ニューロコミュニケーターにおける性能向上の試み

－感性を考慮したブレインマシーンインターフェースの開発－

抄録

In recent years, neurotechnology, the methods of utilizing brain activations based on the fruit of neuroscience, has advanced. Especially, brain-machine/computer interface (BMI/BCI) to provide a direct link between the brain and external devices has been greatly developed. We have developed neurocommunicator system for people with serious motor disabilities to communicate what they think to others using BMI technology. Neurocommunicator is a practical BMI system that interprets the intention in real time based on EEG with a small device and enables the users to convey 512 kinds of message (8*8*8). This system combines the following three core technologies: small wireless EEG amplifier, high-speed and high-accuracy BMI algorithm and user interface for efficient communication. In this study, we compared the performance and fatigue in 4 kinds of methods: two kinds of characters (Japanese character (HIRAGANA) and Chinese characters (LOLO)) and two kinds of pictures (a human face and a house). We used one kind of stimulus in each method (e.g. HIRAGANA:_??__??__??__??_). As a result, our proposed method HIRAGANA, which is simple and familiar to us, is suggested to be the most efficient and fatigue by using the system is not significantly different among 4 methods.