Manipulator operation by using brain-wave signals

Abstract

When we look at a comfortable scene or feel relaxed, our brain waves generally exhibit wave signals in the frequency band of approximately 8 to 13 Hz. These waves particularly exhibit a 1/f fluctuation in which the corresponding power is inversely proportional to frequency f. 1/f -wave signals obtained from test subjects listening to relaxing music was inputted to a robot manipulator. To evaluate the resulting motions, test subjects were asked to complete questionnaires while they watched two types of manipulator motions: a 1/f motion and a white-noise-like motion. The results indicated that 90% of the subjects felt comfortable while watching the 1/f manipulator motions.

SiC for Brain–Machine Interface (BMI)

“…First, it is purely noninvasive and does not require surgery, with the exception of subdermally implanted EEG electrodes, and therefore can be removed when it is not required [4]. The device can extract many different types of potentials, and the positioning of the electrodes for collection of the potentials is well established [54,55]. However there are many drawbacks to the use of EEG interfaced BMI devices.…”

“…Real-time execution of actions is further complicated by the extraction paradigms which limit the execution of commands to only a few times per minute instead of the hundreds of times per minute of natural neuronal signaling [4]. Finally, the noninvasive neural interface recipient must learn to generate an arbitrary signal and relate it to the appropriate command for the BMI device to function correctly and without error [55]. The generation of this arbitrary brain signal requires a great deal of conscious effort from the patient which generally manifests itself as a long training period where both emotional and external environmental situations can interrupt the patients' concentration leading to errors in signal interpretation by the systems on the BMI device [55].…”

“…Finally, the noninvasive neural interface recipient must learn to generate an arbitrary signal and relate it to the appropriate command for the BMI device to function correctly and without error [55]. The generation of this arbitrary brain signal requires a great deal of conscious effort from the patient which generally manifests itself as a long training period where both emotional and external environmental situations can interrupt the patients' concentration leading to errors in signal interpretation by the systems on the BMI device [55]. Even with all of these problems, the EEG neural interfaced BMI devices have provided many therapeutic solutions for people who suffer from total paralysis, like ALS, and its use has developed many invaluable methods for extracting information from neural activity.…”