一个基于稳态视觉诱发电位的两步脑拼写系统的设计与实现  

Design and implementation of a two-step mental spelling system based on steady-state visual evoked potentials

在线阅读下载全文

作  者:龚华云[1] 魏庆国[1] 李茂全[1] 

机构地区:[1]南昌大学电子信息工程系,江西南昌330031

出  处:《南昌大学学报(理科版)》2015年第6期544-551,共8页Journal of Nanchang University(Natural Science)

基  金:国家自然科学基金项目(61365013);江西省科技厅科技支撑计划项目(20132BBE50050;20151BBE50067);江西省教育厅科技项目(GJJ13054)

摘  要:基于稳态视觉诱发电位(steady-state visual evoked potentials,SSVEP)的脑拼写系统在脑机接口(braincomputer interface,BCI)研究领域引起了极大的关注。然而,现存的基于SSVEP的拼写器的性能需要改进,特别是在输入精度和响应速度两个方面。基于16个发光二极管(LED)的视觉刺激,通过提取频率信息设计和实现了一个两步输入的脑拼写器。这个拼写器允许用户输入31个字符和3个功能命令。每个字符输入需要最多两个目标选择,而3个功能键只需要一个选择。7个受试者参加了一个在线实验,他们都成功地完成了字符、单词和一个句子的输入任务。平均的拼写精度、每分钟输入字符数和信息传输率分别为95.8%、7字符/分钟和27.2比特/分钟。结果表明,与现有的脑拼写器相比,就拼写精度和拼写速度而言这个系统具有高的拼写性能。The mental spelling system based on steady-state visual evoked potentials (SSVEP) has drawn great attentions in the field of brain-computer interface (BCI). However, the performance of existing SS- VEP-based spellers need improving,especially the input accuracy and the response speed. This paper pres- ented a two-step mental spelling system based on sixteen LED visual stimuli for extraction of frequency in- formation. The speller allowed users to input 31 characters and 3 function keys. Each character inputting need at most two target selections, and three function keys need only one selection. Seven subjects partici- pated in an online experiment. All of them succeeded in spelling characters, words and one sentence. The average spelling accuracy,letters per minute and information transfer rate reached 95.8 %, 7.0 letters/min and 27.2 bits/min, respectively. The results showed that the proposed system possessed high performance compared to existing mental spellers in terms of spelling accuracy and spelling speed.

关 键 词:脑机接口 稳态视觉诱发电位 脑拼写系统 LED刺激器 典型相关分析 

分 类 号:TP391.4[自动化与计算机技术—计算机应用技术]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象