Wavelet Method for Motor Activity Pattern Analysis on Experimental Data from Human Multichannel Electroencephalography for External Device Control
Abstract
Introduction: The modern fundamental and technical science has a great interest in neural interface development, as it aims at both improving life quality and studying the functioning of a human brain. Neural interface devices are based on recognizing the patterns of recorded brain activity, and their main problems are low efficiency of pattern recognition techniques and long time necessary to train the operators. Purpose: Developing methods for recognizing patterns associated with various motor activity of untrained operators. Results: Based on wavelet technologies, new methods are proposed for studying various patterns of brain activity which correlate with various types of motor activity. Morle basis is modified in order to speed up the calculation. Approaches to assessing energy capacity of various electroencephalography processes are discussed, based on the calculation of skeleton characteristics. We give examples of processing experimental data obtained from untrained volunteers. Characteristic features of patterns for various motor activity (imaginary or real, leg or hand movement) have been discovered. We demonstrate that the electroencephalography patterns are universal for everyone from the volunteering group. Practical relevance: The revealed features of "motor" patterns and the methods of electroencephalography processing can be used in constructing neural interfaces which decode brain electrical activity almost without operator's training.Published
2018-02-01
How to Cite
Runnova, A., Maksimenko, V., Pchelintseva, S., Kulanin, R., & Hramov, A. (2018). Wavelet Method for Motor Activity Pattern Analysis on Experimental Data from Human Multichannel Electroencephalography for External Device Control. Information and Control Systems, (1), 106-115. https://doi.org/10.15217/issn1684-8853.2018.1.106
Issue
Section
Control in medical and biological systems