Measurement errors in multifrequency bioelectrical impedance analyzers with and without impedance electrode mismatch

Abstract

The purpose of this study is to compare measurement errors in two commercially available multi-frequency bioimpedance analyzers, a Xitron 4000B and an ImpediMed SFB7, including electrode impedance mismatch. The comparison was made using resistive electrical models and in ten human volunteers. We used three different electrical models simulating three different body segments: the right-side, leg and thorax. In the electrical models, we tested the effect of the capacitive coupling of the patient to ground and the skin-electrode impedance mismatch. Results showed that both sets of equipment are optimized for right-side measurements and for moderate skin-electrode impedance mismatch. In right-side measurements with mismatch electrode, 4000B is more accurate than SFB7. When an electrode impedance mismatch was simulated, errors increased in both bioimpedance analyzers and the effect of the mismatch in the voltage detection leads was greater than that in current injection leads. For segments with lower impedance as the leg and thorax, SFB7 is more accurate than 4000B and also shows less dependence on electrode mismatch. In both devices, impedance measurements were not significantly affected (p > 0.05) by the capacitive coupling to ground.

Skin-electrode contact area in electrical bioimpedance spectroscopy. Influence in total body composition assessment

“…In this study the obtained EBI spectra agree with the EBI spectra reported in [4] and obtained with electrode mismatch. The deviation observed at high frequencies indicates a kind-of inductive behavior present in the measurements obtained with a single electrode.…”

“…In EBI measurements the presence of electrode impedance mismatch can affect the impedance estimation as shown in [4]. In this study the obtained EBI spectra agree with the EBI spectra reported in [4] and obtained with electrode mismatch.…”

“…The deviation observed at high frequencies indicates a kind-of inductive behavior present in the measurements obtained with a single electrode. Since the same effect is observed in the experiments reported in [4], it is very likely that the inductive effect is related with the presence of skin-electrode impedance mismatch. The inductive character seems to decrease when increasing area of the skin-electrode interface.…”

“…Variations in the quality of the skin electrode interface may affect the electrode polarization impedance (Z ep ) of one or more electrodes. This unbalance of the Z ep can create an electrode mismatch and as it has been reported by Bogónez-Franco in [4] affecting the measurement and producing changes in the impedance spectra, sometimes quite remarkable. In addition to this, the relatively small size of the contact area may contribute to an irregular distribution of the current that is injected in the limb.…”

Simple voltage-controlled current source for wideband electrical bioimpedance spectroscopy: circuit dependences and limitations

“…The results show that at low frequencies the frequency response of the output impedance is dominated by effects predicted by the model, while at high frequencies the determining factor is the unavoidable output parasitic capacitance. The validation has been done only with two structures, but any other current source topology will also suffer this effect and most commercial impedance analysers have high errors at high frequencies (Bogonez-Franco et al 2009). The load-in-the-loop structure has the advantage of being very simple and, with modern operational amplifiers, the output impedance frequency response that can be obtained is enough for bioimpedance applications and is limited at high frequencies by a physical cause that is shared with other structures.…”