NEW YORK (Reuters Health) Jul 12 - A paralyzed patient's brain activity can be harnessed through a brain-computer link to direct movements and restore lost motor function, according to two pilot studies reported in the July 13th issue of Nature. "What is currently available to paralyzed patients is an eye tracker mounted on the headband or wheelchair, which optically looks at the eye, determine where it is pointing, and uses that to control cursor on screen," Dr. Krishna V. Shenoy, co-author of one of the papers, said in an interview with Reuters Health. "But this modality is bulky and commandeers eye movement." "Our goal is to couple natural intention to move with tools that will provide interaction with the patient's environment," continued Dr. Shenoy of the Stanford University School of Medicine in California. He believes the neuromotor systems described in the Nature papers will become widespread in clinical practice, like cochlear implants for hearing loss and deep brain stimulation for Parkinson's disease. "Our research has been invigorated by the microelectronics industry that gave us the ability to build complicated electronic systems that move fast," he added, "as well as advancements in the field of neuroscience." In another paper by Dr. John P. Donoghue, from Brown University, Providence, Rhode Island, and members of his group implanted a sensor comprised of an array of silicon microelectrodes onto the region of the motor cortex that controls arm and hand movement. The patient was a tetraplegic 25-year-old man whose cervical spinal cord had been transected 3 years earlier. This "neuromotor prosthesis" included the implanted sensor, a decoder to translate firing neuron patterns into motor commands and a computer gateway (BrainGate; Cyberkinetics, Inc.). The patient was instructed to imagine arm motions, during which the internal sensor detected brain cell activity that was converted into computerized signals, which permitted the patient's hand and arm to move. Even 3 years after the injury, the researchers note, "motor cortex neurons can still be actively engaged and encode task-related information during the intention to move the limb ordinarily controlled by that motor cortex." So far, the patient has been able to use the control signal to direct computer software, play video games, control a TV and to operate a robot arm and an electric prosthetic hand. Dr. Shenoy said his group is "now addressing improvements in performance of the brain-computer interface," using normal unimpaired rhesus monkeys as a proxy for a paralyzed human. Co-author Dr. Gopal Santhanam described their experiments, in which they trained the monkeys not to move their arm until they receive a "go" cue. The monkeys were implanted with a microarray into the dorsal premotor cortex. When presented with a peripheral target on a computer screen, "the animals stayed still without moving their arm, then reached out to the target on cue. During this process the investigators recorded neural activity that predicted the intended location movements before they were made." "Our next goal is to chain different targets together, to allow multiple movements at the same time, and to do so at increased speeds," Dr. Santhanam said. He believes that despite the increased surgical risk and the danger of infection due to wires leading from the brain to the computer "implanted electrodes provide better assessment of individual neurons in the brain, which we believe will lead to higher performance. Simply using scalp electrodes requires much more training, takes longer to decode the endpoint and delays restoration of independence to patients." He predicted the development wireless telemetry so that the implant can be completely covered, peripheral nerve stimulation of paralyzed muscles, better movement algorithms, and improved feedback from the arrays. Dr. Shenoy added that stem cells may eventually be used for many of the same purposes addressed by the neuromotor prostheses. The two research teams plan to focus on soldiers returning from the Iraq war with amputated limbs or paralysis. "If we fit them with a robotic arm or an esthetic prosthesis, they will need to control their movements, and stem cells will not be of much help there." "What drives us is the patient, wanting to develop instruments to help them interact with the environment and provide independence," Dr. Santhanam said. "Clinicians and patients should stay tuned because a lot of promising things are on the horizon to help patients."
Nature 2006;442:164-171,195-198.

 A group of German researchers at the University of Tübingen's Institute of Medical Psychology and Behavioural Neurobiology, led by Niels Birbaumer, similarly focuses on non-invasive methods. They use self-regulation of slow cortical potentials (SCPs)—shifts of cortical voltage lasting from a few hundred milliseconds to several seconds—for their CBI, called the Thought Translation Device (Hinterberger et al, 2003). Since 1996, ten patients have learned to produce SCPs, enabling them to move a cursor on a computer screen. SCPs do not correspond to movement or feelings, but to the general state of brain activity, and patients learn by reinforcement—using electrically negative and positive potentials—to control the device. On mastering this, they can choose letters from the screen to form sentences, although this is a slow process. The German group teamed up with Wolpaw in 2000 to create a universal platform on which to test existing and new technologies that would allow individuals to choose the signals that work best for them (Schalk et al, 2004). Called the BCI2000, the device can also detect P300 signals, or 'event-related potentials'; these are sharp voltage increases that peak about 300 ms after the brain registers a surprising occurrence, which were discovered by Emanuel Donchin of the University of South Florida (Tampa, FL, USA). These signals are interesting for CBI developers because they allow them to recognize thoughts in a certain category without having to train patients to regulate their brain activity. Despite these advances, John Donoghue believes that non-invasive systems have a disadvantage in that one must be trained to use them. In addition, their signals are more diffuse and inexact because they do not come from the actual neural substrate. Other groups, including scientists led by Ranu Jung, co-director of Arizona State University's Center for Rehabilitation Neuroscience and Rehabilitation Engineering (Tempe, AZ, USA), are therefore exploring ways to bypass the 'thinking' aspect in order to control a neuroprosthetic device in the same way the brain normally controls movement—at the subconscious level. A few years ago, “juggling in my mind” was a mere joke, but now researchers have shown that thinking—with the help of computers—can be translated into action. Ultimately, the type of CBI used—which brain waves it uses, whether implanted or not—may depend on the individual patients. It may be hard to imagine what CBIs will enable patients to accomplish in a few years from now, but it is ultimately the mind that will enable them to escape bodily limitations.

References:

Hinterberger T, Kubler A, Kaiser J, Neumann N, Birbaumer N (2003) A brain–computer interface (BCI) for the locked-in: comparison of different EEG classifications for the thought translation device. Clin Neurophysiol 114: 416–425 [PubMed]. Nicolelis MA, Chapin JK (2002) Controlling robots with the mind. Sci Am 287: 46–53 [PubMed]. Patil PG, Carmena JM, Nicolelis MAL, Turner DA (2004) Ensemble recordings of human subcortical neurons as a source of motor control signals for a brain–machine interface. Neurosurgery 55: 27–38 [PubMed]. Schalk G, McFarland DJ, Hinterberger T, Birbaumer N, Wolpaw JR (2004) BCI2000: a general-purpose brain–computer interface (BCI) system. IEEE Trans Biomed Eng 51: 1034–1043 [PubMed]. Serruya MD, Hatsopoulos NG, Paninski L, Fellows MR, Donoghue JP (2002) Brain–machine interface: Instant neural control of a movement signal. Nature 416: 141–142 [PubMed]. Wolpaw JR, McFarland DJ (2004) Control of a two-dimensional movement signal by a non-invasive brain–computer interface in humans. Proc Natl Acad Sci USA 101: 17849–17854 [PubMed].