Evidence suggests that cerebral blood flow patterns accompanying a mental activity are retained in many locked-in patients. Thus, real-time monitoring with functional transcranial Doppler (TCD) together with a specific mental task could control a brain-computer interface (BCI), thereby providing self-initiated interaction.
The objective of this study was to create an automatic detection algorithm to differentiate hemodynamic responses coincident with one's performance of verbal fluency (VF) versus counting tasks.
We recruited 10 healthy adults who each silently performed up to 30 VF tasks and counted between each. Both middle cerebral arteries were simultaneously imaged using TCD. Linear Discriminant Analyses (LDA) successfully differentiated between VF and both prior and post counting tasks. For every participant, LDA achieved the 70% classification accuracy sufficient for BCIs. Results demonstrate automatic detection of a VF task by TCD and warrant further investigation of TCD as a BCI.
Identifer | oai:union.ndltd.org:TORONTO/oai:tspace.library.utoronto.ca:1807/30587 |
Date | 07 December 2011 |
Creators | Faulkner, Hayley |
Contributors | Chau, Tom |
Source Sets | University of Toronto |
Language | en_ca |
Detected Language | English |
Type | Thesis |
Page generated in 0.002 seconds