基于功能性近红外光谱技术的脑机接口

发布时间：2018-06-08 01:25

本文选题：功能性近红外光谱技术 + 脑机接口　；参考：《上海交通大学学报》2017年12期

【摘要】：为了探究功能性近红外光谱技术(fNIRS)对相同动作的运动想象和运动执行区分可行性以及前额皮层对运动想象和运动执行分类准确率的影响,研究测量了15位受试者手臂伸展和手指敲击的运动想象过程和运动执行过程的前额皮层和运动功能皮层的血氧变化信号.提取均值,斜率,二次项系数和近似熵特征建立基于支持向量机的四分类模型.对应于手臂伸展和手指敲击的四分类模型,分别实现了87.65%和87.58%的分类准确率.相对于单独运动功能皮层区域建立的运动功能皮层-fNIRS-脑机接口,引入前额皮层血氧变化信息能显著提高脑机接口分类性能,且对手指敲击动作的提高效果大于手臂伸展动作.因此,前额皮层区域的血氧响应生理特征能提高fNIRS-脑机接口的分辨性能,同时验证了fNIRS-脑机接口应用于多种肢体动作脑功能活动提取的可行性.
[Abstract]:To explore the feasibility of functional near infrared spectroscopy (FNIRS) in differentiating motion imagination and motion execution for the same motion, and the effect of the prefrontal cortex on the classification accuracy of motion imagination and motor execution, The changes of blood oxygen in the prefrontal cortex and motor functional cortex of 15 subjects during arm extension and finger tapping were measured. Four classification models based on support vector machine (SVM) are established by extracting mean, slope, quadratic coefficients and approximate entropy features. The classification accuracy is 87.65% and 87.58%, respectively, corresponding to the four classification models of arm extension and finger tapping. Compared with the motor functional cortex -fNIRS- brain-computer interface established in the single motor functional cortex region, the classification performance of brain-computer interface can be improved significantly by introducing the information of blood oxygen change in the prefrontal cortex, and the effect of finger tapping is better than that of arm stretching. Therefore, the physiological characteristics of the blood oxygen response in the prefrontal cortex region can improve the resolution of the fNIRS- brain-computer interface, and the feasibility of the application of the fNIRS- brain-computer interface in the extraction of various limb action brain functions is verified.
【作者单位】：中国航天员科研训练中心人因工程国家重点实验室;
【基金】：国家自然科学基金项目(81671861) 中国航天医学工程预先研究项目(YJGF151204) 中国航天员科研训练中心人因国家重点实验室自主课题(SYFD150051805)资助
【分类号】：R318;TN911.7

【相似文献】