Abstract-Positive phase angle is documented and analyzed in a three electrode monopolar needle measurement. Inductance equivalent behavior of the stray capacitance to ground is described as error source in a non-inductive sample measurement.
I. INTRODUCTIONDuring development of new bioimpedance based applications (e.g. as described in Kalvøy [1]), we have had unexplained detections of positive phase angle in some of our measurements. These have been both for in-vitro laboratory setups and in clinical in-vivo experiments. Some of this behavior has been explained for 4-electrode setups by Grimnes and Martinsen [7]), but positive phase angle have also occurred in our monopolar 3-electrode Solartron® 1260/1294 setups. This has typically been at frequencies between 100 kHz and 1 MHz. The magnitude has been higher than expected from the inductive properties of tissue and it seems not to be fully explained by Grimnes and Martinsen, nor have we been able to avoid the positive phase by different optimizations of our 3-electrode setup. However, Aliau-Bonet and Pallas-Areny[4] addressed stray capacitance to ground as a contributor to apparent inductive behavior in 4-electrode impedance measurements. A manipulation of their 4-electrode equations showed that similar phenomenon may also occur in other setups, and in our hospital environment we have a lot of electronic equipment and large conductive surfaces representing possible stray capacitances between a human body surface and ground.From this we postulate that stray capacitance can partly explain the positive phase angle in our 3-electrode Solartron® 1260/1294 measurement setup. In this paper we have tested this hypothesis by evaluating the theory of AliauBonet and Pallas-Areny on an equivalent circuit modified for 3-electrode setups. Laboratory measurements were also performed to confirm the results from the theoretical analysis. The results have been used to propose strategies to avoid or minimize possible errors.H. Kalvøy is with the Clinical and Biomedical engineering department at Rikshospitalet, Oslo University Hospital, Oslo, Norway. (corresponding author; phone: +47 23 07 15; e-mail: havard.kalvoy@ous-hf.no).C. Aliau-Bonet is with Universitat Politècnica de Catalunya, BarcelonaTech (UPC), Castelldefels, Spain, (e-mail: carles.aliau@upc.edu).R. Pallas-Areny is with Universitat Politècnica de Catalunya, BarcelonaTech (UPC) Castelldefels, Spain (e-mail: ramon.pallas@upc.edu).Ø.G. Martinsen is with the Department of Physics at the University of Oslo, Norway and the Clinical and Biomedical engineering department at Rikshospitalet, Oslo University Hospital, Oslo, Norway (e-mail: ogm@fys.uio).
II. MATERIALS & METHODS
A. Three electrodes bioimpedance measurementA large piece of bacon was used as sample. The type of biomaterial used is not critical for this study, and bacon has earlier been suggested as a convenient in vitro model for animal or human tissue [2]. A three electrode measurement was implemented by a 1294 impedance interface connected to a 1260 frequency respon...