From a structural point of view, the trimmed meat from bovine muscle is an anisotropic tissue, which is characterized by a composite network of muscle fiber bundles containing aligned myofibers surrounded by a fine endomysial envelope of connective tissue. Several structural levels that characterize the muscle's morphology have highly contrasted electrical and dielectric properties. In fact, the meat's electrical properties result in part from non-frequency-dependent materials, which define the meat's resistance, i.e. an array of highly elongated structures with high longitudinal conductance due to both intra- and extra-cellular presence of ions, which are surrounded by connective sheaths with a very low conductance. Electrical characteristics of meat are also due to frequency-dependent materials, which define its capacitive reactance due to the cell membrane property of maintaining the separation of the negative from positive electrical charges across it, as in an electrical capacitor. Both these electrical specificities give rise to an electric anisotropy of the muscle, which is closely dependent on its histological characteristics.
In slaughtered meat, it has been found that during the post-rigor period (which begins 2 – 3 days after slaughter) the behavior of its electrical impedance (Zm) reflects major changes occurring in the meat's structure , and these modifications are related to the disruption of the myofibrillar organization of the cytoskeleton and of cell membranes, due to protease activity . In fact, during this time-dependent proteolysis, i.e. the meat's maturation, degradation occurs in proteins with structural tears and myofibril fragmentation, together with degradation in cytoskeleton architecture . These structural modifications of trimmed muscle give rise to a progressive loss of structural anisotropy from which a reduction of Zm also occurs [4,5].
The optimization of the slaughtered meat maturation period, paying close attention that the maturation process takes place properly, is one of the main goals of the meat industry, particularly of the beef sector . In fact, this production sector has to contend with the broad-ranging variability of the raw material and with the low process control on the marketed end product, especially concerning the freeze-thaw cycles that the meat often has to suffer with a worsening of its quality. In order to guarantee good quality of meat products the industry is looking for instrumental systems to assess and certify the product quality of this food supply chain.
In their paper, Banach
Starting from all the above considerations, the aim of this study was that of developing and testing a portable and cheaper device able to safely and easily assess the Zm values of slaughtered beef muscles by injecting an electrical current with a high frequency, the preset value of which surely short-circuits the capacitive reactance of the cell membranes and thereby reduces the Zm to just the intra- and extra-cellular resistive components .
The present work is available on preprints.org (
Fricke and Morse [9,10] have previously described an analogous electrical model which equates biological tissue components to passive electrical elements like resistors and capacitors connected in series and parallel.
So, as is shown in Figure 1, a possible electrical model which could consent the expression of quality attributes of meat may consist of a resistor including extra-cellular fluids together with some connective tissue elements (Re) connected in parallel with the resistor including intra-cellular fluids (Ri) which is in series with the membrane capacitors (Cm) . The meat's electrical impedance (Zm) can be calculated as in Equation (1), where j is the imaginary unity.
It is easy to deduce that as f progressively increases the Xm value tends towards 0, and Zm tends to the expression of Equation (4), or rather, the meat's electrical impedance becomes dependent purely on changes in both intra- and extra-cellular resistive structures, which can be strongly conditioned by the maturation process of the meat  or, when occurring, on ice crystals between and within muscle fibers if the meat has been submitted to freezing/thawing processes [11,12,13].
On the basis of the analogous electrical model of Fricke and Morse  and to reach our goal, we applied a portable and single-frequency device with the capacity to easily assess measurements of the resistive components of the electrical impedance in the trimmed meat from beef carcasses, actuated at the production sites and with the possibility of sending bioimpedance signals to a remote control center. The afore-mentioned device is our own development, called ZMeat, consisting of three separate blocks: a sampling front-end, a command and control system and a data transmission system.
As is shown in Figure 2, the first block of the ZMeat is the sampling front-end, which contains an alternate current supply that injects 1 mA at 65 kHz into the meat sample using two copper electrodes (black arrows) while two other electrodes connect the meat to the amplifier section (gray arrows). The front-end also contains a three-stage filtering section to ensure the elimination of frequencies outside the range of interest from the acquired signals. In the last stage of the front-end a digital conversion operation is performed by an appropriately calibrated 10 bit analog-to-digital converter. The core of the sampling front-end is the implemented and very cheap (about USD 200) ADAS-1000 evaluation board (Analog Devices, USA ). The command and control block supervises the sampling sessions by initializing and driving the ADAS-1000's operations and sends all the information to a controlling software operating on a personal computer by means of a data transmission system which contains a Bluetooth© microchip fully supporting the radio frequency communication protocol, which enables a totally wireless operation mode of the ZMeat.
A chosen signal injected frequency of 65 kHz was used because other authors verified that at 10 kHz the values of impedance and resistance over a longissimus dorsi portion were practically the same . So, a higher frequency such as 65 kHz guarantees that Xm has been eliminated. At the same time, the ZMeat core is identical to another development of the authors dedicated to bioimpedance measurements [15,16,17,18,19].
The conducted research is not related to either human or animal use.
Samples studied were obtained from four steers (half-breed Belgian blue cattle with Friesian) of about 16 months reared in a barn . These were tested by the ZMeat. The considered carcass portions were longissimus dorsi muscles (LDm) of both sides corresponding to the VII–XI thoracic vertebrae.
Trimmed LDm portions (see Figure 3) were 25 cm length with a rostral diameter of about 15 cm and a caudal diameter of about 10 cm, and each weighing from 6 to 8 kg. Each LDm portion was collected from a local abattoir 6 hours after slaughter and was packed in a plastic bag.
In these experiments, inspiration was drawn from the experimental protocols applied in previous experiments concerning the sequence of critical times to be applied just after slaughter LDm to obtain a good maturation . On obtaining the samples from the local abattoir, both sides of LDm were immediately placed in a chilled room at 5 °C. The right sides of LDm stayed in the chilled room up to the 6th day post-slaughter. The left sides of LDm, from the 3rd until to 6th day post-slaughter were frozen at −18 °C and then, at the 7th day post-slaughter, these samples were returned to the initial chilled room at 5 °C for thawing.
Experiments for Zm measurements of both LDm groups were undertaken in the chilling room with the following post-slaughter periodicity: 2nd day (or thereabouts after the rigor period), 7th day. So, in both groups of LDm each of the two Zm measurements, i.e. the first as basal values and the second as post heat treatment values, were made at an identical time after the animals’ slaughtering. Figures 3 and 4 shows that the impedance measurements were carried out by utilizing the tetrapolar scheme , which uses two electrodes to inject a current flow I into the meat sample and two different electrodes to measure the voltage ΔV between these two electrodes and to deduce electric impedance by applying Ohm’s law (Equation (5)).
The four electrodes were annealed and cold-drawn copper wires with a diameter of 0.5 mm. Each electrode was inserted manually in the sample of the meat for a length of 2.5 cm.
Impedance measurements were made by inserting the four electrodes along the same line, ideally located along the muscle fibers and each pair of injection or detection electrodes was placed symmetrically with respect to the ideal point corresponding to one half of the muscle length. (see Figure 4).
From each meat sample 9 measurements of Zm were taken and in each one of these, distances between the injection-detection (Inj-Det) pair of electrodes and the detection (Det-Det) pair of electrodes were modified as shown in table 1.
Distance between electrodes.
Distance refers to each pair of injection (Inj) and detection (Det) electrodes;
Distance refers to the two detection electrodes.
The procedure was carried out in this way to eliminate as much as possible any influences on the measured Zm values due to differences in the relative positions of the electrodes. The time span between two consecutive measurements was 10 minutes.
Sampling tests were set for a minimum duration of 20 seconds each. The ZMeat was configured to always operate at a sampling frequency of 2 kilosamples per second, recording a total of about 40,000 samples during each individual session.
Measured values of Zm were parametrized as Ω per cm (Ω/cm) for scaling the differences between each measurement ‘Test Number’ (see Table 1). A similar data treatment was carried out by Gabriel
Measured values of parametrized Zm from all four tested animals were subdivided into the following three groups: ZmF containing impedance measurements from the left LDm frozen samples made at the 7th day after slaughter; ZmNF containing impedance measurements from the right LDm non-frozen samples made at the 7th day after slaughter. Since no statistical difference was found between the two groups of the Zm measured in both right and left LDm at the 2nd day after slaughter, they were included in the same group; ZmB, which served as the baseline control of the two subsequent heat treatments.
Mean values ± SD were calculated for each data group and differences between them were statistically evaluated by the Student's t-test and, where appropriate, nonparametric rank tests were also applied. Differences among tested data were considered as significant for P<0.05.
Statistical tests were carried out utilizing commercially available software (MedCalc, Belgium).
Table 2 shows that 7 days after the animals’ slaughter the ZmNF samples of LDm which aged 4 days at 5°C in the chilled room after the ZmB measurements, reduced its value slowly (about −12%), even though significantly, with respect to that of the ZmB. On the contrary, the ZmF samples of LDm which stayed 4 days after the ZmB measurements at −18°C before thawing, showed a larger and significant reduction (about −39%) with respect to that of the ZmB.
Mean impedance values among the tested beefs.
Mean ± SD of the electrical impedance values of the slaughtered meat in the base condition (ZmB), not frozen (ZmNF) and frozen and thawed (ZmF) conditions. N: number of measurements;
P<0.0001 with respect to ZmB;
P<0.0001 with respect to ZmNF.
To highlight the different behavior of Zm when LDm were submitted or not to a freezing treatment, Figure 5 shows a graphic representation of a typical behavior of the impedance in the longissimus dorsi muscle samples concerning one of the tested animals. Due to the asymmetric distribution of the data, the visualization of the assessed values of the specific impedance was made by means of box and whiskers plots [26,27] in which we have also included the spread of data. Figure 5 shows a practically complete overlap between the third quartile (Q3) of the ZmB box and the first quartile (Q1) of the ZmNF box, thus their respective median lines are very close (ZmB = 6.6 Ω/cm and ZmNF = 5.8 Ω/cm) with no statistically significant difference between them.
The same Figure 5 clearly shows that the box of ZmB data is placed totally above the box of ZmF data, and the median value of this latter group is lower than half that of the base group of data. For this reason, statistical comparison between these two groups of data showed high significant difference. When comparing both the ZmNF and the ZmF box, it appears that the ZmNF box is also placed above the ZmF one, and the median of non-frozen group data is 1.9 times higher than that of the frozen one. Also, this difference between Zm values turned out to be highly significant.
Results arising from these experimental measures of the meat's electrical impedance reasonably consent us to consider 65 kHz of frequency for the applied signal for Zm monitoring, i.e. that released from the ZMeat, as a reliable frequency for giving information concerning the ageing-dependent maturation of bovine meat.
In fact, it must be taken into account that ageing-dependent maturation of meat is due to a progressive disruption of the composite network of myofibres with their surrounding envelopes of connective tissue. This leads to an irreversible morpho-functional alteration of the cytoskeleton and cell membranes of myocytes  with a progressive loss of the histological anisotrophy of the muscle tissue. This loss of the morphological anisotropy induces a parallel reduction in the muscle's electrical anisotropy  which is revealed by a progressive falling of its electrical impedance  which coincides with its real, i.e. resistive, component.
Considering that in its hardware, the ZMeat implements a current generator which delivers a current flow into muscle tissue with a frequency of 65 kHz, or a frequency which is higher than 10 kHz, which has been found to nullify the muscle capacitive reactance , then it is reasonable to assume that our ZMeat measures the resistive components as described in the muscle's electrical analog model , i.e. this Zm monitor could give information on the ageing-dependent maturation of meat in the form of numerical values.
Paying attention to the results from these experiments, data indicate that, at the 7th day from slaughter, a portion of the beef longissimus dorsi muscle frozen at −18° C in the previous 4 days, when thawed, showed a Zm value which was almost half that of the value assessed in the 2nd day post-slaughter, before its freezing. These results were in good agreement with those previously found by Banach
That described above, from both ours and Banach
In fact, in the mile-stone paper by Rahelić and Puać  it was clearly demonstrated that when samples of beef longissimus dorsi muscle were frozen at temperatures between −10 °C and −22 °C ice crystals formed both between muscle cells and inside these cells. In fact, these authors showed that histological changes in muscles frozen at −10 °C presented in cross-section groups of several tens of fibers. These groups of fibers were separated by large intercellular spaces due to ice crystal formation. In these groups, muscle fibers were attached to one another and individual groups were linked by one or two rows of fibers. Otherwise, when the muscle samples were frozen at −22 °C the fiber groups were smaller with respect to those frozen at −10 °C and were separated by smaller interspaces. However, there are frequent gaps in the middle of fibers due to ice crystals formed inside the cells. This study concluded that damage is most severe in the muscles frozen at −22 °C. To justify this claim they proposed that opposing pressures through cell membranes are created from outside the cell by large ice crystals formed intercellularly and from inside the cell by pressures created by the formation of ice crystals here too. As a result of such opposing pressures, the tearing of the fibers will be greater.
More recently, Egelandsdal
Even though in the present experiment no histological examination of the meat samples was carried out either at the beginning or at the end of the two experimental treatments, on the basis of the previous experimental results as reported above, it is reasonable to assume that in our bovine longissimus dorsi muscle samples that were frozen for 4 days at −18 °C, the histological structure underwent dramatic changes due to formation of ice crystals both outside and inside the muscle fibers.
These occurrences are even more supported by a recent histological study of samples of both longissimus lomborum and semitendinosus beef muscles, which were frozen at −20 °C for 3 days after slaughter . Observing muscle micro slices under the light microscope the authors of these experiments found the presence of ice crystal formation in more elongated shapes and widely distributed patterns. Moreover, in the same experiments, muscle samples frozen at −20 °C for 3 weeks and after thawing, showed more gaps between muscle fibers compared to the ones that underwent maturation at 2 °C for 2 weeks. This occurrence could further justify the disruption of the histological anisotropy, as described by Rahelić and Puać , which could happen when muscle samples are frozen at low temperatures around −18 °C, i.e. what could happen to our left side longissimus dorsi muscles of beef samples.
This evidence would lead us to expect a remarkably better electrolytic conductance in frozen meat, in comparison to the non-frozen one, which is linked with the rapid and profound change in its anisotropic structure, due to the formation of ice crystals, i.e. an alteration of insulating properties of cellular membranes allowing a great mobility of ions could happen [11,33]. The degree of this change depends on the rate of the freezing process and the amount of water inside the structure of muscle tissue . In the present experiment we also found that, as expected, after 7 days post-slaughter the Zm of the sample muscles which had been frozen in the previous 4 days and then thawed were almost a half that of the only-aged one. This latter result reinforces the validity of this 65 kHz ZMeat monitor to also discriminate between fresh and frozen-thawed beef meat.
These experimental results reasonably demonstrate that our ZMeat prototype applied to Zm monitoring is a reliable device to assess resistive components of the electrical impedance in trimmed longissimus dorsi muscle of slaughtered bovine since impedance values assessed by it behaved in agreement with those of previous experiments concerning the same matter.
The use of a custom designed device with a dedicated control software allowed the development of a low cost, compact, portable tool able to automatically and wirelessly perform all the measurements of the resistive components of the electrical impedance in the trimmed meat in real-time. So, this device might meet the needs of the bovine meat industry especially when it is necessary to rapidly check, in the place of production or storage, if such meat is fresh or previously frozen.
A possible criticism for this manuscript could be drawn due to the lack of histological evaluations of our meat samples both at the beginning and the end of the two experimental treatments. However, many previous experiments [29,30,31,32] have already investigated in great depth into the ultra-structural changes that animal elongated muscles undergo when frozen at temperatures close to that used in this experiment, and the results of these previous experiments can reasonably be used as a reference of the possible histological damage that also occurred in our samples of bovine longissimus dorsi muscles. Nevertheless, we are looking for a skilled histological laboratory which could be capable of providing information concerning the ultrastructural changes of beef muscle samples to be compared with our ZMeat data.
Another possible criticism could be drawn due to the relatively small number of animals studied. However, it is quite possible that calves coming from the same breeding and of the same age and weight would also present very similar values of morphological and functional parameters due to the high affinity in genetic architecture of livestock breeds . Nevertheless, the data reported in this paper are based on a pilot study to be followed as soon as possible by a confirmatory study based on a larger sample of tested animals.
Distance between electrodes.
Mean impedance values among the tested beefs.