Electrical Impedance Spectroscopic Studies on Broiler Chicken Tissue Suitable for the Development of Practical Phantoms in Multifrequency EIT
Article Category: Articles
Data publikacji: 16 cze 2011
Zakres stron: 48 - 63
Otrzymano: 16 kwi 2011
DOI: https://doi.org/10.5617/jeb.174
Słowa kluczowe
© 2011 Tushar Kanti Bera, J. Nagaraju, published by Sciendo
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.
Practical phantoms [1, 2, 3, 4, 5, 6, 7, 8, 9, 10] are essential for assessing the system performance in Electrical Impedance Tomography (EIT) [11, 12, 13, 14, 15, 16] for its validation, calibration and comparison purposes. Practical phantom studies are also essential for assessing the EIT-instrumentation [17, 18] and EIT reconstruction algorithms [19, 20, 21]. Saline phantoms with insulator inhomogeneities have their own limitations as they could not mimic the physiological structure of real body tissue in several aspects. Saline or any other salt solutions are purely resistive materials and hence the multifrequency EIT systems [1, 22, 23, 24] cannot be studied properly with saline phantoms because the responses of the purely resistive materials do not change over frequency. On the other hand, electrical impedance of biological materials is a complex quantity [25, 26] which is a function of tissue composition as well as the frequency of the applied ac signal [27]. The Protein-Lipid-Protein structure [28] of the membrane of a biological cell gives a cell capacitance whereas the extracellular fluid, cytoplasm and nucleus contributes to the cell resistance [29]. As a result, animal tissues and plant tissues show a variable response over a wide band of signal frequency due to their complex physiological and physiochemical structures, and hence they can suitably be used as bathing medium [1, 2] and inhomogeneity in the phantoms of a multifrequency EIT system. A proper assessment of a multifrequency EIT system with a real tissue phantom needs a prior knowledge of the impedance profile of the bathing medium as well as the inhomogeneity. In this direction, bioelectrical impedance analysis (BIA) [29, 30, 31, 32] of broiler chicken muscle tissue paste and fat tissue is conducted by electrical impedance spectroscopy (EIS) [33, 34, 35, 36, 37, 38, 39, 40, 41, 42] and their impedance profiles are thoroughly studied. EIS of broiler chicken muscle tissue paste, chicken tissue blocks and chicken fat tissue is conducted using a constant current signal with a wide frequency range using impedance analyzers. Results show that the broiler chicken muscle tissue is less resistive than the fat tissue and hence it can be successfully used as the bathing medium of the phantoms for impedance imaging in multifrequency EIT. On the other hand, the fat tissue is found more resistive than the muscle tissue which makes it more suitable for the inhomogeneity in phantoms of impedance imaging study. Moreover, as there is a sufficient difference between the resistivities of muscle tissue and fat tissue they can be used as either inhomogeneity or background medium. EIS studies show that all the impedance parameters of the chicken tissue samples are functions of frequency and hence they are suitable for use as phantom materials in multifrequency EIT.
Electrical impedance spectroscopy is an electrical impedance analyzing technique in which the complex electrical impedance (Z) and its phase angle (θ) of an object under test are measured at different frequency points within a specified frequency band, from the measured potentials (V) developed by injecting a constant sinusoidal current signal (I) to the object, using a two-electrode or four-electrode method (Fig.-1a). In four-electrode technique (Fig.-1b) of impedance measurement, a constant alternating current is injected through two electrodes (called current electrodes: blue electrodes in Fig.- 1b) and the ac potential developed across two points of the body part under test is measured on other two electrodes (called voltage electrodes: red electrodes in Fig.-1b). Dividing the voltage measurement by the applied current, the complex transfer impedance is estimated.
Figure-1
(a) A schematic of electrical impedance measurement, (b) four electrode measurement of electrical transfer impedance.
On the contrary, in the two-electrode method the current injection and voltage measurement are made using the same electrodes and hence it is known as two-electrode technique. In EIS a constant current signal of variable frequency is applied to the object under test and the potentials developed across the measurement points at each frequency are measured. For a single frequency measurement, complex impedance (Z) is a function of material properties which create a phase displacement (θ) between I and V. In impedance spectroscopy, Z and θ are the functions of material properties at each frequency and hence they are the functions of frequency also. To analyze the EIS data, in present study, the real part of Z (R) and imaginary part of Z (X) at each frequency points are calculated and plotted (along x-axis and y-axis respectively) to obtain the Nyquist Plot (plot of imaginary part of Z versus real part of Z) or R-X spectrum. Any material can be modeled as a combination of electrical circuit elements (as each material has their own electrical properties like resistance (Re), capacitance (Ce), inductance (Le) etc.). Hence, the frequency responses of different circuit elements (resistor, inductor, capacitor etc.) or their combinations are different and hence their corresponding Nyquist plots are different from each other (Fig.-2a-f).
Figure-2
Nyquist plots for different circuit combinations: (a) R and L in parallel, (b) R and C in parallel, (c) r in series with (R and C in parallel), (d) r in series with (R and C in parallel) and (R and C in parallel), (e) R and CPE in parallel, (f) R1 in series with the parallel combinations of (R2 in series with W) and C.
Similarly, biological tissues can also be modeled as the combination of different circuit elements contributing to its complex bioelectrical impedance (Z). Electrical impedance of biological tissues consist of resistive and capacitive components (for low frequency the magnetic behavior is neglected [43, 44]), which are functions of signal frequency giving rise to a frequency dependent complex bioelectrical impedance defined as:
where Re(Z(ω)) = R represents the magnitude of the real part of complex Z and Im(Z(ω)) = X represents the magnitude of the imaginary part of the complex Z.
In biological cells, the cytoplasm, nucleus and extracellular fluids form a purely resistive path (with a good electrical conductivity) to the current signal [45]. On the other hand, the cell membranes (Fig.-3a) are composed of electrically non-conducting lipid bi-layers sandwiched between two conducting protein layers (Fig.-3a) and hence it provides a capacitance [45]. Cytoplasm and the nucleus are mostly made up of solutions of proteins, different chemicals and salts, which are electrically conducting materials. Hence intercellular fluid (ICF) containing the cytoplasm and nucleus offers a resistive path (RICF) to the electric signal (Fig.-1b). The cell membrane (CM) consists of a layer of non-conductive lipid material sandwiched between two layers of conductive protein molecules. Hence it behaves like a capacitance (CCM) contributing a capacitive reactance to an alternating current path (Fig.-1b). Being made up of conducting solution extracellular fluid (ECF) also creates a resistive path (RECF) towards the electric signal (Fig.-1b). Hence, the equivalent electrical circuit of an isolated animal cell consists of the parallel combination of extracellular resistance and the intracellular impedance where intracellular impedance is introduced due to the series combination of the membrane capacitance and the protoplasm (cytoplasm + nucleus) resistance (Fig.-1b). Biological tissue is made up of a three-dimensional array of cells and hence, depending on the tissue composition and structure, the bioimpedance profile changes from tissue to tissue. Moreover, as the frequency changes, the tissue impedance and its phase angle are varied with a corresponding change in its real (R) and imaginary (X) parts.
Figure-3
(a) animal cell membrane structure, (b) chicken muscle tissue paste in rectangular glass tank, (c) EIS study of chicken muscle tissue paste using QuadTech7600, (d) chicken muscle tissue block under EIS study using Agilent 4294A, (e) EIS study of chicken muscle tissue paste using Agilent 4294A.
In bioimpedance spectroscopy, a sinusoidal constant current (with varying frequency of particular range) is injected to the biological object under test through the current electrodes and the voltage developed across a particular domain is measured (Fig.-1b). The bioimpedance (Z) and its phase angle (θ) are calculated from the measured data (Fig.-1a). Z and θ are functions of body compositions as well as the signal frequency (ω = 2πf). The frequency responses of Z(ω), θ(ω), R(ω) and X(ω) and other parameters are studied. The reactive part of the complex bioelectrical impedance is generally of the capacitive type as the inductive response of biological tissues at low frequency is negligible [43, 44]. According to the tissue properties the magnitude of Z(ω), θ(ω), R(ω) and X(ω) are changed with a change in frequency and hence the nature and behavior of the biological tissues can be easily distinguished by plotting the impedance parameters with respect to the signal frequency. Electrical impedance spectroscopy of chicken muscle tissue paste, muscle tissue blocks and fat tissue is conducted and the variations of Z(ω), θ(ω), R(ω), X(ω), resistivity (ρ(ω)), conductivity (σ(ω)), C (ω), permittivity (ε) and relative permittivity (εr) over frequency are studied.
Broiler chicken muscle tissues and fat tissues are collected from the market and cut in small pieces after removing all the bones carefully. Muscle and fat tissues are separately rinsed three times with deionized (DI) water (processed and purified by Millipore Synergy UV Water Purification System with a high accuracy (output water resistivity = 18 ×106 Ωcm)). Cleaned and wet tissues are kept in two separate funnels covered with blotting papers and the external tissue water is filtered and extracted. All the tissues are kept open on Petri dishes for 30 minutes for local drying by natural evaporation of external water, if any. 2% deionized water is added to the chicken muscle pieces and the mixture is crushed for 2 minutes in a mixer grinder (No load speed: 2400 rpm). 1% DI water is further added to the mixture and rotated at the same speed for another 2 minutes.
An electrical impedance measurement set up is developed [21] with a rectangular glass tank (50 mm × 20 mm × 20 mm) containing two identical square electrodes (20 mm × 20 mm). Electrodes are identically cut from a 0.1 mm thick type 304 Stainless Steel (SS) sheet and machined to give all the electrodes an approximately equal surface area of 20 mm × 20 mm. High quality (Type 304) SS sheet is used to avoid the localized pitting corrosion leading to the creation of small holes on the electrode surface. The electrodes are connected by low resistive equal length flexible copper wires and the chicken muscle tissue paste obtained is transferred carefully in the rectangular glass tank (Fig.-3b). The tank is filled with the chicken tissue paste samples and the electrical impedance spectroscopy is conducted at room temperature with a current signal of 1 mA (10 Hz to 2 MHz) using a QuadTech 7600 impedance analyzer. Five chicken muscle tissue samples of identical volume (50 mm × 20 mm × 20 mm) are tested separately within a gap of equal and short time period (time required for a complete scan by QuadTech 7600). Similarly five fat tissue samples of identical geometry (5 mm × 20 mm × 15 mm) are tested separately within a gap of equal and short time period (time required for a complete scan by QuadTech 7600). The complex electrical impedances (ZMuscle and ZFat) and phase angles (θMuscle and θFat) are measured at each frequency using the Quadtech7600 impedance analyzer (Fig.-3c). Real and imaginary parts of the muscle tissue impedance (RMuscle and XMuscle respectively) and fat tissue impedance (RFat and XFat respectively) are calculated at each frequency. The variation of all the impedance parameters of chicken muscle tissue (ZMuscle, θMuscle, RMuscle and XMuscle respectively) and chicken fat tissue (ZFat, θFat, RFat and XFat respectively) are studied over frequency. Muscle tissue resistivity (ρMuscle), muscle tissue conductivity (σMuscle), fat tissue resistivity (ρFat) and fat tissue conductivity (σFat) are also calculated from the real parts of the tissue impedances and plotted over frequency.
For further analysis the EIS of chicken tissue blocks (15 mm × 20 mm × 20 mm) is conducted (Fig.-3d) and the dielectric dispersions associated with biological tissues in which the relative permittivity decreases with increasing frequency [46], are studied. In general three discrete regions of dispersion can be identified in biological tissues [46]. These are commonly defined as:
To study the α and β dispersions of the chicken muscle tissue blocks and muscle tissue paste samples, EIS is conducted from 40 Hz to 20 MHz using an Agilent4294A (Fig.-3e) with a 1 mA constant current signal. The variation of all the impedance parameters of chicken muscle tissue blocks, muscle tissue paste and fat tissue are studied over frequency. Muscle tissue resistivity, muscle tissue conductivity, fat tissue resistivity and fat tissue conductivity are also calculated from the real parts of the tissue impedances and plotted over frequency. Since the phase angle was low, at least above 100 Hz, conductivity was calculated as the inverse of resistivity. Tissue capacitance (CMuscle), permittivity (εMuscle) and relative permittivity (εrMuscle) are also calculated and α and β dispersions are studied. C, ε and εr are also calculated for fat tissue.
KCl solutions are prepared with different concentrations (0.00%, 0.05%, 0.10%, 0.15%, 0.20%, 0.25%, 0.30%, 0.35%, 0.40%, 0.50%, 1.00% and 2.00%) in w/v ratios dissolved in DI water.
The KCl salt samples are separately measured in a high precision weighing machine as per the requirement. The KCl samples are dissolved in deionized (DI) water (processed and purified by Millipore Synergy UV Water Purification System with a high accuracy (output water resistivity = 18 ×106 Ωcm)) and the KCl solution samples are prepared.
The impedance measurement tank is filled with KCl solution and EIS is conducted with a current signal of 1 mA using QuadTech 7600. The EIS study is also conducted for deionized water to study its frequency response as well as the conductivity variation over frequency.
It is observed that the conductivity of the KCl solution remains constant for all the frequencies (Fig.-4a). Fig.-4a also shows that the KCl solutions of different concentration show their individual conductivities which are all constant over frequency. Therefore it is proved that the KCl solution is purely resistive and hence it is independent of signal frequency. It is also observed that (Fig.-4b) the KCl solution conductivity (σKCl) is a linear function of concentration (ηKCl). Results show that rate of increase of σKCl at different frequencies is almost the same up to the KCl concentration of 1% (Fig.-4b). Beyond the KCl concentration of 1%, it is observed that the rate of increase of σKCl reduces slightly due to the higher ion concentration effect (Fig.-4b). It is also observed that, beyond the ηKCl of 1%, the rate of increase of σKCl differs from frequency to frequency up to 100 kHz. Beyond 100 kHz, the slopes of the σKCl curves for different frequencies are the same i.e. the rate of increase of σKCl are the same from 100 kHz to 1 MHz. This is because at the higher frequencies the contact impedance of the electrodes is reduced and ion mobility in the KCl solution increases.
Figure-4
Conductivity profile of KCl solution (a) variation (over frequency) at different concentration, (b) variation (over concentration) at different frequencies.
In the present study, it is to be noted that there is a difference between the geometry of the chicken muscle tissue paste sample (area = 20 mm × 20 mm, length = 50 mm) and the chicken fat tissue sample (area = 20 mm × 15 mm, length = 5 mm). Hence all the impedance parameters depending on the sample length should be understood with a proper scaling.
It is observed that the ZMuscle Paste (Fig.-5a), θMuscle Paste (Fig.-5b), RMuscle Paste (Fig.-5c), XMuscle Paste (Fig.-5d), ZFat (Fig.-6a), θFat (Fig.-6b), RFat (Fig.-6c) and XFat (Fig.-6d) are variable with frequency. Similarly ρMuscle Paste (Fig.-7a), σMuscle Paste (Fig.-7b), ρFat (Fig.-7c) and σFat (Fig.-7d) are variable with frequency. Results show that the variation in ZMuscle Paste is found between 10 Hz – 1 kHz where as ZFat varies throughout the frequency band (10 Hz – 2 MHz) with a larger variation between 10 Hz – 1 kHz.
Figure-5
Variation of impedance parameters of chicken tissue paste over frequency: (a) ZMuscle Paste, (b) θMuscle Paste, (c) RMuscle Paste, (d) XMuscle Paste
Figure-6
Variation of impedance parameters of chicken tissue over frequency: (a) ZFat, (b) θFat, (c) RFat, (d) XFat
Figure-7
Variation of impedance parameters of chicken tissue over frequency: (a) ρMuscle Paste, (b) σMuscle Paste, (c) ρFat, (d) σFat.
Result show that the variations of ZMuscle Paste, θMuscle Paste, RMuscle Paste, XMuscle Paste, ρMuscle Paste and σMuscle Paste are comparatively larger in the low frequency range (10 Hz – 1 kHz) whereas their variations are very low after 1 kHz. On the other hand the variations of ZFat, θFat, RFat, XFat, ρFat and σFat are found throughout the whole frequency range (10 Hz – 2 MHz) and all of them are comparatively larger than the impedance parameters of the muscle tissue.
It is observed that the sample to sample variations of ZMuscle Paste (Fig.-5a), θMuscle Paste (Fig.-5b), RMuscle Paste (Fig.-5c), XMuscle Paste (Fig.-5d) are comparatively lower than the corresponding variations in the ZFat (Fig.-6a), θFat (Fig.-6b), RFat (Fig.-6c) and XFat (Fig.-6d). Similarly the variations in ρMuscle Paste (Fig.-7a) a nd σ Muscle Paste (Fig.-7b) are comparatively lower than the ρFat (Fig.-7c) and σFat (Fig.-7d). It is observed that at 10 Hz the ZMuscle Paste varies (electrode to electrode variations) from 445.01 Ω to 465.54 Ω (Table-1) whereas the ZFat varies from 1168.15 Ω to 1065.09 Ω (Table-1).
Variation of impedance parameters of muscle tissue paste and fat tissue samples (ZMuscle Paste, θMuscle Paste, RMuscle Paste, XMuscle Paste, ρMuscle Paste, σMuscle Paste, ZFat, θFat, RFat, XFat, ρFat and σFat.) for 10 Hz to 2 MHz.
| Parameters | Min | Max | Mean | Change | |
|---|---|---|---|---|---|
| 10 Hz | 445.01 | 465.54 | 455.28 | 4.51 % | |
| 2 MHz | 207.75 | 209.33 | 208.54 | 0.76 % | |
| 10 Hz | 1065.09 | 1168.15 | 1116.62 | 9.23 % | |
| 2 MHz | 730.08 | 783.66 | 756.87 | 7.08 % | |
| 10 Hz | -40.34 | -39.38 | -39.86 | 2.40 % | |
| 2 MHz | 0.66 | 0.70 | 0.68 | 4.71 % | |
| 10 Hz | -18.82 | -18.41 | -18.62 | 2.21 % | |
| 2 MHz | -5.67 | -5.08 | -5.38 | 11.04 % | |
| 10 Hz | 343.76 | 354.96 | 349.36 | 3.20 % | |
| 2 MHz | 207.73 | 209.32 | 208.52 | 0.76 % | |
| 10 Hz | 1008.19 | 1108.41 | 1058.30 | 9.47 % | |
| 2 MHz | 727.22 | 779.82 | 753.52 | 6.98 % | |
| 10 Hz | 282.60 | 301.22 | 291.91 | 6.38 % | |
| 2 MHz | -2.52 | -2.42 | -2.47 | 3.95 % | |
| 10 Hz | 343.47 | 368.77 | 356.12 | 7.10 % | |
| 2 MHz | 64.62 | 77.44 | 71.03 | 18.05 % | |
| 10 Hz | 2.75 | 2.84 | 2.79 | 3.20 % | |
| 2 MHz | 1.66 | 1.67 | 1.67 | 0.76 % | |
| 10 Hz | 60.49 | 66.50 | 63.50 | 9.47 % | |
| 2 MHz | 43.63 | 46.79 | 45.21 | 6.98 % | |
| 10 Hz | 0.3522 | 0.3636 | 0.3579 | 3.19 % | |
| 2 MHz | 0.5972 | 0.6017 | 0.5995 | 0.75 % | |
| 10 Hz | 0.0150 | 0.0165 | 0.0158 | 9.44 % | |
| 2 MHz | 0.0214 | 0.0229 | 0.0221 | 6.99 % | |
Figure-5a shows that at 2 MHz the ZMuscle Paste varies (electrode to electrode variations) from 207.75 Ω to 209.33 Ω (Table-1) whereas the ZFat (Fig.-6a) varies from 730.08 Ω to 783.66 Ω (Table-1). Similarly, the percentage variations (variation between highest and lowest frequencies) of θFat (Fig.-6b), RFat (Fig.-6c), XFat (Fig.-6d), ρFat (Fig.-8c) and σFat (Fig.-8d) are found to be higher compared to θMuscle Paste (Fig.-5b), RMuscle Paste (Fig.-5c), XMuscle Paste (Fig.-5d), ρMuscle Paste (Fig.-8a) and σMuscle Paste (Fig.-8b) as shown by yellow shaded numbers in table-1. It is observed that the variations in Nyquist plots of the fat tissue samples are larger (Fig.-8a-8b) compared to the muscle tissue samples.
Figure-8
Nyquist plots tissue samples (a) muscle Paste, (b) fat.
Fig.-9a shows that for muscle tissue, the variation in ZMean is found between 10 Hz – 1 kHz whereas it remains almost constant within 1 kHz -2 MHz. On the other hand, for fat tissue, ZMean (Fig.-9a) varies throughout the frequency band (10 Hz – 2 MHz) with a larger variation within 10 Hz – 1 kHz. It is observed that for muscle tissue, the variations of θMean (Fig.-9b), RMean (Fig.-9c), XMean (Fig.-9d), ρMean (Fig.-10a) and σMean (Fig.-10b) are comparatively higher in the low frequency range (10 Hz – 1 kHz) whereas their variations are very low after 1 kHz.
Figure-9
Algebraic mean of the Z, θ, R and X of chicken muscle tissue paste and fat tissue: (a) Z, (b) θ, (c) R, (d) X
Figure-10
Algebraic mean of the ρ a nd σ of chicken muscle tissue paste and fat tissue: (a) ρMean, (b) σMean
On the contrary, for fat tissue, the ZMean, θMean, RMean, XMean, ρMean and σMean are found variable throughout the whole frequency range (10 Hz – 2 MHz) and the variations of all of them are comparatively larger than the impedance parameters of the muscle tissue.
Nyquist plot of the average of the ZMuscle Paste and ZFat show that for muscle tissue, the Warburg impedance (ZWMuscle Paste) [34] is found between 10 Hz to 34.2 kHz (Fig.-11a). On the other hand for the fat tissue the Warburg impedance (ZWFat) is found between 10 Hz - 49.5 kHz (Fig.-11b).
Figure-11
Mean Nyquist plots: (a) RMuscle Paste vs -XMuscle Paste, (b) RFat vs -XFat.
It is observed that, STDV of ZMuscle Paste (Fig.-12a), θMuscle Paste (Fig.-12b), RMuscle Paste (Fig.-12c) and XMuscle Paste (Fig.-12d) are less than the STDV of the corresponding impedance parameters of the fat tissue. STDV of the θMuscle Paste is found to be higher than the STDV of the θ Fat up to 100 kHz. It is also observed that the MDR of ZMuscle Paste (Fig.-13a), θMuscle Paste (Fig.-13b), RMuscle Paste (Fig.-13c) and XMuscle Paste (Fig.-13d) are higher than the MDR of the corresponding impedance parameters of the fat tissue. MDR of the θMuscle Paste is found to be less than the MDR of the θFat up to 100 kHz. Results show that STDV ρMuscle Paste (Fig.-14a) and σMuscle Paste (Fig.-14b) are comparatively lower than the STDV of ρ Fat and σFat. It is also observed that, the MDR of ρMuscle Paste (Fig.-15a) and σMuscle Paste (Fig.-15b) are comparatively higher than the MDR of ρFat and σFat.
Figure-12
STDV of the impedance parameters of chicken muscle tissue paste and fat tissue: (a) Z, (b) θ, (c) R, (d) X
Figure-13
MDR of Z, θ, R and X of chicken muscle tissue paste and fat tissue: (a) Z, (b) θ, (c) R, (d) X.
Figure-14
STDV of ρ and σ of chicken muscle tissue paste and fat tissue: (a) ρ, (b) σ
Figure-15
MDR of ρ and σ of chicken muscle tissue paste and fat tissue: (a) ρ, (b) σ
The EIS between 10 Hz to 2 MHz fails to show α, β and γ dispersion [46, 47, 48, 49] of chicken muscle tissue paste. This is due to (i) the muscle tissue is grinded to have a paste in which the tissue and cell structure are disturbed and (ii) the lower frequency range (10 Hz to 2 MHz) in EIS. Hence to obtain the α and β dispersion of chicken muscle tissue, EIS is conducted for a rectangular (15 mm × 20 mm × 20 mm) chicken muscle block (not grinded) within a frequency range of 40 Hz to 20 MHz using the Agilent 4294A with a 1 mA constant current signal. EIS is also conducted for a rectangular (50 mm × 20 mm × 20 mm) chicken muscle paste (not grinded) sample within same frequency range and current signal using the Agilent 4294A.
Z, θ, R, X, ρ, σ, C, ε and εr all are calculated and studied for chicken muscle tissue blocks and chicken muscle tissue paste samples. Results show that the muscle tissue block clearly exhibits its β dispersion in 0.1 – 10 MHz frequency range [48, 49]. The γ dispersion is not obtained as it occurs beyond 1 GHz [49].
In the present study, it is to be noted that there is a difference between the lengths of the chicken tissue muscle block sample (15 mm) and the chicken tissue muscle paste sample (50 mm). Hence all the impedance parameters depending on the sample length should be understood with a proper scaling (area of current conduction is fixed for all the samples).
Figure-16 and figure-17 show that, the variations in ZMuscle Block (Fig.-16a), θ Muscle Block (Fig.-16b), R Muscle Block (Fig.-16c) and X Muscle Block (Fig.-16d) of chicken muscle tissue block are larger than the ZMuscle Paste (Fig.-17a), θMuscle Paste (Fig.-17b), RMuscle Paste (Fig.-17c) and XMuscle Paste (Fig.-17d), respectively. Results show that the variation in ZMuscle Block (Fig.-16a) is found throughout the frequency range (40 Hz to 20 MHz) whereas the variation in ZMuscle Paste (Fig.-17a) is found only at very low (f < 200 Hz) or very high frequencies (f > 2 MHz) due to the electrode polarization. It is observed that the variations in ρMuscle Block (Fig.-18a), σMuscle Block (Fig.-18b), CMuscle Block (Fig.-18c) and εrMuscle Block (Fig.-18d) of the chicken muscle tissue block are heiger than the ρMuscle Paste (Fig.-19a), σMuscle Paste (Fig.-19b), CMuscle Paste and εrMuscle Paste, respectively.
Figure-16
Variations of impedance parameters of chicken muscle tissue block over frequency: (a) ZMuscle Block, (b) θMuscle Block, (c) RMuscle Block, (d) XMuscle Block.
Figure-17
Variations of impedance parameters of chicken muscle tissue paste over frequency: (a) ZMuscle Paste, (b) θMuscle Paste, (c) RMuscle Paste, (d) XMuscle Paste.
Figure-18
Variations of impedance parameters of chicken muscle tissue block over frequency: (a) ρMuscle Block, (b) σMuscle Block, (c) CMuscle Block, (d) εrMuscle Block.
Figure-19
Variations of impedance parameters of chicken muscle tissue paste over frequency: (a) ρMuscle Paste, (b) σMuscle Paste.
Fig.-16 through Fig.-19 show that the α and β dispersions are visible in all the impedance parameters (Z, θ, R, X, ρ, σ, C, ε and εr) of both the tissue samples, but the dispersions are larger for muscle tissue block compared to the muscle tissue paste.
The Nyquist plot (Fig.-20) [34] obtained for the muscle tissue block demonstrates that the equivalent electric model of the tissue sample contains Warburg impedance [34, 50, 51] and a constant phase element [50].
Figure-20
Nyquist plots tissue samples muscle tissue block
As the α and β dispersions are larger for the muscle tissue block, the muscle tissue block is found more suitable for the phantom materials of multifrequency EIT compared to the tissue paste. As the muscle tissue block has larger dispersion in a wide range of frequencies, it is more suitable for multifrequency EIT both for low frequency and high frequency EIT. On the contrary, as the α and β dispersions are smaller for muscle tissue paste, it can also be used for phantom materials in low frequency EIT.
Saline phantoms with insulator inhomogeneities cannot mimic the physiological structure of real body tissue as the saline or any other salt solutions are purely resistive, whereas the biological tissues are always found with a complex electrical impedance. Therefore the saline solution and insulator inhomogeneity are not always suitable materials for practical phantoms to study the multifrequency EIT because the response of the purely resistive materials does not change over frequency. In saline insulator phantoms the NaCl or KCl solution shows constant response over frequency and the insulator inhomogeneity is a pure high-resistive material. As the animal tissues show a variable response over a wide band of signal frequencies due to their complex physiological and physiochemical structures, they can suitably be used as bathing medium and inhomogeneity in the phantoms of multifrequency EIT systems. Moreover, an efficient assessment of a multifrequency EIT system with a real tissue phantom needs a prior knowledge of the impedance profile of the bathing medium and inhomogeneity. Hence it is essential to study the electrical impedance spectroscopy of biological tissues selected for EIT-phantom materials.
Hence the chicken tissues are proposed as the EIT phantom materials and in this direction EIS studies of broiler chicken muscle tissue blocks, muscle tissue paste and fat tissue are conducted in a wide range of frequencies using impedance analyzers and their impedance profiles are thoroughly studied. For a suitable and fair comparison, EIS of a KCl solution is also studied and the results are analyzed. It is observed that the conductivity of the KCl solutions remain constant over the frequency. Therefore it is proved that the KCl solution is purely resistive and hence it is independent of signal frequency. It is also observed that the σKCl is a linear function of ηKCl and the rate of increase of σKCl at different frequencies is almost the same up to the KCl concentration of 1%. Beyond the KCl concentration of 1%, it is observed that the rate of increase of σKCl reduces slightly due to the higher ion concentration effect. It is also observed that, beyond the ηKCl of 1%, the rate of increase of σKCl differs from frequency to frequency up to 100 kHz. On the other hand, the slopes of the σKCl curves for different frequencies are found to be the same beyond 100 kHz. The reason is that at the higher frequencies the contact impedance of the electrodes is reduced and ion mobility in the KCl solution increases.
It is observed that the all the impedance parameters (Z, θ, R, X, ρ, σ, C , ε and ε r) of chicken muscle tissue blocks, muscle tissue paste and fat tissue are functions of frequency. Muscle tissue blocks and fat tissue show a wide range of variations over signal frequency whereas the muscle tissue paste shows a very low variation at high frequencies. Hence, chicken muscle tissue and fat tissue are found suitable for phantom materials in multifrequency EIT systems. Muscle tissue paste can be used for low frequency EIT or single frequency EIT. It is observed that the sample to sample variation of impedance parameters are found to be smaller in chicken muscle tissue paste and hence it is found to be more homogeneous than the fat tissue. Therefore muscle tissue paste is found more suitable as the phantom bathing medium compared to fat tissue. Fat tissue is found more resistive and offers a wide range of variation in its impedance parameters and hence it is found suitable as the inhomogeneity in multifrequency EIT phantoms.
It is observed that the phantom can be used for 4-5 days if we preserved it in low temperature. But in ambient temperature it deteriorates after a day. Membranes of the cells in the muscle tissue paste are destroyed due to the grinding and therefore the muscle tissue paste shows a very low dispersion at frequencies greater than 1 kHz. In the proposed EIT phantoms, the reconstruction parameter is the resistivity which is also frequency dependent and calculated from the real part of the complex impedance and therefore the phase angle is not so important here. Hence, the low phase angle of chicken tissue paste is not a problem for absolute imaging of single frequency static EIT. The EIT phantom with uncrushed tissues (small pieces of tissues) is non uniform and each tissue block will show different frequency responses. Uniform background medium with a chicken tissue block is difficult to make as a single chicken tissue of 150 mm diameter (diameter of the phantom in our present study) is not available. Non-uniform bathing medium will make the EIT imaging study more complex. Hence a chicken tissue paste is taken, which is uniform all over the background region and hence the entire bathing medium (background region) will show a similar frequency response but different from the frequency response of the inhomogeneity. However, studying a multifrequency EIT system needs a medium showing a large dispersion in the high frequency range. In this direction a detailed study with fresh muscle tissue blocks is conducted within a frequency range of 40 Hz to 20 MHz using the Agilent 4294A with a 1 mA constant current signal. Z, θ, R, X, ρ, σ, C, ε and εr all are calculated and studied for chicken muscle tissue blocks and chicken muscle tissue paste samples. Results show that the muscle tissue block clearly exhibits its β dispersion in the 0.1 – 10 MHz frequency range. The γ dispersion is not obtained as it occurs beyond 1 GHz. Results also show that the α and β dispersions are visible in all the parameters of both the tissue samples, but the dispersions are larger for the muscle tissue block. The Nyquist plot obtained for the muscle tissue block demonstrates that the equivalent electric model of the tissue sample contains Warburg impedance and a constant phase element. Hence, the muscle tissue block is suitable for multifrequency EIT compared to the tissue paste. As the muscle tissue block has larger dispersion in the high frequency range and hence it is more suitable for multifrequency EIT as well as for high frequency EIT. Collectively it is concluded that the chicken muscle tissue and fat tissue can be successfully used as the bathing medium and inhomogeneity (though the reverse can also be done), respectively, in the phantoms suitable for impedance imaging in multifrequency EIT.
EIS of broiler chicken muscle tissue paste, chicken muscle tissue blocks and broiler chicken fat tissue are conducted in a wide range of frequencies using impedance analyzers and their impedance profiles are thoroughly studied. Chicken tissues showed a variable response over a wide band of signal frequencies due to their complex physiological and physiochemical structures and hence they are found suitable as phantom materials in multifrequency medical EIT systems. It is found that the all the impedance parameters (Z, θ, R, X, ρ, σ, C , ε and εr) of chicken tissue samples studied are functions of frequency. Moreover the sample to sample variation of impedance parameters is found to be less in chicken muscle tissue paste and hence it is found more homogeneous than the fat tissue. Therefore muscle tissue paste is found more suitable as the phantom bathing medium compared to fat tissue. Fat tissue is found more resistive and offers a wide range of variation in its impedance parameters and hence it is found suitable as the inhomogeneity in multifrequency EIT phantoms. The variations in the impedance parameters of muscle tissue blocks are found to be larger compared to the tissue paste as the cell membrane structures are destroyed during grinding. Results also show that the α and β dispersions are visible in all the parameters of both the tissue samples, but the dispersions are larger for the muscle tissue block. The Nyquist plot obtained for the muscle tissue block demonstrates that the equivalent electric model of the tissue sample contains Warburg impedance and a constant phase element. On the contrary, it is observed that the conductivity of the KCl solutions of all concentrations is independent of signal frequency and σKCl is a linear function of ηKCl. Finally it is concluded that the chicken muscle tissue and fat tissue can be used successfully as the bathing medium and inhomogeneity in the phantoms suitable for impedance imaging in multifrequency EIT.