Implementation of Magnetic Induction Tomography image reconstruction algorithms on Clearspeed Advance co-processor.

磁感应层析成像（MIT，也称为电磁感应层析成像- EMT）是一种新的非接触技术，用于成像物体横截面内的电特性分布。这项技术仍处于开发阶段，到目前为止，市场上还没有MIT商业系统。它具有提供便携式、低成本和非侵入性图像的潜力，可适用于医疗应用，例如在紧急情况下对脑中风等行为进行快速医疗评估。近年来，我们和其他人已经成功地开发了图像重建算法。这些算法的计算量非常大，并且只能在实际时间限制内重建低分辨率图像。在这个项目中，我们建议将我们现有的代码（用FORTRAN编写）移植到一个多加速器板平台（Clearspeed）上，以允许快速计算图像重建并允许更高的分辨率，使MIT更适合医疗应用。新实施的系统将成为麻省理工学院未来研究和商业开发的基础。

Magnetic induction tomography (MIT, also known as electromagnetic induction tomography - EMT) is a new non-contact technique for imaging the distribution of electrical properties within a cross-section of an object.This technique is still at the development stage and to date, there are no MIT commercial systems in the market. It has the potential of providing a portable, low cost, and non invasive imagery which could be suitable for medical applications such as rapid medical assessment of conductions such as cerebral strokes in emergency situations. Algorithms for image reconstruction have successfully been developed in recent years by our group and others. These algorithms are very intensive computationally, and only low resolution images can be reconstructed within practical time limits. In this project we propose to port our existing code (written in FORTRAN) to a multi accelerator board platform (Clearspeed) to allow fast computation of image reconstruction and allow higher resolutions making MIT more suitable for medical applications. The new implemented system will be the basis for future research and commercial development in MIT.