Research on Hihg-Performance Electrical Impedance Tomography

电阻抗断层成像（EIT）能够可视化与活体中的解剖学病症或生理活动相关的电阻抗分布及其变化，以应用于监测诸如脑中的血流变化的生物状况，本研究发展了神经网络方法来解决电导率分布估计的反问题通过一组密集放置在体表的电极测量表面电位分布。为了提高电导率分布的分辨率，当我们想要识别脑中的局部血流增量时，基于MR的反卷积将是有效的;在这种情况下，通过应用动态正则化技术可以很好地改善逆问题的不适定性。我们还开发了一种遗传算法来选择MR分割的最佳指标，这是在自动分割脑的CSF、灰质和白质之后将MR数据并入去卷积网络的关键技术。另一个主题，我们研究的是使用磁控四极方法来监测脑中的脉动阻抗变化：我们对该系统进行了实验测试，以证明该方法在估计心理负荷测试中的血液辉光增量方面的有效性。我们现在计划通过对在磁场连续移位的测量中获得的一系列数据进行去卷积来提高磁控四极法的分辨率。

Electrical impedance tomography (EIT) is capable of visualizing the electrical impedance distribution and its changes related to the anatomical disorders or physiological activities in the living body, to be applied for monitoring the biological conditions such as blood-flow changes in the brain, swellings and urinary bladder content etc.In this research we developed neural network methods to solve the inverse problems in estimating the conductivity distribution from the superficial potential distribution measured via a set of electrodes densely placed on the body surface. For improving the resolution of the conductivity profile, MR-based deconvolution will be effective used when we want to identify the localized blood-flow increment in the brain ; in this case ill-posedness of the inverse problem can be well improved by applying the dynamic regularization technique. We also developed a GA method to select the best indices for MR segmentation, which is a key-technology for incorporating the MR data to the deconvolution network after segmenting the CSF,the gray-matter and the white-matter of the brain automatically.Another topic, we investigated is the use of magnetically controlled tetrapolar method to monitor the pulsatile impedance changes in the brain : we experimentally tested the system to prove the effectiveness of the method in estimating the blood-glow increment during the mental-loading tests.We are now planning to improve the resolution of magnetically controlled tetrapolar method by deconvolving the series of data obtained in the measurement with the magnetic field continuously shifted.