《using temporal evidence and fusion of time-frequency features for brain- computer interfaci》由会员分享，可在线阅读，更多相关《using temporal evidence and fusion of time-frequency features for brain- computer interfaci（121页珍藏版）》请在金锄头文库上搜索。
1、i Using Temporal Evidence and Fusion of Time-Frequency Features for Brain- Computer Interfacing by Gireesh S. Dharwarkar A thesis presented to the University of Waterloo in fulfillment of the thesis requirement for the degree of Master of Applied Science in Electrical and Computer Engineering Waterl
2、oo, Ontario, Canada, 2005 Gireesh S. Dharwarkar 2005 ii I hereby declare that I am the sole author of this thesis. This is a true copy of the thesis, including any required final revisions, as accepted by my examiners. I understand that my thesis may be made electronically available to the public. i
3、ii Abstract Brain-computer interfacing (BCI) is a new method of human-machine interaction. It involves the extraction of information from the electroencephalogram (EEG) through signal processing and pattern recognition. The technology has far reaching implications for those with severe physical disa
4、bilities and has the potential to enhance machine interaction for the rest of the population. In this work we investigate time-frequency analysis in motor-imagery BCI. We consider two methods for signal analysis: adaptive autoregressive models (AAR) and wavelet transform (WAV). There are three major
5、 contributions of this research to single- trial analysis in motor-imagery BCI. First, we improve classification of AAR features over a conventional method by applying a temporal evidence accumulation (TEA) framework. Second, we compare the performance of AAR and WAV under the TEA framework for thre
6、e subjects and find that WAV outperforms AAR for two subjects. The subject for whom AAR outperforms WAV has the lowest overall signal-to-noise ratio in their BCI output, an indication that the AAR model is more robust than WAV for noisier signals. Lastly, we find empirical evidence of complimentary
7、information between AAR and WAV and propose a fusion scheme that increases the mutual information between the BCI output and classes. iv Acknowledgements First and foremost I would like to thank my advisor, Dr. Otman Basir, for his wisdom and for believing in this area of research. His drive and ent
8、husiasm in everything he does has been inspirational and an example for me. I am indebted to my beloved fianc, Andrea, who, during challenging times in this research, convinced me I could accomplish anything; she has been my rock. I am grateful for my parents, Sadashiv and Shobha, and my sister, Nit
9、a, whose encouragement and influence throughout my life has provided me with the foundation to succeed in this research. I would also like to thank my friends at the PAMI lab; I could always count on them for sharing ideas and having many laughs along the way. Lastly, I would like to thank the resea
10、rchers in the BCI group at the Institute of Bomedical Engineering, University of Technology, Graz, Austria for providing data for this research. v This is dedicated to the ones who were the first to teach me a love for knowledge, my parents, Sadashiv and Shobha Dharwarkar. vi Contents Chapter 1_ 1 I
11、ntroduction _ 1 1.1 Introduction to Brain Computing _ 2 1.2 The Electroencephalogram_ 4 1.2.1 The Biology of EEG _ 4 1.2.2 A Brief History of EEG_ 5 1.2.3 Signal Conditioning Challenges _ 6 1.2.4 The International 10-20 System for Electrode Placement _ 8 1.3 Overview of the Brain-Computer Interface
12、_ 10 1.4 Cognitive Tasks and Methods in EEG Communication_ 13 1.4.1 Visual Evoked Potentials _ 13 1.4.2 P300 Evoked Potential _ 14 1.4.3 Slow Cortical Potentials _ 14 1.4.4 -Rhythm and Motor Imagery _ 15 1.4.4.1 Asynchronous Communication Protocols_ 16 1.5 Scope_ 17 Chapter 2_ 20 Background _ 20 2.1
13、 Developments in Time-Frequency Analysis of Motor Imagery EEG 20 2.2 Prospects and Challenges in Motor Imagery BCI _ 25 2.3 Description of Data Set_ 27 Chapter 3_ 29 Overview of Research_ 29 3.1 Problem Statement _ 29 3.2 Thesis Organization _ 32 Chapter 4_ 33 vii Feature Set 1 Adaptive Autoregressive Model _ 33 4.1 The Autoregressive Model_ 33 4.2 The Adaptive Autoregressive Model_ 37 4.2.1 Introduction to Kalman
