Communication via Brain Computer Interface.

Q. Theoretically, can a BCI be used as a translation device?

A. Yes, a BCI could allow a person to spell words using thoughts, which could then be translated into other languages. While brain-based spelling is a reality today, unfortunately, it is rather slow. Thought to text translation is typically at the rate of less then ten characters per minute. Thus, it seems to me, that practical application of BCI technology in the area of translation would require a dramatic increase in this communication rate. It is currently not clear what the upper limit for this communication may be, or how long it will take to achieve this goal.




Q. Please briefly explain how a BCI works, and how thought signals are translated into text.

A. Brain-Computer Interfaces extract specific features from brain signals that reflect the intent of the subject. These features are then translated into device commands that can drive a variety of outputs. Thus, to use a BCI, a subject has to learn how to change particular brain signals in a particular way, which is similar to other types of learning, such as how to play tennis.




Q. Why are BCIs sometimes referred to as thought translators?

A. In popular media, BCIs are sometimes referred to as thought-translation devices. However, this notion is misleading since all BCIs require the active participation of the subject. In other words, a BCI is not a mind-reading device that passively listens in to what the person is "thinking," but rather a communication device that actively engages with the subject.




Q. Please briefly explain how BCIs are being used for communication now, and what web sites do you recommend to learn more on the subject.

A. Most BCIs to date have remained laboratory demonstrations. However, initial efforts are beginning to apply this first generation of BCI systems to the needs of the disabled, e.g., http://braincommunication.org, which uses the BCI2000 system.




Q. Once the thought to text ratio reaches greater than five hundred characters per minute, would a BCI translator then be practical to use for communication?

A. This is too far in the future to contemplate at present.




Q. Please briefly explain the difference between invasive and non - invasive BCIs and the major advantages and dis-advantages of each use.

A. Non-invasive BCIs use electroencephalographic (EEG) activity recorded from the scalp. This recording method is convenient, safe, and inexpensive. EEG electrodes are placed on the scalp and conductive gel is applied between each electrode and the scalp. EEG measures the concerted activity of many millions of brain cells. EEG-based BCIs have been shown to support higher performance than often assumed, including accurate two-dimensional movement control without [18] and with [19] selection capabilities. At the same time, BCI systems based on EEG typically require substantial training to achieve accurate 1D or 2D device control (about 20 and 50 30-min training sessions, respectively). Furthermore, EEG recordings are susceptible to artifacts created by other sources, such as electromyographic (EMG) signals produced by muscle contractions, environmental noise, etc. These problems have thus far impeded the widespread clinical application of this safe and relatively inexpensive technology.

Microelectrode recordings from within the brain have also been used as a basis for BCI systems. These systems use firing rates of individual or multiple neurons, or the overall neuronal activity of multiple neurons recorded within the brain. Thus, these signals have very high resolution. However, the stability of recordings from electrodes implanted within the brain is currently uncertain, because electrodes are subject to different tissue responses. Furthermore, effective operation of these BCI systems depends of substantial initial and continual expert supervision. Thus, despite recent encouraging evidence that BCI technologies based on microelectrodes may be useful to people with paralysis, the difficulties described above currently impede widespread clinical implementation of such invasive BCI technologies.

A final sensor method uses electrodes that are placed on the surface of the brain (electrocorticography (ECoG)). These sensors measure local field potential activity that captures brain signal changes on the order of millimeters (rather than a tenth of a millimeter captured by an implanted microelectrode). ECoG electrodes are in direct contact with the brain, and thus do not need the application of any conductive gel. This methodology could be a powerful but yet practical alternative EEG and microelectrodes. ECoG has higher fidelity than EEG and far less vulnerability to artifacts such as EMG. At the same time, because ECoG electrodes do not penetrate the brain, they are likely to have greater long-term stability compared to implanted microelectrodes.

Q. How do you see BCI technology evolving twenty years into the future?

A. The biggest impediment of BCI technology at present is the lack of a sensor modality that provides safe, accurate, and robust access to brain signals. It is conceivable or even likely that such a sensor will be developed within the next twenty years. The use of such a sensor should greatly expand the range of communication functions that can be provided using a BCI.




Q. Please briefly explain to us about BCI2000 and how this software works?

A. Development and implementation of a Brain-Computer Interface (BCI) system is complex and time consuming. In response to this problem, we have been developing a general-purpose system for BCI research, called BCI2000. BCI2000 has been in development since 2000 in a project led by the Brain-Computer Interface R&D Program at the Wadsworth Center of the New York State Department of Health in Albany, New York, USA.

BCI2000 has already had a substantial impact on BCI research. As of early 2009, BCI2000 has been acquired by more than 350 laboratories around the world. It has been the basis for some of the most impressive BCI studies reported to date. BCI2000 has also received significant attention by scientific and popular media. For example, the initial article on the BCI2000 system (Schalk et al., 2004) has already been cited more than 130 times. The system has also been: referenced several hundred times, including in journal articles, media articles, and personal blogs; used or cited in dozens of Masters Theses or Doctoral Dissertations; mentioned as desirable experience in job postings; and listed as qualification in Curriculum Vitae. This large and increasing success of the adoption of the BCI2000 system provides strong evidence for the substantial demand for and utility of the software.

In summary, BCI2000 is strongly stimulating progress in the field of BCI research. BCI2000 is thus fast becoming, or perhaps has already become, the standard software platform for research in this area.

The BCI2000 system is readily available to research and educational institutions. However, to operate it effectively still requires substantial expertise in a number of BCI-related areas.

Q. Is it possible for non invasive BCIs to interface with brains at a distance? If so, what is the range?

A. No technology at present allows BCI operation from a distance.




Q. Once BCIs can upload information to the brain, do you think this method would be practical for learning new languages?

A. This is too far in the future to contemplate at present.




Q. Please tell us a brief anecdote on the most amazing thing you have seen a BCI accomplish for a human

A. A BCI2000-based system is currently in use by a man who is totally paralyzed by Amyotrophic Lateral Sclerosis. He has used this system for about two years to communicate with his environment. His story was recently featured on 60 Minutes with Scott Pelley.




Q. Have you ever interfaced a non invasive BCI to your brain, if so, how did it feel, or how do others explain the feeling?

A. Most people say that BCI use becomes relatively intuitive with appropriate training.




Q. To close, please tell us more about your influences and role models, and how you became interested, and began your work in this field of study?

A. At present, there are very few educational curricula on brain-computer interfacing. Thus, like most people in the field, I have been educated in a different area, i.e., electrical engineering and computer science. Furthermore, when I began working in this area, BCI research barely existed as its own field. Thus, I was influenced mostly by my immediate supervisor rather than by scientists at other institutions.

 

 

E-mail: info@americantesol.com