Hisaya Tanaka

The development of a brain computer interface device for amyotrophic lateral sclerosis patients

The objective of this research was to develop a brain computer interface (BCI) communication device for amyotrophic lateral sclerosis (ALS) patients. The device was designed to meet the needs of ALS patients, and to be used at a clinical level. Initial tests were performed by ALS patients, and the result was accounted for in the experimental production of the communication device. Lastly, the device was evaluated by able-bodied examinees and ALS patients. For able-bodied examinees, the device scored a high rate of correct sessions. When an ALS patient was the user, the correction rate was not as well, but it would have scored highly if a correct parameter was chosen.

ERD analysis method in motor imagery brain–computer interfaces for accurate switch input

Motor imagery brain–computer interface (MIBCI) can control computers using MI. However, input accuracy is low, partly owing to individual variability in event-related desynchronization (ERD) detection among different subjects. In an earlier study, we determined that using a max power in the mu-band method, i.e., the peak trace method (PTM), is effective for ERD detection. In this study, we compare the PTM to the band power method to determine the most effective method for ERD detection during MI tasks. Experimental results indicate that we could detect ERD using the PTM; however, MI-state estimation was difficult. We also found that the PTM might be effective for ERD detection in subjects with MI experience.

Time-frequency analysis of grasp motion and associated motor imagery for event-related desynchronization

In a brain-computer interface, using event-related desynchronization and synchronization(ERD and ERS), intuitive command input is available. Switching precision was thought to be improved by a detailed examination of the ERD μ rhythm in the time-frequency domain. In the experiment, measured electroencephalograms of motor cortex in the resting state, hand grasp motion, and hand grasp recall were analyzed. Consequently, desynchronization was observed in the band centered at 12 Hz for both the grasp motion and grasp recall.

Study of Japanese Sign Language Semantic Process based on N400 Analysis

We study efficient automatic generation and communication of the Japanese sign language image. Therefore, it is necessary to study language processing mechanism of the sign language of the deaf people. This study describes the method for examining semantic analysis process of sign language using N400 which is a component of event related potentials. From two experiments, N400 amplitude was high in a specific semantic processing. Moreover, it was higher in the character presentation than in the sign language presentation. It is considerated that the the acceptance or the set are different according to type of language or modality though the same content is presented.

Cognitive Function Evaluation of Dementia Patients Using P300 Speller

The character-input BCI has been studying as a communication aid for neuromuscular illness patients such as ALS. We focused on input-error in the character-input BCI. It is necessary that the subject has to concentrate high attention to input correct characters in the character-input BCI. On the other hand, it causes an input-error of far character when the subject is distractedness. Hence, we considered that the number of input error of far character may increase when the dementia patient uses the character-input BCI. Therefore, we conducted a character gazing experiment using the character-input BCI to the elderly. Then, we quantified the result of the character estimation of gazed character by calculating the SEDV (Spelling-Error Distance Value). As a result, in the Modarate1 group, the SEDV was 0.41 characters farther than the Questionably significant group. Hence, there is possibility that the character-input BCI may be able to screening dementia.

Analysis of Physical Feature in the Course Turn While Walking

In this study, we analyzed the physical rotary quantity of the yaw axis while subjects walk on running-machine, because it was considered that this rotary quantity may be available for discrimination of the right and left walk intention on the running-machine. Four subjects were attached a gyro sensor of Enliven3D to a head, a shoulder, a waist and a foot. Then, they walked two kilometers per hour on running-machine. When subjects walked on the running-machine, we requested them to intend changing direction and going straight with looking straight. The shoulder, waist and foot in all subjects turned about 20° under the influence of turn-intention in the changing direction. There was individual difference of about 10° in the rotary quantity between every subject and each part. Each part without heads turned under the effect of turn intention, even if we requested subjects to look at the right and the left.

Visualization of the Muscle Tension in Stand-Up and Sit-Down Motion Using MoCap and EMG

In this study, visualization of the muscle tension of the Vastus lateralis which is the main muscle for standing up and sitting was performed by motion capture and electromyography. It was aimed to confirm the effect of training, with this visualization. First, a human skeletal muscle model was generated from motion capture measurement data based on kinematics theory. Next, the muscle tension model of the Vastus lateralis of the standing and sitting motion was generated from the electromyography data based on the inverse kinetic theory. We used nMotion-Musculous, for estimation of muscle tension. As a result, the muscle tension of the Vastus lateralis was presumed to be 600 to 800 N in both cases of standing up and sitting. Furthermore, a muscle tension of about 100 N was estimated even in a state of maintaining the standing posture. These muscle tension estimates were visualized by CG animation and feed back to the subjects.

Analysis method for ERD in mu-rhythm detection in motor imagery brain-computer interface

Motor imagery brain-computer interfaces (MIBCI) use hand or foot MI to control computers. However, MIBCI control accuracy is low. Previously, we determined that using max power in the mu band method, i.e., the peak trace method (PTM), improves event-related desynchronization (ERD) detection accuracy. Control accuracy may be improved by improving ERD detection accuracy in an MIBCI. In this study, we compare the PTM to the band power method to determine the most effective method for ERD detection during MI tasks. Overall, experimental results indicate that we could not detect ERD more accurately using the PTM. However, the PTM may be effective depending on the mu rhythm occurrence pattern.

Electrodes arrangement on brain-computer interface for the ALS's posture

A brain-computer interface (BCI) can operate a computer by analyzing electroencephalogram data. A BCI does not require body motion; therefore, the physically handicapped, such as those who suffer from amyotrophic lateral sclerosis (ALS), can also operate a computer. However, with ALS patients, attaching electrodes requires significant time and effort. In this study, we examine the number and arrangement of electrodes for a character input BCI in consideration of the limitations encountered with ALS patients. We found that the percentage of correct answers was greater with a smaller number of electrodes than the current number of electrodes. Moreover, we consider that the Cz and CP4 electrode positions are effective for P300 discrimination.

Physiological and psychological evaluations in low- and high-frequency noise using near-infrared spectroscopy

This study aims to evaluate physiological and psychological states using near infrared spectroscopy in noise environments with low or high frequencies. Our system assumes that noise affects brain activity in the frontal lobe. In order to evaluate the subject’s states in a noise environment, we constructed an experimental system that measures the subject’s states. The experimental method adopted here was borrowed from our previous studies. In the present study, we collected experimental data about the subject’s unpleasant or pleasant experiences by producing a noise environment with low and high frequencies. We conclude that noises with low or high frequencies affect our psychological states as well as brain activity in the frontal lobe.