Log in

Login to your account

Username *
Password *
Remember Me
From Brain Science to Intelligent Machines

Enhancing the performance of a practical BCI system


Date: Wednesday 14/3/2012
Venue: MS020
Time: 1.00 pm
Speaker: Mr. Vaibhav Gandhi
Affiliation: Brain-Computer Interfacing and Assistive Technology team
Intelligent Systems Engineering Centre (ISRC), University Of Ulster


Enhancing the performance of a practical BCI system

 

By Mr. Vaibhav Gandhi
Brain-Computer Interfacing and Assistive Technology team
Intelligent Systems Engineering Centre (ISRC), University Of Ulster

Abstract:

Brain-computer interface (BCI) is a direct communication between the brain and an external device, bypassing the traditional pathway of peripheral nerves and muscles. The aim of BCI is to supplement human capabilities by enabling people (particularly disabled) to communicate/control devices by simply “imagining”. There are still many hurdles before this technology can reach its full potential, beginning with the issues in acquiring the electroencephalogram (EEG) signals, unknown noise characteristics embedded within the EEG, lower accuracy and information transfer rate, wide variability amongst subjects as well as over time etc.
This presentation focuses on the introductory concepts, challenges and advancements in the field of the BCI system as undertaken within the Ph.D. framework. The work includes two major aspects; first an intelligent graphical user interface (GUI) design to enhance the overall communication bandwidth of the BCI and secondly, a quantum filtering based pre-processing of the EEG signals to improve the overall system accuracy.
An adaptive GUI has been designed that utilizes the information from the sonar sensors attached to the controlled end device (robot). This enables the BCI user to have a quicker selection of choice thereby improving the time to complete a specific mission/goal. The nature of the human mind is non-classical and prompts the question whether the human mind and the mental processes can be considered as inherently quantum in nature. This led to the designing of a novel de-noising approach using the Schrodinger wave equation (SWE) for enhancement of the raw EEG signal. The raw EEG signal obtained during the motor imagery (i.e., mental imagination) performed by a BCI user is intrinsically embedded with noise while the actual noiseless signal is still unattainable. The raw EEG at every computational sampling instant is encoded in terms of a wave packet using the Recurrent Quantum Neural Network framework. The wave packet can be interpreted as the probability density function of the signal at that instant and evolves as per the SWE to de-noise the raw EEG signal thereby enhancing the information and ultimately the BCI accuracy.


Short biography:
Mr. Vaibhav Gandhi is a Ph.D. student in the Brain-Computer Interfacing and Assistive Technology team of the Intelligent Systems Research Center, University of Ulster, Londonderry. He obtained his B.Eng. in Instrumentation & Control engineering and M.Eng. in Automatic Control & Robotics. After completion of his postgraduate degree, he worked in India first as a Lecturer and later as an Asst. Professor in Instrumentation & Control engineering. After six years of service, he joined the University of Ulster as a Ph.D. student. Vaibhav is a recipient of the UKIERI scholarship for his Ph.D. research and has worked as a Sr. Research Asst. at IIT Kanpur under the UKIERI project. His research interests are in electrophysiological signal processing, neural networks, image processing and robotics.

Challenges in simple and hybrid brain computer interfaces.

Date: Thursday 1/3/2012
Venue: MS020
Time: 1.00 pm
Speaker: Dr. Shahjahan Shahid
Affiliation: Intelligent Systems Engineering Centre (ISRC), University Of Ulster




Challenges in simple and hybrid brain computer interfaces.

 

By Dr. Shahjahan Shahid
Intelligent Systems Engineering Centre (ISRC), University Of Ulster


Abstract:


People who are totally paralysed (e.g., by amyotrophic lateral sclerosis (ALS) or brain-stem stroke) or have other severe movement disorders need alternative methods for communication to express their wishes and control. These individuals need an alternative communication channel that does not rely on the brain's normal output pathways of peripheral nerves and muscles. A brain-computer interface (BCI) is a communication system that can help these users to interact with the outside environment by translating brain signals into machine commands. Since over one decade, the use of electroencephalographic (EEG) signals has become the most common approach for BCI implementation. BCIs are normally categorized on the nature of EEGs (e.g., in terms of the stimulation features for EEG generation and the electrode placement area above brain/skull). These BCIs (here it is addressed as 'simple BCI') basically try to figure out the subject's voluntary intentions and decisions from the measurements of multichannel EEG. Advanced multivariate signal processing algorithms along with artificial intelligent techniques are mainly used in developing these interfaces. Recently, researchers introduce a new type of BCI (known as hybrid BCI) that comprises of multiple type of simple BCIs or a simple BCI and other interface driven by any electro-physiological signal (such as ECG, EMG, EOG, eye-tracker). The main intention of introducing this type of hybrid BCI was to get improved performance in interfacing system. However, in spite of extensive corpus effort for almost two decade, there still exist a number of not-well solved problems in BCI designing these include subjects inability to produce effective EEG signals, signal recoding techniques, nonlinear characteristics of EEG, slow information transfer rate and signal processing techniques.
This presentation will cover the basic principle of simple and hybrid BCI, useful signal processing techniques for BCIs design. A discussion will follow on the challenges in designing these BCIs for rehabilitation..


Short biography:

Dr Shahjahan Shahid is a Research Associate in the Cognitive Robotics Group of Intelligent System Research Centre at Londonderry. He obtained his BSc. And MSc in Applied Physics and Electronics. After completion of his postgraduate degree, he worked for Bangladesh Atomic Energy Commission where he was responsible to repair-maintenance and up-gradation of Nucleonic equipments. After nine years of service, he returned to university for higher studies and completed his PhD in 2004 from University of Limerick, Ireland. His research interests are in electrophysiological signal processing, higher order statistics techniques for time series analysis. Dr. Shahid worked at many laboratories in Singapore, Malaysia Hungary, Austria and UK.