Dr. Gerwin Schalk: In theory, the technique can monitor any part of the brain. The areas most commonly monitored for communication and controlling devices are the areas responsible for controlling motor function. They reflect whether or not you move a hand or even in what direction you are moving the hand. The basic idea is to first identify the signals that relate to those motor functions. For example, take the brain signals associated with the event of moving your hand, or when you are not moving you hand, or even when you simply imagine moving your hand, or when you are not thinking of moving your hand. Once you know that signal you can start associating that signal with a particular output. In a very simple case you would be associating imagining moving your hand with the answer "yes," and another signal with the answer "no." This allows you to answer simple yes or no questions just by imagining hand movements. Obviously there are more complicated versions of this.
GS: I think it's because those areas of the brain are already set up for controlling things and for interacting with the environment, for outputting something which in this case is a motor command. Other areas of the brain are more set up for handling input, like the part of the brain that processes visual input. An area like that would not be particularly well suited to controlling something, because it would be particularly easy for signals there to be modified just by turning on the light. You want an area that can execute a particular function no matter what you're seeing, hearing, or smelling. That's why the motor system is a pretty good area to control devices—because it controls its environment in normal behavior.
ET: Is there a qualitative difference in the brain activity that occurs when you imagine moving a limb and when you actually do it?
GS: There 's been a pretty big debate about that. Recent research has shown conclusively that the exact same areas of the brain are active when you imagine moving your hand as compared to when you actually move it. That's not the case with speech, for example. When you imagine speaking you only mentally rehearse it but you don't necessarily imagine how to move your face and your mouth in order to produce that speech. What you would see is activity in auditory areas of the brain as you mentally rehearse that speech, but you will not see activity in the motor areas of the brain involved in facial movement. This indicates that when you imagine speaking you are not mentally simulating how to move your face and mouth. In contrast, when you imagine walking your motor system is simulating how to move you arms and legs without actually doing it. However there is usually a difference in signal strength between imagining limb movement and doing it, with the latter being noticeably stronger. Paradoxically, a recent study that I participated in demonstrated that once you give actual feedback to those brain signals associated with movement imagery, for example using them to control a cursor on the screen, then the brain signals for imagined movement can become stronger than those for actual movement. What that means is not entirely clear, but it does show that if you train the brain you can get it to produce stronger activity for imagined movement than it would for actual movement.
ET: That's a powerful change in the normal pattern of brain activity. Have any of your subjects reported substantial changes in their dreams after working with your interface?
GS: Not that I know of. The tasks that we ask them to do are pretty similar to their normal daily behavior. You're not going to dream differently just by imagining moving your hand a few times a day. Of course, everything we do changes our future behavior. A promising line of research involves rehabilitating the brain after a stroke. The idea is to try and rehabilitate the brain in a manner similar to how you rehabilitate muscles after an injury. You try to train the brain to producer stronger activity where it is needed to control muscle movement and restore normal function. At least that's the theory.
ET: What would the surgery be like for getting an implanted BCI, and how risky is the procedure?
GS: The surgery involves opening up the skull and placing a sheet of electrodes on the surface of the brain. In the future a wireless transmitter would be added, making the implant virtually undetectable to anyone, but the patient and the procedure could almost be done on an outpatient basis—not completely, but nearly so. The procedure would be less invasive than a breast augmentation. It wouldn't be very risky at all.
No comments:
Post a Comment