Certain Investigation on P300 Extraction and Classification from Single Trial EEG Towards Portable BCI Applications

Abstract

Augmentative and Alternative Communication (AAC) devices have been developed to assist people who are unable to communicate through speech, hand gestures or eye gaze. AAC devices such as joy stick, communication boards, etc. enhance communication in speech impaired people and thus, help people to interact with the environment. But, people who are undergoing stroke rehabilitation or having neuromuscular disorder such as Amyotrophic Lateral Sclerosis (ALS), and Guillian Barre Syndrome (GBS) express difficulty and discomfort in using these devices, as their limb muscles become paralysed. This condition referred to as Locked-In-Syndrome (LIS), is where people cannot use their normal muscular pathway for communication though their cognition abilities remain intact. So they communicate either through eye movements or eye blinks which sometimes lead to misinterpretation. In those situations, Brain Computer Interface (BCI) provides an alternative and promising solution by addressing the above mentioned limitations of AAC devices, to improve the quality of life in LIS patients. BCI refers to the interface that manipulates an external device using brain signals or vice versa. The electrical activity of the brain is monitored and analysed for certain features or patterns which are leveraged to manipulate or control a BCI system. P300 is an endogenous potential generated in response to a rare or infrequent visual or auditory stimulus and is referred to as Event Related Potential (ERP). The challenge lies in the interpretation of this underlying information that presents itself as specific patterns in the scalp EEG signal. Thus, one of the translational research areas of BCI, is the development of various signal processing approaches to extract this significant information from EEG signaThe objective of the proposed research work is to develop and implement signal processing methods that eliminate artefacts from EEG signal and improve detection of endogenous ERP-P300 with better accuracy.