EEG and Eye-Blinking signals through a Brain ... - IEEE Xplore

0 downloads 0 Views 1MB Size Report
and S waves to construct two signals such as meditation and attention. Additionally, we can also extract the eye-blinking signals from BCI. Therefore, attention ...
EEG and Eye-Blinking signals through a Brain­ Computer Interface Based Control for Electric Wheelchairs with Wireless Scheme 2 l' 2 Jzau-Sheng Lin , , Kuo-Chi Chen , and Win-Ching Yang

1' , Department o/Computer Science and Information Engineering, 2 Department 0/Electronic Engineering, National Chin-Yi University o/Technology Taichung, Taiwan 411, R,O.C. *Corresponding Author: [email protected]

Abstract-This paper mainly use simple unipolar electrode to

LED.

capture EEG from the forehead to build a Brain-Computer

learning algorithm for BCI based on combining features to aim

Interface (BCI) based control for electric wheelchairs through

at reducing the training process.

Bluetooth for paralyzed patients. We have normalized and

S

f3,

a,

f)

Additionally,

incorporate interactivity value in a BCI system to control an

we can also extract the eye-blinking

electric wheelchair. And, the BCI system will be constructed to

signals from BCI. Therefore, attention and eye-blinking signals

allow communication between human and machine and to

can be collected as the control signals through a Bluetooth

supply real-time translation of brain and eye-blinking waves

interface and the electrical interface in electric wheelchair. The

into command like "tum to a direction", "moving forward", or

experimental results confirmed that this system can provide a

"stop" to control wheelchair.

convenient manner to control an electric wheelchair.

This paper is divided as follows: Section 2 introduces the

Keywords-EEG; Bel; Wheelchair

I.

[7] demonstrated a semi-supervised

In this paper, we propose a novel and simple technology to

waves to construct two signals such as meditation and

attention.

Liu and Zhao

architecture of BCI; Section 3 demonstrates the proposed hardware and software structures; The steering control of

INTRODUCTION

electric wheelchair is described in Section 4; and Section 5 is the conclusion.

Recently, the research field on Brain-Computer-Interface (BCI) provides a new direction to construct an interactive system which can translate human brainwave into control signals for computer-application devices [1-11]. BCI provides a

II.

communication channel not based on nerves and muscles to allow users to communicate

without

ARCHITECTURE OF BCI

A BCI is a novel communication system which enables

movement with the

users

external world [1]. BCls are the only means to infer user intent

to

translate

Electroencephalography

through direct measures of brain activity usually via EEG

brain (EEG),

actIVIty, into

measured

a control signal

by and

sending to computers. BCI have been exposed as a promising

signals. A BCI system is just to translate EEG signals from a

tool for paralyzed people and healthy people in several

reflection of brain activity into user action through system's

computer applications such as medical assistant devices [8] and

hardware and software. BCI approaches are based on a variety

video games [9].

of strategies to generate control signals which may be the result of auditive or visual stimulation or of imaginary motor tasks.

Fig. 1 shows a schematic of a BCI system. The system involves EEG acquisition, signal processing, and an application

There is a rising tendency in the growing recognition

interface together with the applications. The block of signal

requirements of BCI system. The paralyzed patients can not

processing

operate obiects or communicate their needs, even though their

includes preprocessing,

feature

extraction,

classification.

mental capabilities are integral. These existing BCI systems

Brain-Computer Interface (BCI)

focus on developing new superior control and communication technologies for those with strict neuromuscular disorders.

I

Applications \/R

Videogame

There have been several studies using the signals of brain



activity to control machines. For example, BCI systems were constructed to control electric wheelchair [1-3]. Three BCI

Electric devices

approaches have been implemented at Aalborg University SS­ YEP, MI, and NMI and discussed in reference [4]. A subject­ independent BCI based on motor imagery was discussed by Lotte et. al. [5]. In reference [6], Soraghan et. al. proposed optical brain-computer interfaces to drive triple wavelength

731

I

and

Fig. 1 The brain-computer interface system Fig. 3 The flowchart for Bluetooth receiver and EEGs classifier

III.

THE PROPOSED STRUCTURE

(

)

Start

As shown in Fig. 2, the framework for the signal processing of EEGs with Bluetooth interface is expressed. The EEG signal was extracted from EEG acquisition. In this system, we used the NeuroSky's headset MindSet to capture EEG and eye­

Initialize compass mode to

blinking signals. With headphones on the sensor to read brain waves, the brainwaves were transmitted by the Bluetooth wireless modules. In the receiving part,

,------,---1

we integrated a

Bluetooth module in a personal computer with a software

"direction"

Read EEG value

1---------,

No



interface organized by using of MA TLAB. Two types of EEG signal were classified such as raw EEG and long EEG. The raw

v••

EEG signal was used to convert electrical voltage to control the electric wheelchair. Then we classified two kinds of digital signal such as Attention, Meditation from long EEG brainwave. The flowchart for the Bluetooth receiver and EEGs classifier is shown as in Fig. 3. The flowchart of OSP block in Fig. 2 is shown as in Fig. 4. In this block, a compass was used to decide the forward direction of electric wheelchair.

Bluetooth

EEG acquisition

Transm itter Fig. 4 The flowchart of DSP block in figure 2

Wheelchair

\\\

Bluetooth Receiver



and

OSP Block (MATLAB



h

Control

:



Interface

0 .!!

Fig. 2 The framework for the signal processing of EEG Signals with Bluetooth interface



GOOD

0 '().5

.,

OK]

8 EEGda'a

"0"

594

596

Attention

jC

598 600 Timers Meditation

602

lT�=31':EJ

R eadEEG signal



Error

596

598 Timers

600





596

598 Timers

600



Fig. 5 The GUl for the direction selection,EEG raw data,

Yeo

In Fig. 4, we set the operation modes of electric wheelchair

RAW EEG or LongEEG?

RAW EEG

as "forward" and "direction". Initially, the compass was set to "direction" mode. Then we read EEG signals from Bluetooth interface. If a correct brainwave was received, we rotated

Long EEG

compass and displayed EEG raw data, "meditation", and "attention" signals on the GUI panel. The GUI panel for rotating compass, EEG raw data, "meditation", and "attention" signals is shown as in Fig. 5. Additionally, we checked the correct EEG signal to be an eye-blinking or not. If it was not, then we switched power on to order the wheelchair moving forward if the mode is "forward"

732

and changing the mode to "direction". Otherwise, the system

signal through RE-232C and converted the digital value to

set the mode as "forward" mode. Fig. 6 shows the completed

analog volt to provide the speed and torque.

experimental system. The commands for the operation to control the electric wheelchair are shown as in Table 1. In this system, we used a

CONCLUSION

V.

compass panel and one eye-blinking pulse to decide the

In this paper, EEG and Eye-Blinking signals through a

moving direction. Additionally, it will be stopped if the

BCI interface based control for electric wheelchairs with

electric wheelchair detected an obstacle or two eye-blinking

wireless scheme is proposed. In a conventional system, uses

pulses. Then, the electric wheelchair moves forward if a high

require hand to control electric wheelchair. In the proposed

value of "attention" signal is calculated.

system, a simple unipolar electrode was used to capture EEG

In the experimental system, a notebook computer was

and eye-blinking signals from the forehead and transmitted to

used to as the block of Bluetooth receiver and EEGs Classifier,

a notebook computer through a Bluetooth interface. The brain

DSP block, and wheelchair control interface in Fig. 2. Then

and eye-blinking signals were calculated by the notebook

the commands are transmitted to the electric wheelchair

computer and transmitted through RS-232C interface.. The

system through RS-232C interface.

proposed

system

is

a

low

cost

and

easy

control

with

"attention" and eye-blinking signals. NeuroSky's headset MindSet

RS-232C

r

Sunplus SPCE061A

Fig. 7 The architecture of the wheelchair system

ACKNOWLEDGMENT

This work was supported by the National Science Council,

Fig. 6 The experimental system

Taiwan, under the Grant NSC98-2221-E-167-016-MY2. TABLE I THE COMMANDS FOR THE PROPOSED ELECTRIC

RE FERENCES

WHEELCHAIR

IV.

Commands

Extracted signals

Tum to a direction

1 eye-blinking pulse

Moving forward

attention

Stop

2 eye-blinking pulse

THE STEERING CONTROL OF ELECTRIC WHEELCHAIR

[I]

S-. Y. Cho, A. P. Vinod, and K. W. E. Cheng, "Towards a Brain­ Computer Interface Based Control for Next Generation Electric Wheelchairs ",Int. Can! on Power Electronics Systems and Applications, pp. 1-5,2009.

[2]

K. Tanaka, K. Matsunaga, and H. O. Wang, "Electroencephalogram­ Based Control of an Electric Wheelchair ",IEEE Trans. on Robotics, vol. 21,no. 4,pp. 762-766,2005.

[3 ]

F. Gal'an, M. Nuttin, E. Lew,P. W. Ferrez, G. Vanacker,J. Philips,and J. d. R. MiII'an, "A brain-actuated wheelchair: Asynchronous and non­ invasive brain-computer interfaces for continuous control of robots," Clinical Neurophysiology, vol. 119,pp. 2159-69,2008.

[4]

A. F. Cabrera, O. F. do Nascimento, D. Farina, and K. Dremstrup, "Brain-Computer Interfacing: How to Control Computers with Thoughts," First International Symposium on Applied Sciences on Biomedical and Communication Technologies, pp. 1-4,2008. F. Lotte,C. Guan, and K. K. Ang, "Comparison of Designs Towards a SUbject­ Independent Brain-Computer Interface based on Motor Imagery," 31" Annual Int. Can! of the IEEE EMBS Minneapolis, Minnesota, pp. 4543 4546,Sep. 2-6,2009.

[5]

C. J. Soraghan, C. Markham, F. Matthews and T. E. Ward, "Triple Wavelength LED for Optical Brain-Computer Interfaces," Electronic Letters, vol. 45,no. 8,pp. 3 92-394,April,2009.

[6]

M. Liu and M. Zhao, "A Semi-Supervised Learning Algorithm for Brain-Computer Interface Based on Combing Features," 4'h Int. Can! on Neural Computation, pp. 3 86-390,2008.

The wheelchair is controlled trough two motor drivers to operate

two

DC

motors

providing

both

of steering

and

propulsion. The electric wheelchair has two 12-volt batteries with 31-50 Amp-hr to support 24 V for the whole system motors. The 24 V is directly connected to the motor driver to provide the main power. They have a continuously variable speed from 0 to 4 MPH. The motor driver provides the power to the motor. The block diagram of electric wheelchair system is shown as in Fig. 7. In which the sunplus SPCE61A is used as a

central

control

unit

(CCU)

platform

that

contains

all

algorithms and electronics. The CCU accepted commands

733

[7]

J. R. Wolpae,N. Birbaumer,D. J. Mc Farland,G. Pfurtschelier,and T. M. Vaughan, Brain-Computer Interfaces for Communication and Control, Clin. Neurophsiol., vol. 113,no. 6,pp. 767-791,2002.

[8]

A. Lecuyer, F. Lotte, R. B. Reilly, R. Leeb, M. Hirose, and M. Slater, Brain-Computer Interfaces, Virtual Reality and Videogames: Current Applications and Future Trends, IEEE Comput. , vol. 41, no. 10, pp.6672,2008.

[9]

[10]

K. Nielsen,A. Cabrera, and O. Nascimento,"Eeg based bci - towards a better control : Brain-computer interface research at aalborg university," IEEE Transactions on Neural Systems and Rehabilitation Engineering. , vol. 14,no. 2,pp. 202-204,2006.

[11] G. E. Fabiani, D. J. Mc Farland, J. R. Wolpaw, and G. Pfurtscheller "Conversion of EEG activity into cursor movement by a brain-computer interface (BCI), IEEE Trans. on Neural Systems and Rehabilitation Engg.,vol. 12,no. 3,pp. 331-338,Sep. 2004.

A. Cabrera and K. Dremstrup, "Auditory and spatial navigation imagery in brain computer interface using optimized wavelets," Journal of Neuroscience Methods,vol. 174,no. 1,pp. 135-146,2008.

734