A brain computer interface (BCI) is a device that enables people with motor or cognitive limitations to control computers and other devices. BCIs make use of different combinations of sensors, recording methods using signal processing algorithms, as well as machine learning to translate brain signals into operation commands. They can be used to control a diverse variety of devices, from cursors on computer screens to robotic arms and wheelchairs.
Most BCIs measure activity from the scalp, such as electroencephalographic (EEG) or functional near-infrared imaging (fNIRS) signaling. The signals are recorded by a sensor which detects the activity of neurons. Software converts them into operating commands.
In many BCI systems users have to undergo a process of instruction how to generate signals that the system recognizes. For example for example, a BCI designed to type letters requires the user to imagine that they are moving their left or right hand to generate the desired image of movement.
The most exciting BCI developments involve implantable, or noninvasive systems that record directly from neural tissue, not the scalp. These devices are more precise than noninvasive BCIs, but they also require surgical intervention and carry some risks.
BCIs that are invasive are still experimental and patients should be aware of the risks and benefits. Security and privacy are also major issues, since BCIs detect neural signals that may contain sensitive information about health or behavior. Some people are not a fan of the technology because they fear that hackers might hack their minds and control them.