Originally published in Spanish in El País
What is a Brain-Computer Interface?
Brain-computer interfaces (BCIs) or brain-machine interfaces (BMIs) are technologies that establish direct communication between the human brain and an external machine, usually a computer or an electronic circuit. These devices can be used both in basic and clinical research, or for personal consumption. BCIs record neuronal activity directly or indirectly, and can be electrical, optical, magnetic, and even acoustic. There are two types of BCIs: invasive and non-invasive. Invasive devices require neurosurgery and are inserted into the brain and communicate with the exterior via cables or wirelessly. Non-invasive devices, on the other hand, do not require surgery and are placed on top of the skull, as if they were a kind of cap or headband.
Why are BCIs important?
BCIs allow external access to brain activity and enable the brain’s modification. Since brain circuits generate mental activity and human behavior, BCIs can allow the decipherment of cognitive activity and its selective manipulation. Experiments in laboratory animals over the last decade have shown the possibility of altering and manipulating sensory perception, memory, and behavior. Although BCIs were initially designed to assist neurologically-impaired patients that suffered from blindness, paralysis or other disabilities, non-invasive BCIs could replace iPhones, since, after all, smartphones serve simply to connect the user to the Internet. With BCIs, the bandwidth of that connection would be much greater, and the connection would be immediate, without having to use your fingers or your eyes.
BCI in the Academic world
DARPA, the research agency of the United States Department of Defense, is the main driver of the development of BCIs within the BRAIN initiative. In 2017, a team from Columbia University received a $15.8 million grant from DARPA to manufacture a two-centimeter-square flexible ultra-thin silicon (CMOS) chip, with one million neural recording electrodes and 100,000 neuronal stimulation electrodes. This wireless chip is designed as a prosthetic for the blind, directly connecting their visual cortex to a camera. But, in principle, it could be implanted in any area of the cerebral cortex to connect it to computers and machines and in people who are not necessarily neurologically-impaired. This chip is currently being tested in monkeys to be used in human patients later on. DARPA is also funding the University of California, which received $21.6 million to develop a microscope capable of allowing us to observe 1 million neurons while simultaneously stimulating 1,000 of them with maximum accuracy. Brown University received $19 million from DARPA to create "neurograins," tiny, wireless devices capable of interacting with individual neurons.
BCIs in the Industrial World
Elon Musk’s Neuralink focuses on creating invasive BMIs. First, Neuralink has already raised $158 million in funds and the company has developed a neurosurgical robot that can insert electrodes into the brain with extreme precision without damaging vasculature. Additionally, Neuralink has created a 4x5mm implantable chip packed with electrodes that aims to both “read” and “write” brain data. In other words, the chip would theoretically be able to collect and decode brain data, while also stimulating the brain in specific ways to control behavior. In February 2020, Musk declared that Neuralink plans to soon test its technology by implanting four of its chips on a human.
Kernel, a neuroscience startup founded and led by Silicon Valley multimillionaire Bryan Johnson launched last May a new technology called: “Neuroscience as a Service (NaaS)” to provide on-demand non-invasive brain recording capacities for scientific research organizations. NaaS consists of a software and a headset that weighs less than 1.5 Kg and is covered in 48 modules. Kernel’s aim is to provide good software that doesn’t need to be connected to heavy equipment in laboratory settings.
Iota, founded by UC Berkeley researchers Jose Carmena and Michel Maharbiz, has raised $15 million to develop “neural dust,” a technology that consists of implantable, microscopic, wireless chips the size of a grain of sand that collect data from specific areas of the peripheral nervous system. Once placed in a specific desired muscle fiber, the neural dust can be activated using a beam of ultrasound, which then runs between the electrodes, and interacts with the electrical activity of the tissue, thus reflecting back a slightly different wave. A reader can then interpret these wave changes, converting them into precise monitoring data.
Facebook is developing the project Thought-to-test, a non-invasive a Brain Computer Interface (BCI) that would allow consumers to type directly from their brain. The aim of this new technology would be to take consumers’ conscious thoughts and display them directly on a screen, without having to narrate them to Siri. In March 30th 2020, Facebook announced that a group of researchers they sponsor at UCSF had successfully developed a brain-to-texting algorithm that can decode brain data at unprecedented speech and with minimal error rates.
CTRL-Labs, launched by creator of Microsoft Internet Explorer Thomas Reardon, is developing a bracelet that connects via a cable to a computer with a digital avatar. he bracelet is lined with chips that can sense the body’s electrical neurons. hen a person wearing the bracelet moves his or her arm, the electrical impulses travel down the wire, and are translated into the computer. Once the electrical impulse has been receivedt, the digital avatar on the screen is able to mirror the movement of the bracelet-wearing-person. CTRL-labs aims to continue developing this technology so that the bracelet doesn’t just sense neuronal activity that results in movement, but also senses and interprets when a person is simply thinking of a specific movement. If this is successful, the avatar on the screen would move according to the person’s thoughts moving. In 2019, Facebook acquired CTRL-Labs, providing them with somewhere between $500 million and $1 billion to continue developing the digital avatar, according to people acquainted with the deal.
Translated by Clara Baselga-Garriga and Paloma Rodríguez Páramo