Brain computer interface based wheelchair for disable people using electroencephalography signal
Brain computer interface causes direct operation between brain and computer. Interfacing of the EEG signal produced by the brain with any control or communication device produces unidirectional communicating channel. Among the non-invasive techniques for probing human brain dynamics, EEG provides a...
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Zes Rokman Resources
2018
|
| Subjects: | |
| Online Access: | http://irep.iium.edu.my/60571/ http://irep.iium.edu.my/60571/ http://irep.iium.edu.my/60571/13/60571-BRAIN%20COMPUTER%20INTERFACE%20BASED%20WHEELCHAIR.pdf |
| Summary: | Brain computer interface causes direct operation between brain and computer. Interfacing of the EEG signal produced by the brain with any control or communication device produces unidirectional communicating channel. Among the non-invasive techniques for probing human brain dynamics, EEG provides a direct measurement of cortical activity i.e., intention of a human being; with millisecond temporal resolution. However, the well-off interconnectivity between the various cortical areas may allow for events in one area to be preceded or accompanied by detectable patterns in other unrelated areas. To develop a practical BCI system, three components should be considered. These are i) to establish an appropriate multivariate signal processing technique to extract multiclass features from multi-channel EEG signals, ii) to look up suitable pattern classification technique to improve the performance of BCI and finally iii) to develop an approprite interfacing circuit to control a user device. Due to poor classification acuracy, practical BCI system has not been fully materialised yet. However, an advanced and simple classification algorithm for motor imagery related BCI system has already been developed with Mahalanobis Discriminant Analysis (MDA) technique. It obtains 93% of kappa accuracy in evaluation phase, which is validated and acceptable, whereas the accuracy with others is maxmimum 86%. Moreover, the developed technique needs a very low computational requirement that makes it suitable for real-time BCI based system to control a wheelchair for the disabled people. To have a fruitful result, the next phase of hardware realization research and interfacing with users are essential which is highly desired factor in a practical/commercial BCI system development. |
|---|