Hook artifact correction of localized electrical bioimpedance for improved agreement between different device measurements
Hook artifact correction of localized electrical bioimpedance for improved agreement between different device measurements
复制标题
DOI：
10.1088/2057-1976/aa971b
复制
发表时间：
2018-01-01
影响因子：
1.4
通讯作者：
Critcher, Shelby
中科院分区：
文献类型：
--
作者：
Freeborn, Todd J.;Crenshaw, Tim;Critcher, Shelby
理想情况下，生物阻抗测量仅捕获所研究的组织，但实际上包含其他贡献，包括测量设置中的寄生电容。这些寄生效应可以作为钩状伪影观察到，即高频下电抗不断增加。这种伪影会妨碍组织的准确高频表征以及从不同测试配置收集的结果的比较。目的：评估从表现出和不表现出钩状伪影的同一组织收集的实验局部电阻抗数据集的钩状伪影校正方法。方法：通过两台设备从研究参与者的前臂收集七组测量结果，其中一台设备在所有测量中都显示出钩子伪影。通过应用非线性最小二乘优化例程来确定数据的等效电路参数并从等效表示中去除寄生分量，对钩子伪影数据集进行了校正。主要结果：除去评估的寄生电容，生物阻抗的校正等效电路表示显示电阻和电抗与收集的未表现出钩伪影的相同组织的测量值的一致性显着改善。这种改进对于 > 100 kHz 的频率最为显着，校正测量值与没有钩伪影的实验测量值之间的偏差降低了一个数量级。意义：这些结果进一步验证了实验生物阻抗数据中的高频钩伪影可以建模为与所研究的组织的并联电容。使用该模型的校正技术可以从阻抗数据的等效电路表示中消除并联电容的影响，从而提高从不同设备收集的测量结果的一致性。该过程可用于改善组织的表征以及使用不同测试配置收集的测量结果之间的比较。
Ideally, bioimpedance measures capture only the tissue under investigation but in practice contain other contributions including parasitic capacitances in the measurement setup. These parasitics can be observed as a hook artifact, an increasing reactance at high frequency. This artifact can prevent accurate high frequency characterization of a tissue and comparison of results collected from different test configurations. Objective: To evaluate a hook artifact correction method on experimental localized electrical impedance datasets collected from the same tissue that do and do not exhibit the hook artifact. Approach: Seven sets of measurements were collected by two devices from the forearm of a study participant, with one device exhibiting the hook artifact for all measurements. The hook artifact datasets were corrected by applying a nonlinear least squares optimization routine to determine the equivalent circuit parameters of the data and remove the parasitic component from that equivalent representation. Main results: Removing the assessed parasitic capacitance, the corrected equivalent circuit representation of the bioimpedance showed a significant improvement in agreement of both resistance and reactance with the collected measurements of the same tissue that did not exhibit the hook artifact. This improvement was most significant for frequencies > 100 kHz, with an order of magnitude decrease in the deviation between the corrected and the experimental measurements without the hook artifact. Significance: These results further validate that the high frequency hook artifact in experimental bioimpedance data can be modeled as a parallel capacitance with the tissue under investigation. And that correction techniques using this model to remove the contribution of the parallel capacitance from equivalent circuit representations of the impedance data improve the agreement of measurements collected from different devices. This process can be applied to improve characterization of a tissue and comparisons between measurements collected with different test configurations.