University of Twente Student Theses


Mirror therapy in virtual reality by a brain-computer interface for amputees experiencing phantom limb pain

Waardenburg, F.H. (2021) Mirror therapy in virtual reality by a brain-computer interface for amputees experiencing phantom limb pain.

[img] PDF
Abstract:Although there are different therapies to relieve Phantom Limb Pain (PLP) in amputees, these do not work for or are not available to every amputee. Mirror Therapy (MT) is an effective therapy that is inaccessible for bilateral amputees. Therefore the goal of this graduation project is to design, implement and evaluate a Virtual Reality (VR) system, used for rehabilitation through MT, that is inclusive for people experiencing PLP after amputation. By using literature, brainstorming and analytical assessment the implementation was designed and created. This resulted in a modification of MT and bringing it into the virtual world. Therefore the mirror is replaced by a VR environment, where the user views themselves through the body of an avatar. The operating method is based on Motor Imagery (MI), imagining movement, in particular pointing. In short, the implementation uses Functional Near-Infrared Spectroscopy (fNIRS) to measure brain signals, MATLAB to classify these brain signals and VR glasses to show the output in VR. An experiment (N=10) with two groups was used to evaluate the implementation. In the experiment, the fNIRS based Brain-Computer Interface was used to make the avatar’s arm point in Virtual Reality (VR). The fNIRS group (N=5) saw the feedback of the avatar pointing in VR based on their own oxygenated haemoglobin (O2Hb) levels. Whereas the control group (N=5) saw the avatar point automatically based on a time interval. The classification of the BCI was calculated by using the mean of O2Hb of the baseline (a clear/empty state of mind) and comparing this to the O2Hb value in the task phase (imagining to point). If this result exceeds the threshold the avatar started pointing. The fNIRS group had an activation rate, the times that the avatar pointed, of 71%. Whereas, the control group experienced, where the data were classified after the experiment, an activation rate of 60%. In addition, the immersion and activation showed a strong correlation. This indicates that if the avatar pointed more often (higher activation rate), the immersion experienced by the participant is higher. This work shows that using an fNIRS BCI to operate VR is a promising step towards making MT inclusive. In addition, this thesis is a stepping stone towards applications combining BCIs and VR. Furthermore, an interesting relation between the immersion and activation rate has been found. For future development, the classification can be improved by using additional filters and assessing the signal quality by the Signal Quality Index (SQI) algorithm
Item Type:Essay (Bachelor)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:30 exact sciences in general, 42 biology, 44 medicine
Programme:Creative Technology BSc (50447)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page