Clustering of EEG Signal to Optimize Number of Electrodes in BCI Applications
(0) With the advent of low-noise,highly sensitive probes,and being able to analyze the multiple signals in real-time facilitated by increased memory and computational power available,recording of the electrical signals on our scalp is getting reliable,and cheap.This leads to a lot of attention to the research and development of Brain-Computer Interface(BCI)applications,not only for medical applications(like epileptic seizures,monitoring anesthesia or brain function etc.)but also for entertainment purpose.Yet,it is not widely used for several reasons-cost,time required for correct placement of dozens of probes on the head,cleaning the scalp,applying saline solutions or gels to make a clean contact etc.We do need significant computational power to real-time process multiple signals received from the probes.It so happens that for a target application,signals from a few probes would suffice.If we can reduce the number of probes,it will lead to cost reduction,as well as lighter computation.One way to reduce the number of probes,without losing any information,is to cluster the signals received from the probes and use only one from each cluster,the one nearest to the cluster center.In this work,we investigated signal pre-processing techniques as well as clustering algorithms,and evaluated the results after reducing the number of probes.Results show that the accuracy is not hampered even when the number of probes are reduced from 21 to 4.The application was to detect Event Related Potential ERP300.
Electroencephalogram (EEG Brain Computer Interface (BCI) Pulse Code Modulation (PCM) Fuzzy c-means Wards method
Hikaru Yokoha Goutam Chakraborty Daigo Kikuchi
Gard.School of Software & Information Science.Iwate Prefectural University,Iwate,Japan Dept.of Software & Information Science Iwate Prefectural University,Iwate,Japan Gard.School of Software & Information Science Iwate Prefectural University,Iwate,Japan
国际会议
The 2014 ICME International Conference on Complex Medical Engineering (CME2014)ICME复合医学工程国际会议
台北
英文
278-283,141
2014-06-26（万方平台首次上网日期，不代表论文的发表时间）