Frequency Recognition of Selected Features of EEG Signals with Neuro-Statistical Method

Abstract

The advent of research work to analyze massive multiway oriented electroencephalogram (EEG) signals with low configurable computer is a great challenge. This thesis presents an algorithm for extracting underlying frequency components of such EEG data. Frequency components of these data play a vital role to realize brain-body condition. Usually, a huge amount of time and specially built computers are essential to process these EEG data having different subjects. It also restricts to visualize inherent frequency of EEG for a general practitioner. An algorithm is developed using two-stage cascaded architecture of canonical correlation analysis with neural network named neural canonical correlation analysis (NCCA) to address three major challenges for extracting frequency components from EEG data, such as: (a) It processes massive data which are feed sequentially into neural network, rather than feeding whole data at a time, (b) It uses the conventional personal computer instead of special computer built for such application, (c) It spends very short time for a moderate data set consisting of several ways (time, trials and channels). (d) It considers the nonlinear correlation among the data groups while statistical CCA ignores it. In order to get reliable and robust result, the experimental are carried out with different structures of network such as linear, nonlinear and nonlinear feedback structures. The inherent dominant frequency of 1 Hz having a quite resemblance with EEG landscape has been found. This provides a great opportunity in analyzing brain-body function. Although it is possible to recognize frequency of massive EEG data at shorter time with NCCA than statistical CCA, but subjects differentiation is still a great challenge. In this view, this paper presents a new feature selection (FS) approach based on NCCA. In order to get robust features subset having maximum correlation and minimum redundancies, NCCA is devised to search highly correlated subsets by maximizing correlation among several subdivisions of raw data and pruning the features of lightly scored weights of CCA network. The result of NCCA is very robust in terms of accuracy. In this sense, frequency recognition is very easy using selected EEG features than original features which are inspected from correlation profiles. The computational complexity is also greatly reduced if selected features are used to recognize frequency which is proved theoretically and experimentally. In this connection, elapsed time is calculated and observed that NCCA is about 2 to 33 times faster to recognize frequencies from selected EEG features than original set.