Analysis and Implementation of Brain Waves Feature Extraction and Classification to Control Robotic Hand

Mohammed Z Al-Faiz,Ammar A Al-Hamadani

doi:10.31987/ijict.1.3.35

Abstract

In this paper, (i) time domain, frequency domain and spatial domain feature extraction methods were investigated. (ii) Two dimensionality reduction methods were proposed, implemented and compared. (iii) The method pair (feature extraction + dimensionality reduction) that owns the lowest classification error rate will be used to learn a machine learning algorithm to control robotic hand in offline mode. Two classes EEG dataset of three bipolar channels was used. The extracted feature vectors were fed into Support Vector Machine with Radial Basis Function kernel (SVM-RBF) to train the classifier. The experimented time domain feature extraction methods were: Mean Absolute Value (MAV), integrated Absolute Value (IAV), Zero Crossing (ZC), Root Mean Square (RMS), Waveform Length (WL) and Slope Sign Change (SSC). Frequency domain feature was the Autoregressive Feature (AR). Finally, the spatial domain feature was the Common Spatial Patterns (CSP). Matlab codes for Principal Component Analysis (PCA) and channel selection algorithm were designed and used to reduce the dimensionality of the features vector. Results showed that CSP features got the lowest error rate for both dimensionality reduction technique with 2.14%. Results recommends to use channel selection algorithm over PCA since it owns the lowest processing time of 8.2s over 8.5s for PCA.

Highlights

As an important practice in the brain computer interface (BCI) systems, in humanoid robotic hand (HRH)applications, is to uncover the intention of human to move his hands
The highest classification accuracy has been achieved with subject A08 at 97.21New Brain Machine Interface (BMI) system has been developed based on Imagined Body Kinematics (IBK) in [7]
The EEG signal has been filtered in the range of 0.16 and 1 Hz, such low frequency information regarded as features of the IBK

Summary

Introduction

As an important practice in the brain computer interface (BCI) systems, in humanoid robotic hand (HRH)applications, is to uncover the intention of human to move his hands. In [3], novel method of online feature extraction and classification for BCI based EEG signal has been implemented. CSP feature extraction and LDA classifier have been used, the obtained classification accuracy was 95.4% after five runs In another approach such as [5], BCI system with adaptive decoder and discrete wavelet transform (DWT) feature has been proposed. Where the classification error rate was calculated by using one of the feature methods with one of the dimensionality reduction algorithms at a time. The pair of algorithms (feature extractions plus dimensionality reduction) that have the lowest classification error rate were used to control robotic hand as an offline BCI paradigm

Objectives

Methods

Results

Conclusion