Prospects for high fidelity imaging in nonlinear EIT using high performance computing

Caton, M.J, Cox, S.J., Blott, BH, Daniell, G.J. and Nicole, D.A. (1999) Prospects for high fidelity imaging in nonlinear EIT using high performance computing. Holder, D. (ed.) pp. 1-6 .

Abstract

Linear image reconstruction in EIT has been widely used in medical and process applications, and nearly all of these have been based on approximately uniform distributions of the electrical conductivity in the object under study. Successful applications in medicine have been reported although the markedly non-uniform conductivity distribution in the human body, which is not known a priori, means that artefacts produced by the linear imaging approximation are difficult to exclude. It is important therefore to pursue the full nonlinear problem in order to provide a firm basis for subsequent linear treatments; we have already shown that effective nonlinear EIT can be achieved when appropriate choices are made to constrain the form of the images sought [Blott et al 1998]. In this paper we explore the image resolutions potentially achievable with the availability of improved signal-to-noise in the data. The advent of modern high performance computing allows the production of accurate reference images in practicable time scales, and these can then be used as the basis of high fidelity linear imaging systems in real-time applications. In addition, it becomes feasible to tackle the full 3-d problem.

This record has no associated files available for download.

Contributors

Download statistics