Brain-Computer Interfaces: Beyond Medical Applications

By Jan B.F. van Erp, Fabien Lotte, and Michael Tangermann

NOTE: This is an overview of the entire article, which appeared in the April 2012 issue of Computer magazine.
Click here to read the entire article.

Brain-computer interface (BCI) technology is a potentially powerful communication and control option in the interaction between users and systems. At present, most BCI applications focus on assistive care, providing an alternative communication medium for those who cannot use a keyboard or mouse, but applications have the potential to include any task that would benefit from interaction beyond the keyboard.

In the past few years, BCI applications have broken out of laboratories and hospitals to include nonmedical applications, such as gaming. But nonmedical BCIs are still relatively embryonic, and interest is growing in identifying the breakthroughs that will enable this technology’s broad introduction in nonmedical applications – any implementation for healthy users that would benefit from brain-computer interaction. In addition to transferring knowledge from medical BCI applications, researchers also must deal with a range of issues that are more critical outside assistive care.

The article addresses challenges facing BCI systems in non-medical applications (many of which are also present in life sciences applications outside the area of assistive care. BCI designers must address a range of usability, integration, standardization, and ethical issues to enable the growth of BCI applications.

Figure 1. Two BCI products in use. (1) The user is wearing electrodes while running IntendiX. (2) The user is wearing a band of dry EEG sensors developed by Emotiv.

Hardware and open source software currently in use in BCI applications are reviewed. The authors conclude that out of a number of measurement technologies, EEG is the most usable, although it is not ideal. Five open source software tools for processing brain signals are listed as noteworthy.

Seven application areas, as outlined below, are explored in the article.

• Device control
  Research on BCIs to assist users lacking full limb development has matured to the point that such users are already benefiting, even though the devices offer limited speed, accuracy, and efficiency.

  Nonmedical device control is more problematic. Users with full muscular control cannot benefit as easily because a BCI lacks the bandwidth and accuracy to compete with a standard input device, such as a mouse or keyboard. Introducing a shared control scheme would enable the user to give high-level, open-loop commands while the device takes care of low-level control.

  Additional control channels or hands-free control could benefit users such as drivers, divers, and astronauts, who must keep their hands on controls to operate equipment. Brain-based control paradigms could supplement other forms of hands-free control, such as a voice command or eye movement.

• User-state monitoring
  Future interfaces must be able to understand and anticipate the user’s state and intentions. Automobiles could alert sleepy drivers, or virtual humans could convince users to stick to their diet.

  BCIs might also be useful in neuroscientific research. Because they can monitor the acting brain in real time and in the real world, BCIs could help scientists understand the role of functional networks during behavioral tasks.

• Evaluation
  Evaluation applications can be either online or offline. The former continuously provide evaluations, in real or near real time; the latter provide evaluations only once, after the experimental study is finished. Neuroergonomics and neuromarketing are two application subareas.
• Training and education
  Most training aspects relate to the brain and its plasticity – the brain’s ability to change, grow, and remap itself. Measuring plasticity can help improve training methods and individual training regimens.
• Gaming and entertainment
  Over the past few years, companies such as Neurosky, Emotiv, Uncle Milton, MindGames, and Mattel have released numerous products. Most developers are convinced that BCIs will enrich the gaming and entertainment experience in games tailored to the user’s affective state – immersion, flow, frustration, surprise, and so on.
• Cognitive improvement
  A common nonmedical application involving a BCI is neurofeedback training, in which operant conditioning alters brain activity to improve attention, working memory, and executive functions.

  The line between medical and nonmedical neurofeedback applications is likely to be thin, but a nonmedical application might be the optimized presentation of learning content.

• Safety and security
  Safety and security EEG alone or combined EEG and eye movement data from expert observers could support the detection of deviant behavior and suspicious objects. Also, image inspection might be faster than is possible with current methods.

The authors address several categories of technological challenges that need to be met if wider use of BCIs are to find wide acceptance in non-medical (or perhaps more broadly, non-assistive-care) applications. One key area is usability. Users in these areas don’t want to have to undergo extensive training or calibration sesssions, and won’t accept using gelled electrodes on their scalp. In the long term, they suggest, alternative sensors to EEG electrodes should be developed.

Software must be robust to environmental noise and nonstationarity. It will be essential to reduce calibration times. And new algorithms to perform user-state monitoring will be required.

Finally, progress will be needed in integrating BCI equipment into existing systems. (Think ‘plug-and-play’). They suggest that hardware and software standardization will be key in this area.

The prospects are bright for rapid growth of BCI in non-medical areas. This is especially true of gaming, with its large economic impact. The authors propose measures that can assist the realization of BCI’s potential. One observation is that the coordination of medical and nonmedical BCI research efforts is vital. Such alignment could produce a shared roadmap and research agenda that would benefit both areas.

ABOUT THE AUTHORS

Jan B.F. van Erp is a senior scientist and program manager in the Department of Perceptual and Cognitive Systems at The Netherlands Organization for Applied Scientific Research (TNO) and is the scientific director of the BrainGain consortium. His research focuses on advanced human-computer interaction including multimodal interaction, augmented reality, and brain-computer interfaces. Van Erp received a PhD in computer science from Utrecht University, the Netherlands. Contact him at jan.vanerp@tno.nl.

Fabien Lotte, a research scientist at INRIA Bordeaux Sud-Ouest, France, was a research fellow at the Institute for Infocomm Research, Singapore, when he started work on this article. His research interests include BCIs, signal processing, pattern recognition, and virtual reality. Lotte received a PhD in computer science from the National Institute for Applied Sciences, Rennes, France. He is a member of the International Association for Pattern Recognition. Contact him at fabien.lotte@inria.fr.

Michael Tangermann is a postdoctoral researcher in the machine learning department at the Berlin Institute of Technology (Technische Universität Berlin) and is a member of the Berlin Brain-Computer Interface group. His research interests include brain-computer interfaces, machine learning, and auditory attention processes. Tangermann received a Dr. rer. nat. in computer science from Tübingen University, Germany. Contact him at michael.tangermann@tu-berlin.de.