Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.633652
Title: Musical spacialised auditory feedback for a brain-computer interface
Author: McCreadie, Karl A.
Awarding Body: University of Ulster
Current Institution: Ulster University
Date of Award: 2013
Availability of Full Text:
Access through EThOS:
Abstract:
A brain-computer interface (BCI) offers an alternative method of communication and control for those who have suffered neuromuscular damage due to disease or injury. Imagined movement can be used to modulate the brain activity of an individual which may then be translated into a control signal which is regulated using sensory feedback. However, this feedback is predominantly presented visually and excludes those with a disability coupled with vision problems, a typical BCI beneficiary target group. Presenting feedback using an alternative sensory pathway, such as hearing, can allow for a more inclusive method of control. Additionally, the diagnosis of those with disorders of consciousness (DOC) can be problematic due to the lack of overt motor responses and inherent ocular instabilities. Moreover, auditory BCI allows the possibility of freeing up the visual channel in able-bodied users permitting other tasks. Although audio based BCI systems exist, variations in the methods of auditory feedback presentation have not been fully investigated. There is a clear need for an auditory motor imagery based BCI which has the potential not only to offer an alternative communication channel for the disabled, but also an additional method of assessing and training those with a DOC. Firstly, the most suitable form of auditory feedback is investigated before a between group comparison study examines the differences in performance comparing visual and auditory able-bodied feedback groups (N=20) when presented with stereophonic broadband noise. Prior to this research stereophonic auditory sensorimotor feedback with broadband noise had not been tested. Secondly, performance of able-bodied participants is assessed in a within group study (N=7) where the amount of spatial information is varied. Additionally, the development of an auditory asteroids avoidance game is detailed and the performance compared both with a visual equivalent and with other auditory presentation methods. All studies to this point present audio through loudspeakers, but a subsequent study including two patients with a DOC, shows results when presenting musical feedback through earphones, which is paralleled with results obtained using broadband noise. Subsequently a modified earphone presentation method including spatial cues is used to present musical feedback to two able-bodied participants to assess performance. Results show that stereo pink noise is a suitable replacement for the visual equivalent, that there is no statistical difference in performance when spatial information is varied, and that there is a correlation between certain psychological, factors and performance. This has implications not only for those previously mentioned, but also for those with conditions resulting in a reduction in information from the traditional pathways allowing for a possible sensory extension. Finally, suggestions are made to other possibly fruitful research directions, including the need for user specific music libraries which have shown to enhance engagement and could improve BCI performance in future developments, the potential for crowd sourced datasets, the prospect for an eyes-closed system, and highlights the need for a greater focus on training the mind rather than just training the system.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.633652  DOI: Not available
Share: