FrD5.4

Proceedings of the 4th International IEEE EMBS Conference on Neural Engineering Antalya, Turkey, April 29 - May 2, 2009

DEVELOPMENT OF BRAIN-COMPUTER INTERFACE (BCI) MODEL FOR REAL- TIME APPLICATIONS USING DSP PROCESSORS N.Sriraam Center for Biomedical Informatics and Signal Processing, Department of Biomedical Engineering SSN College of Engineering, Chennai 603110 Email: [email protected]

G.Karthikeyan Undergraduate student, Department of Biomedical Engineering SSN College of Engineering, Chennai 603110 Email: [email protected]

Abstract: Brain computer interfaces (BCIs) are systems that provide translation of the electrical activity of the brain into commands which can control devices in real-time. It is known that most BCI based application involves non-invasive EEG signals in combination with machine learning algorithms for control of devices. Although there has been some real- time implementation of BCI including cursor movements controlled by a person’s thoughts, they have not been done as a portable device, but rather as a laboratory experiment controlled using multi- core processors and computers. This paper focuses on real- time implementation of BCI for motor imagery tasks using a floating point DSP processor (TMS320C6713DSK). The proposed work involves FIR based preprocessing filter to remove extraneous noise including Electrooculogram (EOG) and Electromyographic (EMG) signals followed by feature extraction using time-frequency domain features. The implications of this kind of a development in Brain- Computer Interface would be tremendous since the whole system can be made into a portable device or a System on a Chip (SoC).

II. A.

EEG Data

Figure 1 Time sequence specification considered for data recording The data set used in this study was recorded at the Biomedical Instrumentation Laboratory of the SSN College of Engineering; Chennai, India. The recording procedure is described below: The electrodes were placed with accordance to the 10-20 electrode placement system and the recordings were done from a single subject for three sessions with each session yielding 300 usable trials and recordings from each of the considered channels were sampled at a rate of 256 Hz. Fig. 1 details the recording time sequence followed. The characteristic of the filter used was that of a band pass filter with the filter range being 0.1Hz to 100Hz. The impedance value of the recording set up was limited within 20 KΏ. The channels recording the frontal activities were considered for the study as the frontal electrodes are the most sensitive ones for the recording of imagery motor activities Different combinations of frontal electrodes were taken into account and the study conducted and the results evaluated accordingly. The frequency range of the signals used in the study was that of the characteristic β-waves of the imaginary thought processes. Fig. 2 shows the sample recordings.

Keywords— Brain- Computer Interface, DSP Processor, Neural Networks, Linear Discriminate Analysis (LDA).

I. INTRODUCTION Brain-computer interface (BCI), a rapidly developing technology, has been constantly improving and new techniques being constantly added to the already existing ones. The main goal of Brain- Computer Interface is creating a new pathway of communication for persons with severe motor disabilities [1-2,4-5]. Most of the present day BCI research is carried out using Electroencephalogram (EEG) signals owing to its non- invasive nature. [1-2]. Though EEG is predominantly diagnostic in nature, their inherent uniqueness in different features can aid in development of BCI devices. This paper focuses on the real- time application of Brain Computer Interface by using a DSP processor for filtering the raw EEG data and using this filtered data for feature extraction and classification.

978-1-4244-2073-5/09/$25.00 ©2009 IEEE

MATERIALS AND METHODS

419

Authorized licensed use limited to: Technische Universiteit Delft. Downloaded on September 3, 2009 at 09:11 from IEEE Xplore. Restrictions apply.

III.

facilitates development of a high- level language compiler. The internal program memory is structured so that a total of eight instructions can be fetched every cycle.

EEG FILTER DESIGN COMPARISON

Fig 3 shows the overview of the proposed process. Filter design is done using both MATLAB and the DSP processor and the EEG data was filtered using the filter designed and the filter characteristics of both the filters were found to be the same. The filter transfer function has been given in Fig 4. The Cut- Off frequencies for the Band- Pass filter of order 400 was given as 1 Hz and 30 Hz. Fig 4 shows the magnitude response of the filter designed.

Features of the C6713 include 264 kB of internal memory (8kB as L1P and L1D Cache and 256kB as L2 memory shared between program and data space), eight functional or execution units composed of six arithmetic-logic units (ALUs) and two multiplier units, a 32-bit address bus to address 4 GB (gigabytes), and two sets of 32-bit general-purpose registers. Fig.5 shows the typical DSP flow process.

Trial 1 right hand signal Am p(m icro volts)

2000

IV.

1000 0 -1000 -2000

0

10

20

30

40 50 60 Time (seconds) Trial 1 left hand signal

70

80

90

Amp(micro volts)

1000

Feature Extraction was done using time- domain, frequency domain and non- linear chaotic parameters. The feature extraction methods used were, Levinson- Durbin (LD) method, power spectral density using the Burg (PSD Burg) method and the Hurst exponent (HE) method. The LD method is used to calculate the autoregressive (AR) Coefficients of an nth- order AR linear process is calculated by using the relation: H(z)=1/A(z)=1÷(1+a(2)z^1+....+a(n+1)z^-1) ->(1)

0

-1000

-2000

FEATURE EXTRACTION

0

10

20

30

40 50 Time (seconds)

60

70

80

90

Hurst exponent which possesses spectral characteristics is one of the key parameters in non- linear dynamics. The values of HE is obtained using the statistical technique called “Rescaled range analysis [3]. Figs 6-7 show the feature extraction results obtained for the test data considered for this work.

Figure 2 Plot of the data sets of both right and left hands used in the study A.

TMS320C6713DSK- Overview

The C6713 DSK is a low-cost standalone development platform that enables users to evaluate and develop applications for the TI C67xx DSP family. The DSK also serves as a hardware reference design for the TMS320C6713 DSP.

Figure 4 Magnitude Response of the FIR filter (DSK o/p)

V.

CLASSIFICATION

Classification of the extracted features were carried out using feedback (Elman Network) and feedforward (multilayer perceptron) neural network classifiers. The results obtained using the classifiers were compared and the most suitable classifier for a given set of feature extraction techniques was studied. The performance of the classsifiers were evaluated in terms of classification accuracy (CA) which is defined as in (2)

Figure 3 Block set up for the study The TMS320C6XXX processors are based on the Very Long Instruction word (VLIW) architecture. This architecture

420

Authorized licensed use limited to: Technische Universiteit Delft. Downloaded on September 3, 2009 at 09:11 from IEEE Xplore. Restrictions apply.

CA= CD/TA

DSP processor and a fully functional real- time BCI system can be developed based upon the preliminary works proposed.

(2)

Where, • CD refers to total number of correctly detected left hand and right hand patterns by the neural network • TA refers to total number of applied patterns.

0.7 Imaginary LH Movement-Pburg Imaginary RH Movement-Pburg

0.6

For this study, 400 features are used for training and 600 for testing. P S D-B urg

0.5

0.4

0.3

0.2

0.1

0

0

5

10

15

20 25 30 EEG patterns

35

40

45

50

Figure 7 Feature extraction results using PSD Welch Method Figure 5 DSP process flow

Elman network with Levinson 1

Figs 8(a)-(c) shows the classification results obtained using Elman neural network. It can be observed an overall classification accuracy of 99.5% is achieved using the Hurst exponent with Elman neural network.

0.9 0.8 0.7

EN output

0.6

0.03 Imaginary LH Movement-Levinson Imaginary RH Movement-Levinson 0.025

0.5 0.4

Levinson-Durbin Coefficients

0.3

0.02

0.2 0.1

0.015

0

0

100

200

300 No of samples

400

500

600

0.01

Figure 8(a): Classifier output using Levinson- Durbin 0.005

Elman network with Pburg 1

0

0.9

0

5

10

15

20 25 30 EEG patterns

35

40

45

50 0.8

Figure.6 Feature extraction results using Levinson- Durbin method

0.6 EN output

VI.

0.7

CONCLUSION

This paper discusses the design and comparison of preprocessing filters for BCI applications using real- time DSP processors. The different types of features extracted were classified using Neural Network classifier. For the filter designed using the DSP processor, it was found that the transfer characteristics of the filter was quite similar to the transfer characteristics of the filter designed using MATLAB. The work carried out on MATLAB can be extended to the

0.5 0.4 0.3 0.2 0.1 0

0

100

200

300 No of samples

400

500

600

Figure 8(b): Classifier output using PSD Burg’s Method

421

Authorized licensed use limited to: Technische Universiteit Delft. Downloaded on September 3, 2009 at 09:11 from IEEE Xplore. Restrictions apply.

[13] Non-invasive classification of cortical activities for brain computer interface: A variable selection approach Besserve, M, Biomedical Imaging: From Nano to Macro, 2008. ISBI 2008. Elman network with HE

[14] A DSP based Brain Computer Interface system, Morrison, Masters Thesis, University of Florida.

1 0.9

[15] Digital Signal Processing and Applications with the C6713, Ralph Chassaing.

0.8 0.7

[16] Bill Davidson, Matlab program for Hurst exponent, October 2003

EN output

0.6 0.5 0.4 0.3 0.2 0.1 0

0

200

400

600 No of samples

800

1000

1200

Figure 8(c): Classifier output using HE

REFERENCES [1 ]N.F.Incea,A.H.Tewfika, S.Aricab,Extraction subject-specific motor imagery time–frequency patterns for single trial EEG classification,Computers in Biology and Medicine 37 (2007) 499 – 508 [2] J.Wolpaw. Brain-computer interfaces as new brain output pathways. J Physiol, 579:613--619, 2006. [3] Bassingthwaighte JB, Raymond GM, Evaluating rescaled range analysis for time series, Ann Biomed Eng 1994 Jul-Aug; 22(4):432-44. [4] An adaptive filter bank for motor imagery based Brain Computer Interface, Thomas, Engineering in Medicine and Biology Society, 2008. [5] J. Wolpaw, N. Birbaumer, D. McFarland, G. Pfurtscheller, and T. Vaughan. Brain-computer interfaces for communication and control. (Invited review). Clin Neurophysiol, 113:767--791, 2002. [6] V.Srinivasan, C.Eswaran, and N.Sriraam. 2005. Artificial Neural Network based epileptic detection using Time-domain and Frequency-domain features , Journal of Medical Systems, Kluwer Academic Publishers, New York, USA Volume 29, Number 6. 647-660. [7 ]N.J. Huan and .Palaniappan,Electroencephalogram Signal Classification Using Linear Discriminant Analysis for Brain-Computer Interface Design, Proceedings of M2USIC 2004, 9-12 ,2004. [8] N. Sriraam, G. Karthikeyan, J. Kamala Kannan Performance evaluation of imaginary motor activities for Brain- Computer Interface Applications. International Symposium on Global Trends in Biomedical Informatics, Chennai, Jan 2008. [9] Model-based responses and features in Brain Computer Interfaces, Kamrunnahar, Engineering in Medicine and Biology Society, 2008. EMBS 2008. [10] Classification of EEG with structural feature dictionaries in a brain computer interface,Goksu, Engineering in Medicine and Biology Society, 2008. EMBS 2008. [11] An Intelligent Brain Computer Interface of Visual Evoked Potential EEG, Chen, Intelligent Systems Design and Applications, 2008. ISDA '08. [12] Linear Discriminant Analysis on Brain Computer Interface, Perez, Intelligent Signal Processing, 2007. WISP 2007.

422

Authorized licensed use limited to: Technische Universiteit Delft. Downloaded on September 3, 2009 at 09:11 from IEEE Xplore. Restrictions apply.

Development of Brain-Computer Interface (BCI) Model ...

kind of a development in Brain- Computer Interface would be tremendous since the .... feedback (Elman Network) and feedforward (multilayer perceptron) neural ...

401KB Sizes 0 Downloads 181 Views

Recommend Documents

Development of Brain-Computer Interface (BCI) Model ...
It is known that most BCI based application involves non-invasive EEG ... kind of a development in Brain- Computer Interface would be tremendous since the ...

A Brain–Computer Interface (BCI) for the Detection of ...
is still the primary technique for finding and eliminating these objects. Even with the aid of underwater robotics to capture data, the task of processing the imagery ...

BCI Oecol.pdf
Montañana, 1005, 50192 Saragossa, Spain. Author's personal copy. Whoops! There was a problem loading this page. BCI Oecol.pdf. BCI Oecol.pdf. Open.

BCI Rules Inconsistent.pdf
Government of India,. South Block, Raisina Hill,. New Delhi-110011. The Hon'ble Law Minister. Government of India,. 4th Floor “A” Wing,. Shashtri Bhawan, Rajendra Prasad Road,. New Delhi-110001. The Hon'ble Chairman. Legal Education Committee,. B

An elastic–plastic interface constitutive model - Semantic Scholar
Available online 27 February 2004. Abstract. An interface constitutive ... +1-617-253-1635; fax: +1-617-258-8742. E-mail address: ... occurs naturally across the interface, and traction-free cracks form and propagate along element boundaries.

Matlab-Based Graphical User Interface Development ...
exploiting Matlab's abundant GUI tools. The proposed ... Even though third party GUI tools are available for ... the user to plot the data without the ability to, for example, filter it. ..... visualize system response within a GUI-based environment.

A Biological Development model for the Design of ...
Development involves cell division, the emergence of pattern, change in form, cell ... At present only combinational applications are considered, hence the EUs are ..... Evolvable Hardware Workshop, IEEE Computer Society, Los Alamitos, Ca, ...

Development of a mathematical model for simulating ...
+81 89 946 9828; fax: +81 89 946 9916. .... ј рnpDC юAf KCЮ PC АPT. V CрtЮ. V OрtЮюV ..... For model predictions, the initial free volume of the film package ...

Model-based Toolchain for the Efficient Development of ...
Apr 12, 2011 - and software. ... targets on the definition of vendor-neutral tool interfaces based ... RTP-ModelBus is taking a service-oriented approach into.

A Biological Development model for the Design of ...
Development involves cell division, the emergence of pattern, change in form, cell ... The Execution Unit (EU) is the circuit incorporated to do the real calculation of the target application. The inputs to each EU come from its immediate west and no

Research Article Development of Novel In Silico Model ...
drug companies. 2.2. ... 150mm, 5 μm, Waters, USA) under a gradient flow of ... software in training sets of 32 diverse noncongeneric com- ..... log PC 240 min.

Development and internal validation of a multivariable model to ...
Download. Connect more apps... Try one of the apps below to open or edit this item. Development and internal validation of a multivariable model to predict.pdf.

A Tool for Model-Driven Development of Collaborative Business ...
In [13, 15] a model-driven development method for collaborative business processes, which is based on the Model-Driven Architecture (MDA) [10], has been ... encourages a top-down approach and supports the modeling of four views: ..... Workshop of Req

Model-Based Test Driven Development of the Tefkat ...
testing in traditional systems [3] or object-oriented systems ... simple data types or traditional objects. .... this reference points to an ECore XML file that contains.

A Method for the Model-Driven Development of ...
prototype tool for model transformation that we are developing. In this tool, model ...... on Data Management Issues in E-Commerce, 31(1) (2002). [CompTIA] ...

Development of Novel In Silico Model to Predict Corneal Permeability ...
Oct 26, 2010 - drug companies. 2.2. Preparation of .... software in training sets of 32 diverse noncongeneric com- pounds including ... 6.1.12.33, Japan), ACD/Chemsketch (Freeware version 10), ..... database of structures that avoids the need to calc

Development of Software for Feature Model ... - Semantic Scholar
Dec 2, 2006 - this master thesis is to analysis existing four popular feature diagrams, to find out ..... Clustering. Cluster descriptions. Abstraction. Abstract descriptions. Classification. Classify descriptions. Generalization. Generalize descript

Development of Novel In Silico Model to Predict Corneal Permeability ...
Oct 26, 2010 - 2 Department of Pharmacology, All India Institute of Medical Sciences, New Delhi 110029, India. 3 Department of Ocular .... drug companies. 2.2. Preparation of the .... software in training sets of 32 diverse noncongeneric com-.

Development of Spoken Language Model for Automatic ...
from the sentences with respect to the context using the language model. This has various applications in various situations like for say if you want to certain ...

A Biological Development model for the Design of ...
using evolution that have the ability to “recover” themselves from almost any kinds of .... The digital organism employed in this application is made up of 3x3 identical digi- tal cells. ... sized by ISE 6.1i from Xilinx, downloaded into the hard

03 A Model for Leadership Development of Private University ...
03 A Model for Leadership Development of Private University Student Leaders.pdf. 03 A Model for Leadership Development of Private University Student ...