Detection of Electroencephalography Artefacts using Low Fidelity Equipment

Abstract

The use of Electroencephalography (EEG) signals in the field of Brain Computer Interface (BCI) has gained prominence over the past decade, with the availability of diverse applications especially in the clinical sector. The major downside is that the current equipment being used at medical level is specialized, complex and very expensive. Our research goals are to further increase accessibility to this technology by providing a unique approach in data analysis techniques, which in return will allow the usage of cheaper and simpler EEG hardware devices targeted for end users. We use non-invasive BCIs designed on EEG, mainly due to its fine temporal resolution, portability and ease of use. The main shortcoming of EEG is that it is frequently contaminated by various artefacts. In this paper we provide vital groundwork by identifying and categorizing artefacts using low fidelity equipment. This work forms part of a wider project in which we attempt to use those artefacts constructively, when others try to filter them out. The main contribution is to create awareness of the extent to which artefacts can be encountered, identified and categorized using offthe shelf equipment. Our results illustrate that we are able to adequately identify and categorize the most commonly encountered artefacts in a non-clinical environment, using low fidelity equipment.