可攜式Wi-Fi無線傳輸64通道高解析度 腦波量測系統開發與驗證

Abstract

腦機介面技術(Brain-Computer Interface, BCI)是一種溝通介於大腦與外在世界的工具，此技術已被廣泛研究。然而傳統收錄高品質腦波所使用的標準化量測儀器往往笨重且巨大，對腦波收錄與實際應用造成不便。近年來伴隨著高度系統晶片(System on Chip, SOC)整合以及感測器的發展，無線可攜式裝置常見於腦機介面研究當中，但仍有不足的地方。隨著神經科學研究發展，腦電訊號(EEG)分析需更多且精確的數據資料，由於藍芽無線傳輸技術頻寬有限，使用16通道以上腦波量測系統時需降低取樣率，否則將造成資料遺失。因此本研究目的在發展多通道、高解析度(24位元)、高取樣率、Wi-Fi傳輸EEG訊號擷取裝置。期望此系統能使腦波技術在神經科學研究具有更多貢獻。 本論文提出的腦波量測系統是以Cortex-M4內建Wi-Fi子系統的微控器為核心，致力於多通道腦波接收的設計，且同時提升訊號品質。本系統可使用傳統濕電極、貼片電極，亦可以使用新式乾式頂針電極。腦波資料則以Wi-Fi傳輸至Labview使用者介面並紀錄以便於腦波離線分析，而在認知實驗中常需要的事件傳輸則以RS232來進行事件紀錄。本系統經過SSVEP、ERP等驗證實驗的測試，顯示系統能實際運用於腦神經科學研究。

The brain-computer interface (BCI) technique is a method of communication between the brain and the external world. This technology had been extensively studied. However, the BCI instruments with high measured signal quality are always heavy and huge. Thus, recording EEG tends to be an inconvenience task. In recent years, a high level of system chip (System on Chip, SOC) integration and development of sensors commonly used in wireless portable devices in BCI research, but there are still not enough. With the development of neuroscience, EEG signal analysis needs more accurate data. Due to the limited bandwidth of Bluetooth wireless transmission technology, the use of more than 16 channel EEG measurement system need to reduce the sampling rate, otherwise it will cause data loss. Therefore, this proposed study aimed at developing a multi-channel, high-resolution (24 bits), high sampling rate, Wi-Fi transmission EEG brain-computer interface devices. Hope this system can make contributions in neuroscience research. The EEG acquisition system proposed in this thesis is based on Cortex-M4 microcontroller with Wi-Fi subsystem. Committing to the multi-channel design, and also enhance the signal quality. This system can be used with wet sensor, Ag/AgCl electrode or dry sensors. The user-interface (Labview) receives EEG data via Wi-Fi transmission and also saves raw data for off-line analysis. On the other hand, RS-232 would be used for sending event’s tag in some cognitive experiments. The system had been validated by SSVEP, ERP experiments, etc. Results of these experiments demonstrate that this system can be used well for EEG measurements and is feasible for practical applications.