Efficient Electrical Impedance Tomography Software Based on the Dual Reciprocity Boundary Element Methods

这个小型企业创新第一阶段研究项目将开发高效的计算算法，使医学和工业中的二维(2-D)和三维(3-D)应用能够使用电阻抗断层成像(EIT)进行精确成像。EIT试图通过在物体表面施加特定的电流并测量那里感应的电压来确定物体内部的电导率分布。与其他成像方法相比，如果成功，EIT在经济、方便和安全方面为广泛的临床和工业应用提供了显著的优势。虽然在实验EIT技术和EIT数学模型方面已经取得了很大进展，但它仍然局限于实验室情况和2-D成像对象。将EIT扩展到实际情况的主要绊脚石似乎是缺乏有效的算法来执行准确的数据反演。传统的EIT反演算法要么基于仅适用于二维环形几何的启发式技术，要么基于有限元的全数值解。在拟议的研究中，通过使用使用双互易技术的新算法，将计算效率更高的边界元方法(BEM)扩展到EIT。这些算法将允许在复杂的2-D和3-D几何图形中进行有效的成像，并将显著减少问题大小，避免繁琐的网格/网格生成，并提供数量级的加速。第一阶段研究将开发利用双互易边界元反演EIT数据的数值算法和软件。开发的软件将在分析生成的正问题数据上进行测试。第二阶段将在第一阶段开发算法，使用并进一步完善第一阶段算法，开发一个原型EIT系统。第三阶段将把开发的系统和软件商业化。

This Small Business Innovation Phase I Research project will develop efficient computational algorithms that will enable accurate imaging using Electrical Impedance Tomography (EIT) for both two-dimensional (2-D) and three-dimensional (3-D) applications in medicine and industry. EIT attempts to determine the distribution of electrical conductivity inside objects by applying specified currents at their surface, and measuring the voltage induced there. In comparison to other imaging methods, if successful, EIT offers significant advantages in economy, convenience, and safety for a wide variety of clinical and industrial applications. While much progress in experimental EIT techniques and in the mathematical modeling of EIT have been made, it remains restricted to laboratory situations and imaging objects in 2-D. The chief stumbling block to extending EIT to practical situations appears to be the lack of efficient algorithms that can perform accurate inversion of data. Conventional inversion algorithms in EIT are based on either heuristic techniques that are only applicable to 2-D ring-like geometries, or on full numerical solution using Finite Elements. In the proposed research the much more computationally efficient Boundary Element Method (BEM) will be extended to EIT by using new algorithms that use dual reciprocity techniques. These algorithms will allow efficient imaging in complicated 2-D and 3-D geometries and will significantly reduce problem size, avoid cumbersome grid/mesh generation, and offer order of magnitude speed-up. Phase I research will develop the numerical algorithms and software for inverting EIT data using Dual Reciprocity BEM. The software developed will be tested on analytically generated forward problem data. Phase II will develop algorithms in the develop a prototype EIT system using and further refining the Phase I algorithm. Phase III will commercialize the developed system and software.