKinsey’s estimates, the potential market size for global medical applications of brain-computer interfaces (BCIs) is expected to reach $40–145 billion between 2030 and 2040. Within this, the potential market for serious medical applications is estimated at $15–85 billion, while that for consumer healthcare applications is projected at $25–60 billion.

Currently, exploration in the consumer healthcare sector is equally vibrant. For instance, a Canadian brain-computer interface company has launched an EEG headband that helps users enhance their meditation experience through real-time audio feedback.

In China, companies have already launched products such as brain-controlled sleep aids, brain-controlled mice, and meditation brain-state monitors to address insomnia and depression. For instance, Rouling Technology has released Airdream, a miniaturized EEG monitoring device that, in addition to assessing sleep quality via EEG signals, incorporates a home-use EEG-based sleep management system to help users reduce stress, relax, and fall asleep more quickly. Hangzhou Enter Tech has also introduced an EEG-enabled smart eye mask.

Overall, current brain-computer interface devices can only act on the cerebral cortex, which simplifies many of the issues involved in in-depth research. The cerebral cortex primarily processes a variety of direct signals, such as sensory information directly derived from motor intentions, as well as auditory and visual inputs.

Theoretically, this could resolve many issues such as blindness, paralysis, and hearing loss, simply by correctly connecting the device to the cerebral cortex.

According to Musk, Neuralink’s electrodes are currently implanted approximately 3–4 mm into the brain, sensing signals from multiple layers within the cortex. Beneath the cerebral cortex lie deeper brain structures, such as the hypothalamus. These deeper brain systems are the ultimate targets for achieving bidirectional information exchange and could also be used to treat conditions such as depression, addiction, and anxiety.

This still requires substantial research investment and output, with a long road ahead.

/ 03 /

Time-Dependent Response

If the ultimate goal of AI is to achieve the pinnacle of inorganic intelligent life, then brain-computer interfaces aim to break down the information boundary between organic and inorganic intelligence, enabling their integration.

Regarding brain-computer interfaces, countless individuals have proposed application scenarios, such as facilitating communication and regulating emotions. In the short term, they are overly overestimated; in the long term, they are overly underestimated.

Elon Musk once stated that he aims to restore sight to the blind and mobility to people with disabilities, and has even discussed the integration of humans and artificial intelligence. This has propelled brain-computer interfaces into the mainstream spotlight.

But in fact, “making the blind see, the paralyzed move, and the deaf hear again” has been a cliché circulating for 25 years. And countless scientists are working toward this ultimate goal.

The current challenge is that restoring sensory input (such as vision) involves electrical stimulation in the brain, which is fundamentally different from merely recording single-neuron activity. Currently, there is no evidence to suggest that existing neural implant devices can create a sensory system in any way.

In other words, as an emerging field of research, brain-computer interface (BCI) technology is still in its early stages of development. It encompasses multiple disciplines, including computer science, neuroscience, cognitive psychology, biomedical engineering, mathematics, signal processing, clinical medicine, and automatic control, with numerous challenges yet to be addressed.

For example, how to handle the vast and complex number of neurons.

Brain-computer interfaces can be categorized into various types to provide diverse functionalities. However, all scientists researching brain-computer interfaces are striving to address two fundamental questions: How to accurately decode information from the brain? And how to precisely encode information into the brain?

These two processes are constantly occurring naturally in your brain. As you read this sentence, your eyes are performing a series of specific horizontal movements. This is because neurons in the brain output information to a machine (your eyes), which receives commands and responds accordingly.

Input and output information is the work of brain neurons. What brain-computer interfaces need to do is intervene in this process.

This does not sound difficult. However, the total volume of the cerebral cortex is approximately 500,000 cubic millimeters, containing roughly 20 billion neuronal cell bodies within this space. On average, each cubic millimeter of cortex contains about 40,000 neurons. Yet, the neuronal cell body constitutes only a small fraction of the neuron's overall structure.

In addition, the brain contains glial cells in numbers comparable to neurons, as well as blood vessels. The total length of capillaries per cubic millimeter of cortex can reach one meter.

For brain-computer interface (BCI) technical engineers to achieve highly precise capture or feedback of brain signals, they must detect signals emitted by specific neuronal cell bodies within this one-cubic-millimeter region, or stimulate certain specific cell bodies to generate the signals required by the engineers. The considerable difficulty involved is thus evident.

Compared to non-invasive approaches, invasive brain-computer interfaces can better capture neuronal signals but require cables to transmit large volumes of data.

Furthermore, greater engineering challenges include cost control, specifically whether costs can be reduced through optimized processes and technologies to achieve commercialization.

Another example is Moore’s Law for brain-computer interfaces.

Statistical data indicate that, given the current rate at which brain-computer interface technology doubles the number of simultaneously recordable neurons every 7.4 years on average, it will take until 2100 to achieve simultaneous recording of one million neurons, and until 2225 to record all neurons in the human brain.

Therefore, how brain-computer interfaces address the bandwidth issue has become a key focus of academic research and industrial breakthroughs. Donoghue stated that Neuralink’s proposed solution to the massive data volume is more practical than anticipated.

Rather than pursuing high-bandwidth, high-speed data acquisition methods, Neuralink employs Bluetooth. This approach yields a lower volume of information; due to bandwidth limitations, it cannot isolate the activity of every individual neuron. Nevertheless, it is effective, as this method is sufficient to achieve the goal of controlling computers via neural activity.

In addition, improving the accuracy of signal recognition and refining signal processing methods to make them systematic and universal are also issues that need to be addressed.

Certainly, these technical challenges are being overcome one by one by scientists and engineers.

The journey of a technology from its initial invention to final application is long and arduous, involving not only technical challenges but also considerations of humanities, ethics, and philosophy. After all, humanities without science are ignorant, while science without humanities is dangerous.

When human brain activity can be accurately decoded, it means that the rich privacy data within the brain may potentially be leaked or stolen. With the development of brain-computer interface technology, in the future, there will undoubtedly be a need to provide sufficiently secure measures to protect users' private data security.

“For me, the brain is a safe haven for thought, imagination, and freedom of dissent. Without any safeguards, we are on the verge of crossing the final frontier of privacy,” said Nita Farahany, Professor of Neuroethics at Duke University in the United States.

/ 04 /

Summary

Much like the narrative in the film *Oppenheimer*, the first detonation of an atomic bomb carried two potential outcomes: either a massive explosion or the destruction of the Earth. At the time, the answer was unknown; even today, top scientists cannot accurately predict the future of brain-computer interfaces. After all, many developments are unforeseeable.

The engineering challenge in brain-computer interfaces lies in whether it is possible to develop low-power information transmission methods with smaller form factors and higher bandwidth.

The answer in this regard appears to be affirmative. However, there are other technical challenges that remain difficult to resolve at present, and no one can know what the final outcome will be until the results of pivotal trials are available.

But this is precisely the essence of science: you never know what lies around the corner.

PharmDive Media Business Cooperation

Media Relations | News &Press Release for the Conference

Manager Zhang: 18600036371 (WeChat ID same as phone number)

Click “Yaodu” below to follow for more exciting content

Disclaimer

This article reposted by the “Yao Du” official WeChat account originates from other WeChat public account platforms, with the primary purpose of sharing industry-related knowledge and disseminating the latest information. The copyrights of images and articles belong to the original authors. In case of any infringement, please notify us promptly, and we will remove the relevant content within 24 hours.

The push notification rules of WeChat Official Accounts have changed yet again. If you don’t click“Watching”Or not set to"Starred", we might just fade into the vast sea of literature~ Click here, and don't miss out on the latest news from Pharmadule! ??