What is a brain-computer interface (BCI)?

Brain-computer in­ter­faces enable direct com­mu­ni­ca­tion between the brain and a computer. They work by capturing neural signals, in­ter­pret­ing them and trans­form­ing them into commands that computers or machines can un­der­stand.

What is meant by a brain-computer interface?

A brain-computer interface (BCI) is an interface between the human brain and a computer that allows direct in­for­ma­tion transfer between humans and machines. This neu­rotech­nol­o­gy fa­cil­i­tates the con­nec­tion without engaging the pe­riph­er­al nervous system, which means that it operates in­de­pen­dent­ly of speech and movement.

Brain-computer in­ter­faces are based on the un­der­stand­ing that merely imagining an action is enough to produce a mea­sur­able change in the brain’s elec­tri­cal activity. For example, imagining moving a finger already triggers a response in the motor cortex, which plans and initiates voluntary movements. Through a training process, the brain-computer interface learns which brain ac­tiv­i­ties cor­re­spond to specific thoughts or mental commands. This allows the extracted brain signals to be used as neu­rotech­no­log­i­cal input systems. However, due to numerous tech­no­log­i­cal chal­lenges, BCI de­vel­op­ment involves sig­nif­i­cant time and financial in­vest­ments.

How does a brain-computer interface work?

Brain-computer in­ter­faces capture and analyze brain activity to convert it into control commands for computers. The mea­sure­ment of elec­tri­cal brain activity is done using elec­trodes. Sub­se­quent­ly, spe­cial­ized al­go­rithms process the captured signals to recognize patterns that correlate with specific thoughts and mental images. In the next step, the brain-computer interface trans­lates these patterns into commands that machines can un­der­stand. Re­searchers use machine learning and ar­ti­fi­cial in­tel­li­gence to recognize and analyze signals due to the com­plex­i­ty of the data.

Non-invasive vs. invasive brain-computer in­ter­faces

Brain activity can be recorded either using manually applied and removable BCIs or through sur­gi­cal­ly implanted BCIs:

• Non-invasive brain-computer in­ter­faces capture brain activity using elec­troen­cephalog­ra­phy (EEG). This method measures voltage fluc­tu­a­tions on the surface of the scalp through elec­trodes placed on the head. These BCIs typically involve a cap fitted with sensors. Al­ter­na­tive­ly, mag­ne­toen­cephalog­ra­phy (MEG) is used to record magnetic brain activity, producing a three-di­men­sion­al image of various areas.
• Invasive brain-computer in­ter­faces use elec­trodes implanted directly into the brain to measure elec­tri­cal impulses via EEG. This ob­ser­va­tion method offers the highest signal res­o­lu­tion but currently carries the risk of medical com­pli­ca­tions such as neural damage. There are also semi-invasive methods where elec­trodes are placed on the cerebral cortex, which is con­sid­ered less risky.

What is the current state of brain-computer interface de­vel­op­ment?

Due to intensive research efforts, the quality of the extracted brain signals is con­tin­u­ous­ly improving. This is es­pe­cial­ly true for implanted BCI systems, which have a high trans­mis­sion rate and are in­creas­ing­ly the focus of sci­en­tif­ic research and studies. Non-invasive BCIs, however, offer limited accuracy because the skull filters the signals. Although the first invasive BCI system was implanted in a human as early as 1998, the high com­plex­i­ty of the procedure has resulted in very few BCI implants since then, with only about 50 worldwide in over 25 years.

Due to recent funding ini­tia­tives for fun­da­men­tal neu­ro­log­i­cal research in the U.S. (BRAIN Ini­tia­tive) and Europe (Human Brain Project), sig­nif­i­cant ad­vance­ments in BCI tech­nolo­gies are expected in the coming years. Research teams are currently working on bidi­rec­tion­al in­ter­faces, capable of trans­mit­ting signals from external sources into the brain. Ad­di­tion­al­ly, con­tin­u­ous progress is being made in in­ter­pret­ing brain activity, owing primarily to modern an­a­lyt­i­cal methods like neural networks, big data, and deep learning, which can process large data sets ef­fi­cient­ly.

Who is working on BCI tech­nolo­gies?

As of 2024, numerous gov­ern­men­tal in­sti­tu­tions, uni­ver­si­ties and private companies are con­duct­ing research on BCI tech­nolo­gies. In 2020, re­searchers at Zhejiang Uni­ver­si­ty (China) implanted a brain-computer interface in a para­plegic patient, enabling him to control robotic arms and operate devices with his thoughts. In January 2024, a team from the Mass­a­chu­setts Institute of Tech­nol­o­gy (MIT) in­tro­duced a non-invasive brain-computer interface that can control Boston Dynamics’ robot dog.

The United States and China are currently leaders in BCI im­plan­ta­tion, while German research focuses on non-invasive BCIs due to the lower as­so­ci­at­ed risks. Notable BCI companies include:

• Neuralink develops invasive Brain-Computer In­ter­faces. The U.S.-based company’s BCI implant contains over 1,000 elec­trodes attached to hair-thin wires. Besides better treatment for severe brain disorders, Neuralink aims to even­tu­al­ly enhance mental ca­pa­bil­i­ties.
• Blackrock Neurotech is based in Utah and has been active in the brain-computer interface sector since 2008. Blackrock devices are among the most widely used BCI implants, primarily aimed at in­creas­ing the in­de­pen­dence of people with severe paralysis.
• BrainGate in­tro­duced the first-ever human-im­plantable BCI chip in 2004 and is con­sid­ered a pioneer in the field. The latest implants consist of two or more units with up to a hundred elec­trodes each, placed on the cerebral cortex.
• Synchron has developed a minimally invasive BCI that is not implanted directly into the brain but instead sits in the blood vessels in the head. The im­plan­ta­tion is done using a tiny metal stent, through which the BCI is in­tro­duced into the head.

Current and future ap­pli­ca­tions of BCI tech­nolo­gies

To date, BCIs’ most critical ap­pli­ca­tion is sup­port­ing people with sig­nif­i­cant physical dis­abil­i­ties. BCIs are already being used to assist in­di­vid­u­als with dis­abil­i­ties or specific con­di­tions like locked-in syndrome (LiS) in their mobility, com­mu­ni­ca­tion and in­de­pen­dence. Medically used BCIs, for example, enable patients to move a robotic arm, com­mu­ni­cate using a spelling machine or control devices through thought. However, medical BCI ap­pli­ca­tions are still in the prototype phase. In the en­ter­tain­ment and wellness sector, some products are already ready for the market. For example, non-invasive BCI headsets, which reduce stress using biofeed­back systems, are already available.

Various other scenarios are con­ceiv­able in the future. Brain-computer in­ter­faces may drive the de­vel­op­ment of neu­ro­pros­thet­ics, allowing users to ex­pe­ri­ence sensation or connect with robots to complete complex tasks. Bidi­rec­tion­al BCIs could make it possible to com­mu­ni­cate brain-to-brain, upload thoughts to cloud servers and connect directly to the internet. Whether BCI tech­nolo­gies will establish them­selves in the long term depends not only on tech­no­log­i­cal progress but also on societal ac­cep­tance.

What are the op­por­tu­ni­ties and risks as­so­ci­at­ed with brain-computer in­ter­faces?

Brain-computer in­ter­faces have the potential to lead to dis­rup­tive changes in various societal areas, not only in medicine but also when used for op­ti­miza­tion in areas like work, school and daily life as well as in fields such as virtual reality. The­o­ret­i­cal­ly, BCIs could activate skills and ca­pac­i­ties pre­vi­ous­ly unimag­in­able—like learning a language by directly down­load­ing it to your brain. However, several tech­no­log­i­cal chal­lenges remain to be addressed before this is possible.

Despite these ad­van­tages, brain-computer in­ter­faces also pose sig­nif­i­cant risks. Reading brain activity enables the analysis of highly sensitive personal data. Critics warn that BCIs could be misused to influence in­di­vid­u­als’ thoughts and behavior. Ad­di­tion­al­ly, BCIs are still tech­ni­cal­ly immature and error-prone, which can result in un­de­sir­able con­se­quences. To ensure user safety, ethical, legal and social im­pli­ca­tions must be carefully con­sid­ered.