Graduate Institute of Automation Technology, National Taipei University of Technology, Taipei 10608, Taiwan.
Department of Mechatronics Control, Industrial Technology Research Institute, Hsinchu 310401, Taiwan.
Biosensors (Basel). 2022 Sep 20;12(10):772. doi: 10.3390/bios12100772.
Most people with motor disabilities use a joystick to control an electric wheelchair. However, those who suffer from multiple sclerosis or amyotrophic lateral sclerosis may require other methods to control an electric wheelchair. This study implements an electroencephalography (EEG)-based brain-computer interface (BCI) system and a steady-state visual evoked potential (SSVEP) to manipulate an electric wheelchair. While operating the human-machine interface, three types of SSVEP scenarios involving a real-time virtual stimulus are displayed on a monitor or mixed reality (MR) goggles to produce the EEG signals. Canonical correlation analysis (CCA) is used to classify the EEG signals into the corresponding class of command and the information transfer rate (ITR) is used to determine the effect. The experimental results show that the proposed SSVEP stimulus generates the EEG signals because of the high classification accuracy of CCA. This is used to control an electric wheelchair along a specific path. Simultaneous localization and mapping (SLAM) is the mapping method that is available in the robotic operating software (ROS) platform that is used for the wheelchair system for this study.
大多数行动障碍者使用操纵杆来控制电动轮椅。然而，患有多发性硬化症或肌萎缩性侧索硬化症的患者可能需要其他方法来控制电动轮椅。本研究实施了一种基于脑电图（EEG）的脑机接口（BCI）系统和稳态视觉诱发电位（SSVEP）来操纵电动轮椅。在操作人机接口时，监视器或混合现实（MR）护目镜上会显示三种涉及实时虚拟刺激的 SSVEP 场景，以产生 EEG 信号。典型相关分析（CCA）用于将 EEG 信号分类为相应的命令类，并使用信息传递率（ITR）来确定效果。实验结果表明，所提出的 SSVEP 刺激会产生 EEG 信号，因为 CCA 的分类精度很高。这可用于控制电动轮椅沿特定路径行驶。本研究中的轮椅系统使用机器人操作系统（ROS）平台中的同时定位与地图构建（SLAM）作为映射方法。
