*** 9531798 Duraiswami This Small Business Innovation Phase II Research project will develop practical, efficient, fast and accurate algorithms and software for imaging using Electrical Impedance Tomography (EIT) for both two and three-dimensional applications in medicine and industry. EIT images the inside of an object by determining the spatial distribution of impedance. This is achieved by applying specified currents at their surface, and measuring the voltage induced. Compared to other imaging methods promises significant advantages in economy, convenience, and safety for a wide variety of clinical and industrial applications. The main obstacle to making EIT more widely applicable is the lack of an efficient algorithm that can perform accurate data inversion. Inversion algorithms currently in use are based on either heuristic techniques of limited applicability, or on full numerical solution using finite elements. In Phase I of this research the computationally efficient Boundary Element Method (BEM) was extended to EIT by using new algorithms that use dual reciprocity techniques. Further improvements in efficiency and accuracy were obtained by developing a new methodology for parametrizing the inverse problem. Phase I results demonstrated the speed and accuracy advantages of EIT images obtained using the new algorithms. Phase II research will further improve these algorithms. They will be specialized to particular applications, incorporate more efficient ways to perform the multiple parameter minimization, and account for available a priori information. The developed algorithms will be validated and refined by performing experiments. The software developed will be integrated into a user-friendly package BEM software for Electrical Impedance Tomography will have a very large commercial market in medical imaging, two-phase flow imaging, and industrial nondestructive evaluation and testing. The developed software will be licensed to manufacturers of EIT systems, while we w ill develop EIT based multiphase flow instrumentation. ***
