Affiliations: [a] Key Research Centre in Health Technologies, University of Technology, Sydney, Broadway, NSW, Australia | [b] Rehabilitation Research Unit, Northern Clinical School, Faculty of Medicine, The University of Sydney, NSW, Australia | [c] The Brain Resource Company, Ultimo, NSW, Australia
Correspondence:
[*]
Address for correspondence: Professor Ashley Craig, Rehabilitation Studies Unit, Northern Clinical School, Faculty of Medicine, The University of Sydney, Ryde, PO Box 6, NSW 1680, Australia. E-mail: [email protected].
Abstract: Research and development has been conducted on a “hands-free” brain wave activity (electroencephalography-based) assistive technology device called the Mind Switch. It has been designed to provide a severely disabled person (such as someone with tetraplegia) with the ability to switch using their brain activity and thus activate devices in the environment. An area of concern has been the time it takes to complete set-up of the system before individual use, as well as the occurrence of errors in switching rates. In order to improve the user friendliness of the system, this study compared the efficacy of detecting electroencephalography (EEG) changes in brain signals that occur with eye closure and used as a “switch”, with two different methods of processing the EEG. Using data from spinal cord injured participants, fractal analysis, that is, the change in fractal dimension of the EEG wave during eye closure was compared to the original method of using spectral analysis change. In comparison to the spectral technique, the fractal dimension technique improved set up time, and was substantially improved in reducing the number of false positive and false negative errors when activating the device. Furthermore, the time to switch was also marginally reduced when using fractal dimension processing. Implications for improving the viability of an EEG based environmental control system for the severely neurologically injured are discussed.
Keywords: Spinal cord injury, assistive technology, electroencephalogram (EEG), environmental control system (ECS), fractal dimension, spectral analysis
