Using bio impedance to measure and analyse tissue data
With correct instrumentation, bioimpedance theory can be used to measure and discriminate between different biological materials, including human tissue. Some examples of bioimpedance in use in medical applications are imaging of newborns lungs, measuring respiration rate and measuring sweat activity.
Published: 30. December 2014
By: Patrick Hisni Brataas, Development Engineer, Data Respons
A non-medical application can be a fingerprint sensor where it is possible to measure the difference between living and dead tissue. This can in some cases be important as an increased security measure to make sure the finger is still attached to the person it belongs to or to ensure that it is not some kind of copy. In this article, we shortly explain how we can use a needle as the measurement tool to distinguish between e.g. fat and muscle tissue or determine the quality of a piece of meat. This article will highlight the most important bioimpedance theories, how bioimpedance can be measured and how to analyze the data.
What is bioimpedance?
Bioimpedance is the term used when referring to the electrical impedance of a biological material. Bioimpedance describes the passive electrical properties, which in an electrical perspective, differs in many ways compared to metals. The charge carriers in metals are freely moving electrons, while in biological materials it is the ions outside and inside the cells. When doing measurements, the transition between the electrons and ions is at the electrode. In addition, the cells have significant mass and therefore a current flow implies a movement of mass, which in itself results in changes in the biological material. The cell membrane separating the outside from the inside has a very low conductance and therefore acts as a dielectric and the whole cell acts as a capacitor. Because of this, biological materials have capacitive properties. At low frequencies (less than 10 kHz) the path outside the cells dominate, but at higher frequencies the cells capacitive properties gradually allows alternating current to pass through. A general trend is that the impedance decreases as the frequency increases. When the biological material is polarized by an external power source, ionic polarization occurs. This is the displacement, not the change in charge, of the positive ions with respect to the negative ions and does not occur instantaneously. The frequency of the applied signal is a critical factor of how much polarization impedance will be measured and depends on the time the ions need to change their position in the material. If the frequency is low enough, it means that all ions have enough time to change their position and the resulting polarization impedance will be maximal. This also means that polarization will decrease as the frequency increases. Dispersion is another important phenomenon and is the dependency between the permittivity and frequency in a polarized material. There are four known dispersion areas, and different biological materials exhibit a varying degree of dispersion. For example, muscle tissue exhibits a large beta-dispersion while in fat there is none. This can be used in algorithms to discriminate between different biological materials such as human tissue.
Electrodes and measurement setup
To be able to measure bioimpedance, an interface between the biological material and the electronics is needed. Electrodes provide this interface. At the electrode is where the shift in charge carriers occur, between the free flowing electrons in the metal to the ions in the biological material, and vice versa. In most measurement setups, electrode polarization impedance is measured in series with the measured bioimpedance. It is therefore important to understand how the measurements are influenced by this. In general, the electrode polarization impedance decreases as the frequency increases. Electrodes with a gel containing ions are often used. If the measuring electrode has a large metal-electrolyte interface area, it results in a low electrode polarization impedance. A large electrolyteskin interface area implies an averaging effect and a loss in spatial resolution, but with less noise in the measurements
A monopolar measurement means that the impedance contribution mainly comes from one of the electrodes. When using two or more electrodes one may achieve a monopolar measurements by making one of the electrodes dominant. In practice, this can be done in a two-electrode setup by increasing or decreasing the size of one of the electrodes relatively to the other. A three-electrode configuration is inherently slightly monopolar and can be further enhanced by changing the electrode sizes.
By using a relatively small measurement electrode, for example an insulated needle electrode with a small exposed area on the tip, a very monopolar measurement is achieved. If we also combine this with a three-electrode configuration, the measurement would be even more monopolar. It has been shown in published papers that approximately 97 % of the contribution originates from a few millimeters radius from the needle tip. This means that it is possible to accurately measure the bioimpedance at the tip of the needle when it is inserted into biological materials.
There are several ways to measure bioimpedance, of which only one method will be explained here. The explained method assumes a two-electrode or three-electrode configuration. The biological material is excited with a stable sinusoidal voltage signal with a known frequency and amplitude through the reference electrode, or reference and counter electrode for a three-electrode configuration.
The measurement needle electrode is inserted into the biological material and picks up the sinusoidal excitation current after it has propagated through the material. A voltage is excited upon the biological material and the current is measured, therefore we have the current-to-voltage ratio, which is the admittance or inverse of impedance. IQ demodulation can now be used to retrieve the real and imaginary components of the admittance and the impedance modulus and phase can be calculated using complex numbers theory. Different biological materials will alter the excitation signal in different ways at different frequencies. It might be necessary to calibrate the system depending on the implementation of the IQ demodulation, analog electronics and digital electronics.
On the designed device an ARM microcontroller was used with a minimum of analog components. A direct digital synthesizer (DDS) was used as the signal source to excite the biological tissue. The device is designed to be portable, powered by a standard lithium battery and uses Bluetooth for control and data transfer to a computer or tablet. The computer or tablet displays real-time measurement data. One graph shows impedance modulus and phase as a function of time at a single frequency, which can be set by the operator. Another graph shows the impedance modulus and phase as a function of frequency. A two-point phase calibration was performed to remove known linear phase errors. Calibration was performed by measuring a resistor, which was assumed to be an ideal resistor with no phase delay. Also the microcontrollers internal ADC was calibrated before use.
Interpretation of the measured data makes it possible to discriminate between different biological materials. The most evident method is comparing the impedance modulus and phase of the biological materials one wants to distinguish from each other, and determine a valid range for the values. This simple approach can work in many cases, but is susceptible to errors due to electrode impedance polarization, change in the material over time and other biological factors.
The next step is to do exactly the same, but compare at two or more different frequencies to get a better discrimination. This means it will be possible to discriminate between two biological materials that have similar characteristics at one frequency, but not at another. A step further is to use the knowledge about the dispersion regions. These regions in different biological materials can affect the measured values in a narrow frequency area, and be a very characteristic trait. By combining these parameters and knowledge about dispersion regions, it is possible to design smart algorithms that can discriminate between many different types of biological materials.
As an example, a measurement on a boiled egg was performed. By measuring the phase difference at two frequencies, 3kHz and 30kHz, it is possible to discriminate the egg white from the egg yolk, as well as accurately determine when the tip of the needle electrode is at the boundary in between. The algorithm for discriminating between the different parts of a boiled egg has used the phase parameter at two frequencies in the beta-dispersion region, which is strongly present in egg yolks, but not as much in egg whites. Areas of application for bioimpedance is yet to be discovered, but has a lot of potential.
Some of its medical applications are measuring respiration rate or sweat activity, lung imaging of newborns, diagnosis tool for skin disease and as a parameter for detecting restrictions in blood supply to organs. Some of the less medical and non-medical applications include body composition analysis, sometimes performed at your local gym, determining meat, wood or soil quality, monitoring the fermentation process in breweries and monitoring volcanic activity.
Some of its medical applications are measuring respiration rate or sweat activity. This could be used while conducting lie detector tests.
Examples of use of bioimpedance
Skin water content
Tissue ischemia monitoring
Find joint angle
Skin cancer detection
More information about the device can be found in the master thesis by Patrick Hisni Brataas  and general information about bioimpedance can be found in the “Bioimpedance and bioelectricity basics” book .  “Wireless Embedded Microcontroller System for Bioimpedance Measurements”, Master Thesis, Patrick Hisni Brataas, 2014. Link: http://urn.nb.no/ URN:NBN:no-45539  “Bioimpedance and Bioelectricity Basics”, 3rd Edition, Sverre Grimnes and Ørjan Grøttem Martinsen, Elsevier Academic Press, 2014.  “Needle Guidance in Clinical Applications based on Electrical Impedance”, Phd paper, Håvard Kalvøy