Fabrizi L, McEwan A, Woo E, Holder D S
Department of Medical Physics and Bioengineering, UCL, London, UK.
Physiol Meas. 2007 Jul;28(7):S217-36. doi: 10.1088/0967-3334/28/7/S16. Epub 2007 Jun 26.
Electrical impedance tomography measurements in clinical applications are limited by an undesired noise component. We have investigated the noise in three systems suitable for imaging epileptic seizures, the UCH Mark1b, UCH Mark2.5 and KHU Mark1 16 channel, at applied frequencies in three steps from 1 to 100 kHz, by varying load impedance, single terminal or multiplexed measurements, and in test objects of increasing complexity from a resistor to a saline filled tank and human volunteer. The noise was white, and increased from about 0.03% rms on the resistor to 0.08% on the human; it increased with load but was independent of use of the multiplexer. The KHU Mark1 delivered the best performance with noise spectra of about 0.02%, which could be further reduced by averaging to a level where reliable imaging of changes of about 0.1% estimated during epileptic seizures appears plausible.
临床应用中的电阻抗断层成像测量受到不期望的噪声成分的限制。我们研究了三种适用于癫痫发作成像的系统中的噪声,即UCH Mark1b、UCH Mark2.5和KHU Mark1 16通道系统,在1至100 kHz的三个应用频率步骤中,通过改变负载阻抗、单端或多路复用测量,以及在从电阻器到充满盐水的水箱和人类志愿者的复杂度不断增加的测试对象中进行研究。噪声是白色的,从电阻器上约0.03%的均方根值增加到人体上的0.08%;它随负载增加,但与多路复用器的使用无关。KHU Mark1表现最佳,噪声谱约为0.02%,通过平均可进一步降低到一个水平,在该水平下,对癫痫发作期间估计的约0.1%的变化进行可靠成像似乎是可行的。