Tamura Hiroki, Yan Mingmin, Sakurai Keiko, Tanno Koichi
Department of Environmental Robotics, University of Miyazaki, Miyazaki 889-2192, Japan.
Organization for Promotion of "Center of Community" Program, University of Miyazaki, Miyazaki 889-2192, Japan.
Comput Intell Neurosci. 2016;2016:7354082. doi: 10.1155/2016/7354082. Epub 2016 Jun 21.
The aim of this study is to present electrooculogram (EOG) and surface electromyogram (sEMG) signals that can be used as a human-computer interface. Establishing an efficient alternative channel for communication without overt speech and hand movements is important for increasing the quality of life for patients suffering from amyotrophic lateral sclerosis, muscular dystrophy, or other illnesses. In this paper, we propose an EOG-sEMG human-computer interface system for communication using both cross-channels and parallel lines channels on the face with the same electrodes. This system could record EOG and sEMG signals as "dual-modality" for pattern recognition simultaneously. Although as much as 4 patterns could be recognized, dealing with the state of the patients, we only choose two classes (left and right motion) of EOG and two classes (left blink and right blink) of sEMG which are easily to be realized for simulation and monitoring task. From the simulation results, our system achieved four-pattern classification with an accuracy of 95.1%.
本研究的目的是呈现可作为人机接口使用的眼电图（EOG）和表面肌电图（sEMG）信号。为患有肌萎缩侧索硬化症、肌肉萎缩症或其他疾病的患者建立一种无需明显言语和手部动作的高效替代通信渠道，对于提高他们的生活质量至关重要。在本文中，我们提出了一种用于通信的EOG-sEMG人机接口系统，该系统使用面部相同电极上的交叉通道和平行线通道。该系统可以将EOG和sEMG信号作为“双模态”进行记录，以便同时进行模式识别。尽管可以识别多达4种模式，但考虑到患者的状态，我们仅选择了两类EOG（向左和向右运动）和两类sEMG（左眨眼和右眨眼），这对于模拟和监测任务来说易于实现。从模拟结果来看，我们的系统实现了四种模式的分类，准确率为95.1%。
Zotero 插件
