Neural Control using EEG as a BCI Technique for Low Cost

Prosthetic Arms

Shyam Diwakar

, Sandeep Bodda

, Chaitanya Nutakki

, Asha Vijayan

Krishnashree Achuthan

and Bipin Nair

Amrita School of Biotechnology, Amrita Vishwa Vidyapeetham (Amrita University),

Clapana P.O., Amritapuri, Kollam, Kerala, India

Amrita School of Engineering, Amrita Vishwa Vidyapeetham (Amrita University),

Clapana P.O., Amritapuri, Kollam, Kerala, India

Keywords: Brain Computer Interface, Prosthetic Arm, Kinematics, EEG, Iterative Algorithm, Machine Learning.

Abstract: There have been significant advancements in brain computer interface (BCI) techniques using EEG-like

methods. EEG can serve as non-invasive BMI technique, to control devices like wheelchairs, cursors and

robotic arm. In this paper, we discuss the use of EEG recordings to control low-cost robotic arms by

extracting motor task patterns and indicate where such control algorithms may show promise towards the

humanitarian challenge. Studies have shown robotic arm movement solutions using kinematics and machine

learning methods. With iterative processes for trajectory making, EEG signals have been known to be used

to control robotic arms. The paper also showcases a case-study developed towards this challenge in order to

test such algorithmic approaches. Non-traditional approaches using neuro-inspired processing techniques

without implicit kinematics have also shown potential applications. Use of EEG to resolve temporal

information may, indeed, help understand movement coordination in robotic arm.

1 INTRODUCTION

Brain Computer Interfaces (BCI) is a novel fast

evolving technology connecting the brain to a

computing devices (Birbaumer, 2006; Wolpaw et al.,

2002), now seen as a ubiquitous detection and

diagnostics tool. The domain of EEG-based BCIs

include several applications like controlling a cursor

on the screen (Yuanqing Li et al., 2008), selecting

letters from keyboard playing games (Donchin et al.,

2000), controlling a prosthetic arm (Bi et al., 2013;

Muller and Blankertz, 2006). BCI devices are used

in multiple modes including invasive or non-

invasive (Leuthardt et al., 2004; Owen and Coleman,

2008; Pfurtscheller et al., 2010), synchronous and

asynchronous (Md Norani et al., 2010) modes in

current BCI applications. Prosthetic articulators

based on EEG play a vital role in the area of haptics

and sensorimotor control (Wolpert and Flanagan,

2010). In this position paper, we discuss the

evolution of EEG-based BCI techniques for control

of neuro-prosthetic articulators and include our case

study on a low-cost robotic arm model.



Electroencephalography (EEG) is a widely used

neuroimaging technique, owing to its high temporal

resolution, low cost, high portability and has become

a practical choice for BCI. The quality of EEG

signals are usually affected by noise from scalp,

skull and a significant contribution from

background noise (Nicolas-Alonso and Gomez-Gil,

2012). Various EEG-based BCIs differ based on

user intent to extract neuro-electrical activity.

Techniques commonly used are based on

recognition of specific pattern in EEG to a particular

task (Millán et al., 2002; Pfurtscheller et al., 2003;

Wolpaw et al., 2002), identification of characteristic

waveforms in EEGs which follow an event

(Birbaumer et al., 2003; Farwell and Donchin,

1988), and the presence of periodic waveforms in

EEGs in the range of frequencies of an oscillatory

signal corresponding to a light flash stimulus

(Friman et al., 2007). EEG signals, based on

specific responses related to a task-related stimulus,

serve as an input for BCI systems to control

prosthetic arms (Figure 1). EEG patterns can be

extracted using Sensory Motor Rhythms (SMR).

Motor movement or imaginary movement changes

the oscillatory patterns of EEG, resulting in

270

Diwakar S., Bodda S., Nutakki C., Vijayan A., Achuthan K. and Nair B..

Neural Control using EEG as a BCI Technique for Low Cost Prosthetic Arms.

DOI: 10.5220/0005134802700275

In Proceedings of the International Conference on Neural Computation Theory and Applications (NCTA-2014), pages 270-275

ISBN: 978-989-758-054-3

 2014 SCITEPRESS (Science and Technology Publications, Lda.)

suppression of amplitude (ERD) or enhancement in

amplitude (ERS) for mu or beta rhythms.

(McFarland et al., 2000; Pfurtscheller et al., 2006;

Wolpaw et al., 2002).

1.1 Implementation Issues regarding

EEG Based Techniques

Although EEG is portable (Nicolas-Alonso and

Gomez-Gil, 2012; Tanaka et al., 2005) and cost

effective (Bi et al., 2013; Vespa et al., 1999) for

research purposes, poor signal to noise ratio or

artefacts are recorded during signal acquisition. For

statistical significance, EEG analysis require

complex data analytics and significantly large

dataset with a fair number of subjects (Schlögl et al.,

2002). Due to low spatial resolution, EEG signals

need elaborate interpretation in order to functionally

hypothesize on areas activated by particular

response (Srinivasan, 1999). Pre-recording setup

times are also significantly long.

Noise in the signals plays an important role in

EEG signal analysis and interpretation of data

(Repovs, 2010). There is a necessity for efficient

strategies towards noise prevention and removal.

1.2 Neurological Mechanisms Used in

BCI

Control signals generated by BCI methods

correspond into 5 main categories namely

sensorimotor activity (ERD/ERS), VEP, P300, SCP,

activity of neural cell (Wolpaw et al., 2002), and

furthermore into two additional categories, mental

arithmetic tasks (non-movement) and multiple

neural mechanisms (Anderson, 1995; Gysels et al.,

2005).

Previous studies (Anderson, 1995; Choi, 2012;

Craig and Nguyen, 2007; Leeb et al., 2007; Pires et

al., 2008; Tanaka et al., 2005) have shown that these

neurological mechanisms were used in different

motor-related tasks. A previous work (Tanaka et al.,

2005) had extracted ERD/ERS neurological

phenomena for pattern classification of turn-left and

turn-right events concerning a BCI-enabled wheel

chair. Similar methods were employed for moving-

forward and moving-backward tasks but used SVM

(Choi, 2012), Linear classifier (Leeb et al., 2007),

Artificial Neural Network(ANN) (Craig and

Nguyen, 2007). Methods using EEG-based

techniques involved low-pass filtering(7 Hz) of the

P300 wave and feature-extraction using windowing

or normalization (Pires et al., 2008).



Figure 1: Schematic representation of a BCI-controlled

low-cost robotic upper arm model.

Datasets were then classified using Bayesian

classifier for categorizing multi-class movement

data. A motor–task study using SSVEP-based

methods (Middendorf et al., 2000), employed feature

extraction using Welch periodogram (512 FFT

points) and involved SVM-based classification of

turning right/going forward and stopping (Dasgupta

et al., 2010). Neural Networks with back-

propagation learning have been shown to classify

arithmetic calculation task features, extracted using

Burg method/Levison algorithm (auto-regression

models) (Anderson, 1995). Studies on word-

generation and motor activity used Butterworth filter

(1-40 Hz) coefficients and phase locking

values(PLV) (Lachaux et al., 1999) as features and

classified the dataset using SVM (Gysels et al.,

2005). Laplacian algorithms have also been used for

movement-related tasks (McFarland et al., 2000).

1.3 Prosthesis & Control with EEG

Signal

Brain computer interface (BCI) have been employed

to control prosthetic arms (Wolpert and Flanagan,

2010) in order to do specialized tasks, namely,

reaching the target with an optimal feedback

(Mitrovic et al., 2010), examining the different

parameters of an object (Saal and Vijayakumar,

2010). Prosthetic or robotic arms consist of a series

of links which were equipped by an end-effector to

move in a 3D space (AbuQassem, 2010; Wolpert

and Flanagan, 2010). Links were driven by motors

NeuralControlusingEEGasaBCITechniqueforLowCostProstheticArms

271

changing the orientation of the manipulator

(Megahed, 2013).

Modeling of robotic arm behavior for prosthetic

devices involved kinematics of the robotic links

(Kay, 2005). Different algorithms have been

proposed to solve the robotic kinematics; DH

method (Iqbal et al., 2012), homogenous method

(Mitra, 2012) for forward kinematic model and

analytical method (Iqbal et al., 2012) for inverse

kinematics model. Analytical methods are

subcategorized into geometric and algebraic

approaches. Geometric approach has been applied to

simple robotic structures with few DOF, whereas

algebraic approach were used for greater DOF

(Kucuk and Bingul, 2006). Triangulation and CCD

(Cyclic Coordinate Descent) algorithms (Muller-

Cajar and Mukundan, 2007) have been used to solve

inverse kinematics with a scalable number of links.

Quaternion algebraic approach (Sahu et al., 2008)

has been shown to be computationally cost-effective

compared to homogenous methods.

A feedback system was shown to control grip

force of a gripper/ grasper (Westling and Johansson,

1984). Controllers like PID (open loop

optimization), OFC (closed loop optimization) have

been used to optimize the motor commands with the

help of cost functions like iterative linear quadratic

Gaussian (ILQG) which use trajectory planning and

execution in multiple steps (Mitrovic et al., 2010)

instead of using sensors. Stochastic optimal control

is another application which is now widely being

used for planning and controlling of robotic systems

(Rawlik et al., 2012). It has been demonstrated that

Kullback-Leibler divergence minimization algorithm

could present a solution towards stochastic optimal

control (Rawlik et al., 2012).

1.4 Non-Traditional BCI Methods

Models of spiking neural networks (SNN) take

advantage of precise timing of spikes to produce rich

dynamic behavior(Kasabov et al., 2013). The study

of enriched cognitive systems embodied interaction

with environment could be achieved by employing

SNN. Futuristic design of hybrid architecture

inspired by the working human brain have led to

complex structures and significant models of

internal dynamics has seen in representation of the

model kinematics structures such as the cerebellum

(Furber et al., 2014; Shepherd et al., 1998).

The key contribution to this direction of study

will be a method for simulating a spiking neural

network with high parallelism relying on data

organization has seen in internal representation

mimicking the motor circuit in the brain. An

evaluation of user configurable structures resemble

primary circuit of movement coordination such as

the cerebellum or the V1 motor cortex may suggest

discrete spike based transformation models

generating responses appropriate to kinematic

algorithms via data classification technique such

circuits may have ranges modified by nature of input

and delays configuring the plasticity of adaptive

responses as seen in biological neural circuit.

CMAC (Albus, 1975) had proposed a pattern

separation algorithm based on internal representation

model of the cerebellar neurons that perform

movement coordination tasks. While spiking neuron

models of CMAC-like algorithms are being

elaborated, benchmark nonlinear tests have shown to

function using simple neural microcircuit models

(Joshi and Maass, 2005).Such neural circuits may

perform indifferently to the kind of feedback

received compared to the control performance of

traditional techniques. A trend of novel spiking

neural circuit based methods for SLAM techniques

may help bridge the gap for the BCI devices and

such low-cost articulator models.

1.5 Implementation Issues for Low

Cost Prosthetic Devices

For sensorimotor control, the primary aim was to

accomplish a task of reaching a specified target.

Targets for low economic cost prosthetic arms

include an ability of generalization of tasks (say

grasp task) without significant precision or high

accuracy. Adapting variability models in kinematic

algorithms and using learning methods, some of the

issues may be overcome (Vijayan et al., 2013).

Position feedback is measured in some robotic arms

using sensors (Mitrovic et al., 2010). In devices

without sensor-based feedback of real-time

localization, effective prediction-correction schemes

may be needed (Kalanovic et al., 2000).

Although a major humanitarian necessity, the

major challenges faced when designing a low cost

prosthetic devices include the economic cost for

research and development, local availability of

components, device functionality, prediction of time

of failure, design simplicity (D’Apuzzo et al., 2012).

Avoiding sensory feedback decreased cost but

increased the localization variability in models.

However, with a low-cost prosthesis implementation

issues such as position control, simultaneous

localization models and power management pose

additional challenges post-design. We have however

regarded a task-based control model in the context

of this paper.

NCTA2014-InternationalConferenceonNeuralComputationTheoryandApplications

272

2 CASE STUDY: LOW-COST

PROSTHETIC ARM CONTROL

USING EEG

In our study, we used a indigenously designed

robotic ARM (Amrita Robotic Manipulator), a part

of remotely-triggered experiments available online

(Vijayan et al., 2013) with 6 DOF as a prototype for

a prosthetic upper arm. We tested the kinematic

behavior of the manipulator with DH method and

algebraic method (AbuQassem, 2010) for forward

kinematics and triangulation algorithm (Muller-

Cajar and Mukundan, 2007) for inverse kinematics.



We used F3 and F4 channels data for extracting

movement patterns. Signal pre-processing was done

using band-pass filtering from 6 to 30 Hz to obtain

Mu and beta rhythms.

To extract features, we used power spectral

density and cross-correlation analysis from

preprocessed signal data (Hosni et al., 2007).

Movement of the arm to left or right directions were

categorized as two classes. We used classification

(Vijayan et al., 2013) for translating imagery to

robotic articulation (Figure 2). In our study,

prosthetic arm had a localization variability of ±5

cm variability. Mapping and localization of end-

effector positions were corrected using error

minimization algorithms (work in progress).

Previous studies on pattern classification of

signals related to motor tasks used to train prosthetic

devices like wheelchairs have shown a high level of

accuracy (~80%) (Tanaka et al., 2005).The

scalability of similar techniques on high-end devices

like the DLR JUSTIN arm or DARPA ARM may

need detailed studies and outreach modifications.



Figure 2: Accuracy of robotic arm dataset for different

classification algorithms.

3 CONCLUSIONS

The paper aims to highlight the current progress in a

humanitarian challenge of redesigning a low

economic cost neuro-prosthetic arm that could be

controlled using EEG-based signal re-classification.

Usual techniques include applications of machine

learning and adaptive feature extraction methods to

process EEG real-time and using a learned system to

control the arm using kinematic techniques. Such

methods have their performance and training issues.

While data reliability and over-learning can cause

issues, the device variability requires prediction-

correction or other iterative approaches. The issues

may be solved using feedback via sensors but that

would add considerable financial and computational

overload to the design and implementation. To keep

the low-cost target, internal representation models

may be needed to help the prediction-correction

process. Our case-study using a home developed

(<$50 ARM) suggests the common issues seen with

any low-cost project while allowing us to use the

platform for testing the potential solutions. The

suggestions are as follows: Firstly, while EEG based

tests are reliable for some event-related tasks, a

learned feature extraction approach may help reduce

the noise in the dataset. Secondly, classification has

to have simple mechanisms such as testing using

SVM or ANNs. Thirdly, better approximations are

to be favored over precision. Iterative processing of

kinematics movements may substitute the sensory-

motor feedback model. Spiking neural network

based internal representation models may help

overcome some of the internal representation issues.

While traditional approaches have their own

performance and implementation issues, a novel

non-traditional approach seems inevitable.

As a final word, we indicate that control

mechanisms using BCI may change the design of

kinematics for robotic articulators. It may, therefore,

need a dual-styled approach of classification and

interpretation from EEG to the arm and an internal

representation model to predict the kinematics of the

arm based on the feature-triggered categorization of

movement dynamics.



ACKNOWLEDGEMENTS

This work derives direction and ideas from the

chancellor of Amrita University, Sri Mata

Amritanandamayi Devi. This work is partially

supported by grants SR/CSI/49/2010 and

SR/CSI/60/2011 and Indo-Italy POC 2012-13 from

DST, BT/PR5142/MED/30/764/2012 from DBT and

by Sakshat Virtual Labs, NMEICT, MHRD,

Government of India. Authors would like to thank

Hareesh Singanamala and Dhanush Kumar for their

help with the robotic arm.

NeuralControlusingEEGasaBCITechniqueforLowCostProstheticArms

273

REFERENCES

AbuQassem, M.R., 2010. Simulation and Interfacing of 5

DOF Educational Robot Arm. Islamic University of

Gaza.

Albus, J.S., 1975. A New Approach to Manipulator

Control: The Cerebellar Model Articulation

Controller(CMAC). J. Dyn. Syst. Meas. Control.

Anderson, C., 1995. EEG signal classification with

different signal representations. In: Neural Networks

for Signal Processing. IEEE, Cambridge, MA, pp.

475–483.

Bi, L., Fan, X.-A., Liu, Y., 2013. EEG-Based Brain-

Controlled Mobile Robots: A Survey. IEEE Trans.

Human-Machine Syst. 43, 161–176.

Birbaumer, N., 2006. Breaking the silence: brain-computer

interfaces (BCI) for communication and motor control.

Psychophysiology 43, 517–32.

Birbaumer, N., Hinterberger, T., Kübler, A., Neumann, N.,

2003. The thought-translation device (TTD):

neurobehavioral mechanisms and clinical outcome.

IEEE Trans. Neural Syst. Rehabil. Eng. 11, 120–3.

Choi, K., 2012. Control of a vehicle with EEG signals in

real-time and system evaluation. Eur. J. Appl. Physiol.

112, 755–66.

Craig, D.A., Nguyen, H.T., 2007. Adaptive EEG thought

pattern classifier for advanced wheelchair control. In:

Annual International Conference of the IEEE

Engineering in Medicine and Biology Society. pp.

2544–7.

D’Apuzzo, M., Liccardo, A., Bifulco, P., Polisiero, M.,

2012. Metrological issues concerning low cost EMG-

controlled prosthetic hand. In: 2012 IEEE

International Instrumentation and Measurement

Technology Conference Proceedings. IEEE, pp. 1481–

1486.

Dasgupta, S., Fanton, M., Pham, J., Willard, M.,

Nezamfar, H., Shafai, B., Erdogmus, D., 2010. Brain

controlled robotic platform using steady state visual

evoked potentials acquired by EEG. In: 2010

Conference Record of the Forty Fourth Asilomar

Conference on Signals, Systems and Computers. IEEE,

pp. 1371–1374.

Donchin, E., Spencer, K.M., Wijesinghe, R., 2000. The

mental prosthesis: assessing the speed of a P300-based

brain-computer interface. IEEE Trans. Rehabil. Eng.

8, 174–179.

Farwell, L. a, Donchin, E., 1988. Talking off the top of

your head: toward a mental prosthesis utilizing event-

related brain potentials. Electroencephalogr. Clin.

Neurophysiol. 70, 510–23.

Friman, O., Volosyak, I., Gräser, A., 2007. Multiple

channel detection of steady-state visual evoked

potentials for brain-computer interfaces. IEEE Trans.

Biomed. Eng. 54, 742–50.

Furber, S.B., Galluppi, F., Temple, S., Plana, L.A., 2014.

The SpiNNaker Project. Proc. IEEE 102, 652–665.

Gysels, E., Renevey, P., Celka, P., 2005. SVM-based

recursive feature elimination to compare phase

synchronization computed from broadband and

narrowband EEG signals in Brain–Computer

Interfaces. Signal Processing 85, 2178–2189.

Hosni, S.M., Gadallah, M.E., Bahgat, S.F., AbdelWahab,

M.S., 2007. Classification of EEG signals using

different feature extraction techniques for mental-task

BCI. 2007 Int. Conf. Comput. Eng. Syst. 220–226.

Iqbal, J., Islam, R., Khan, H., 2012. Modeling and

Analysis of a 6 DOF Robotic Arm Manipulator. Can.

J. Electr. Electron. Eng.

Joshi, P., Maass, W., 2005. Movement generation with

circuits of spiking neurons. Neural Comput. 17, 1715–

1738.

Kalanovic, V.D., Popoviic, D., Skaug, N.T., 2000.

Feedback Error Learning Neural Network for Trans-

femoral Prosthesis. IEEE Trans. Rehabil. Eng. 8, 71–

80.

Kasabov, N., Dhoble, K., Nuntalid, N., Indiveri, G., 2013.

Dynamic evolving spiking neural networks for on-line

spatio- and spectro-temporal pattern recognition.

Neural Netw. 41, 188–201.

Kay, J., 2005. Introduction to Homogeneous

Transformations & Robot Kinematics.

Kucuk, S., Bingul, Z., 2006. Robot Kinematics: Forward

and Inverse Kinematics. In: Industrial-Robotics-

Theory-Modelling-Control. pp. 117–148.

Lachaux, J.P., Rodriguez, E., Martinerie, J., Varela, F.J.,

1999. Measuring phase synchrony in brain signals.

Hum. Brain Mapp. 8, 194–208.

Leeb, R., Friedman, D., Müller-Putz, G.R., Scherer, R.,

Slater, M., Pfurtscheller, G., 2007. Self-paced

(asynchronous) BCI control of a wheelchair in virtual

environments: a case study with a tetraplegic. Comput.

Intell. Neurosci. 79642.

Leuthardt, E.C., Schalk, G., Wolpaw, J.R., Ojemann, J.G.,

Moran, D.W., 2004. A brain-computer interface using

electrocorticographic signals in humans. J. Neural

Eng. 1, 63–71.

McFarland, D.J., Miner, L. a, Vaughan, T.M., Wolpaw,

J.R., 2000. Mu and beta rhythm topographies during

motor imagery and actual movements. Brain Topogr.

12, 177–86.

Md Norani, N.A., Mansor, W., Khuan, L.Y., 2010. A

review of signal processing in brain computer

interface system. In: 2010 IEEE EMBS Conference on

Biomedical Engineering and Sciences (IECBES).

IEEE, pp. 443–449.

Megahed, S.M., 2013. Mechanical Design & Production

Department.

Middendorf, M., McMillan, G., Calhoun, G., Jones, K.S.,

2000. Brain-computer interfaces based on the steady-

state visual-evoked response. IEEE Trans. Rehabil.

Eng. 8, 211–4.

Millán, J., Franzé, M., Mouriño, J., Cincotti, F., Babiloni,

F., 2002. Relevant EEG features for the classification

of spontaneous motor-related tasks. Biol. Cybern. 86,

89–95.

Mitra, A.K., 2012. Joint Motion-based Homogeneous

Matrix Method for Forward Kinematic Analysis of

Serial Mechanisms. Int. J. Emerg. Technol. Adv. Eng.

2, 111–122.

NCTA2014-InternationalConferenceonNeuralComputationTheoryandApplications

274

Mitrovic, D., Nagashima, S., Klanke, S., Matsubara, T.,

Vijayakumar, S., 2010. Optimal Feedback Control for

Anthropomorphic Manipulators. In: Int. Conf. Robot.

Autom. 2010). Alaska, USA.

Muller, K.-R., Blankertz, B., 2006. Toward noninvasive

brain-computer interfaces. IEEE Signal Process. Mag.

23, 128–126.

Muller-Cajar, R., Mukundan, R., 2007. Triangulation : A

new algorithm for Inverse Kinematics. In:

Proceedings of Image and Vision Computing New

Zealand. pp. 181–186.

Nicolas-Alonso, L.F., Gomez-Gil, J., 2012. Brain

computer interfaces, a review. Sensors (Basel). 12,

1211–79.

Owen, A.M., Coleman, M.R., 2008. Using neuroimaging

to detect awareness in disorders of consciousness.

Funct. Neurol. 23, 189–94.

Pfurtscheller, G., Allison, B.Z., Brunner, C., Bauernfeind,

G., Solis-Escalante, T., Scherer, R., Zander, T.O.,

Mueller-Putz, G., Neuper, C., Birbaumer, N., 2010.

The hybrid BCI. Front. Neurosci. 4, 30.

Pfurtscheller, G., Brunner, C., Schlögl, A., Lopes da Silva,

F.H., 2006. Mu rhythm (de)synchronization and EEG

single-trial classification of different motor imagery

tasks. Neuroimage 31, 153–9.

Pfurtscheller, G., Neuper, C., Müller, G.R., Obermaier, B.,

Krausz, G., Schlögl, A., Scherer, R., Graimann, B.,

Keinrath, C., Skliris, D., Wörtz, M., Supp, G.,

Schrank, C., 2003. Graz-BCI: state of the art and

clinical applications. IEEE Trans. Neural Syst.

Rehabil. Eng. 11, 177–80.

Pires, G., Castelo-Branco, M., Nunes, U., 2008. Visual

P300-based BCI to steer a wheelchair: a Bayesian

approach. In: Conference Proceedings : ... Annual

International Conference of the IEEE Engineering in

Medicine and Biology Society. IEEE Engineering in

Medicine and Biology Society. Conference. pp. 658–

61.

Rawlik, K., Toussaint, M., S, V., 2012. On Stochastic

Optimal Control and Reinforcement Learning by

Approximate Inference. In: Robotics: Science and

Systems Conference.

Repovs, G., 2010. Dealing with Noise in EEG Recording

and data analysis. Inform. Medica Slov. 15, 18–25.

Saal, H.P., Vijayakumar, S., 2010. Active estimation of

object dynamics parameters with tactile sensors. In:

2010 IEEE/RSJ Int. Conf. Intell. Robot. Syst. IEEE,

pp. 916–921.

Sahu, S., Biswal, B.B., Subudhi, B., 2008. Computational

Intelligence , Control and Automation A Novel

Method for Representing Robot Kinematics using

Quaternion Theory A Novel Method for Representing

Robot Kinematics using Quaternion Theory. IEEE

Spons. Conf. Comput. Intell. Control Comput. Vis.

Robot. Autom.

Schlögl, A., Slater, M., Pfurtscheller, G., 2002. Presence

research and EEG Properties of EEG recordings.

Shepherd, G.M., Mirsky, J.S., Healy, M.D., Singer, M.S.,

Skoufos, E., Hines, M.S., Nadkarni, P.M., Miller, P.L.,

1998. The Human Brain Project: neuroinformatics

tools for integrating, searching and modeling

multidisciplinary neuroscience data. Trends Neurosci.

21, 460–468.

Srinivasan, R., 1999. Methods to Improve the Spatial

Resolution of EEG. Int. J. Bioelectromagn. 1, 102–

111.

Tanaka, K., Matsunaga, K., Wang, H.O., 2005.

Electroencephalogram-Based Control of an Electric

Wheelchair. IEEE Trans. Robot. 21, 762–766.

Vespa, P.M., Nenov, V., Nuwer, M.R., 1999. Continuous

EEG monitoring in the intensive care unit: early

findings and clinical efficacy. J. Clin. Neurophysiol.

16, 1–13.

Vijayan, A., Nutakki, C., Medini, C., Singanamala, H.,

Nair, B., 2013. Classifying Movement Articulation for

Robotic Arms via Machine Learning. J. Intell.

Comput.

Westling, G., Johansson, R.S., 1984. Factors influencing

the force control during precision grip. Exp. brain Res.

53, 277–84.

Wolpaw, J.R., Birbaumer, N., McFarland, D.J.,

Pfurtscheller, G., Vaughan, T.M., 2002. Brain-

computer interfaces for communication and control.

Clin. Neurophysiol. 113, 767–91.

Wolpert, D.M., Flanagan, J.R., 2010. Q&A: Robotics as a

tool to understand the brain. BMC Biol. 8, 92.

Yuanqing Li, Chuanchu Wang, Haihong Zhang, Cuntai

Guan, 2008. An EEG-based BCI system for 2D cursor

control. In: 2008 IEEE International Joint Conference

on Neural Networks (IEEE World Congress on

Computational Intelligence). IEEE, pp. 2214–2219.

NeuralControlusingEEGasaBCITechniqueforLowCostProstheticArms

275