This dissertation asserts that the visualisation of electroencephalographic (EEG) data a personal computer would provide a low cost tool for examining brain function. It is proposed that visualisation, coupled with an effective mapping algorithm, will prove to be an important aid in the interpretation of polygraph recordings. Based on a comprehensive survey of the literature, this study isolates the most feasible course for implementing a useful three dimensional display of EEG data on a personal computer. It is suggested that volumetric data calculated using the Cortical Imaging Technique and displayed in a symbolic representation composed of pregenerated pixel bitmaps presents the most practical solution at this time. The Cortical Imaging Technique has been previously shown to extract detail not apparent in surface maps and is readily adapted to yield volumetric information. Although the significance of CIT data remains to be ratified, the technique appears to yield better three dimensional data than other algorithms for mapping EEG data. The detailed rendering of potential fields is found to be unacceptably slow for interactive visualisation in the hardware environment tested. Significant improvement in render time is possible by using precalculated bitmaps of shaded spheres. The spheres convey an adequate impression of three dimensions, allowing the user to assess spatial relationships in the image while improving update time to limits acceptable for user interaction. This speed improvement is achieved at the expense of image detail. It is proposed that a clinical study will quantify the usefulness of EEG data visualisation and will dictate what improvements to pursue with the advent of more powerful hardware.
|Degree Type||Masters degree|