Source of tissues

In this study, a total of eight ovine adrenal gland samples (N = 8) were dissected from slaughtered animals in a local abattoir. After excision, the glands were sealed in closed plastic containers within their fat capsule to prevent tissue dehydration. Samples were transported to our lab within 24 h. Since the adrenal glands were obtained from animals slaughtered for human consumption within a regulated abattoir, therefore, no ethical approval was necessary. Once the samples were received in the laboratory, the adrenal glands were separated from the surrounding structures, such as the kidney and fat. The size of the samples varied, with lengths ranging between 25.0 ± 7 mm, widths of 10.6 ± 2 mm, and thicknesses of 5.4 ± 0.9 mm. The adrenal gland samples were sectioned to provide access to both the adrenal cortex and the medulla layers, and measurements were performed on the internal surface of the tissue to minimize the effect of surface dehydration. Figure 1 shows the photograph of one of the adrenal gland samples and the measurement sites on the adrenal cortex and medulla tissues.

Figure 1. Photograph of ex vivo ovine adrenal gland showing adrenal cortex and medulla. The open-ended coaxial (OECL) probe is in contact with the adrenal cortex. The measurement points are labeled.

Dielectric properties measurements

The dielectric properties were measured using a slim-form OECL probe (Keysight 85070E, Santa Rosa, CA, USA). The reflection coefficient (${S_{11}}$) at the calibration plane of the probe was recorded using a vector network analyzer (VNA) (Keysight VNA E5063A, Santa Rosa, CA, USA). The data were recorded in the frequency range of 0.5–8.5 GHz over 101 linearly spaced frequency points. A commercially available software suite (Keysight N1500A, Santa Rosa, CA, USA) was used to convert the ${S_{11}}$ parameters to real ($\varepsilon {^{\prime}}\left( \omega \right)$) and imaginary ($\varepsilon {^{\prime \prime}}\left( \omega \right)$) parts of the complex permittivity [Reference Bottiglieri, Shahzad, Donlon, Dennedy, Lowery, O'Halloran and Farina11, Reference Amin, Shahzad, Farina, Parle, McNamara, O'Halloran and Elahi20, Reference La Gioia, Porter, Merunka, Shahzad, Salahuddin, Jones and O’Halloran26]. The dielectric properties were recorded by precisely placing the probe on the tissue sample as shown in Fig. 2. The slim-form probe was firmly attached to the VNA to minimize the measurement inaccuracies caused by repositioning the probe. A lift stand as shown in Fig. 2 was used to place the tissue samples during the measurements to avoid sample movement and to ensure proper contact of the probe with the tissue sample. The temperature of each sample was recorded using a digital infrared thermometer with dual-laser targeting (Precision Gold, N85FR). The experimental setup for measuring the dielectric properties is shown in Fig. 2.

Figure 2. Dielectric properties measurement setup.

The temperature of the samples varied between 22.80 ± 0.70°C. At each measurement location of the sample, a total of five measurements were performed. Accordingly, for the eight ovine adrenal glands, a total of 335 dielectric measurements were recorded and analyzed (total measurement points = 67; adrenal cortex measurement points = 35; and adrenal medulla measurement points = 32). The probe was cleaned with an alcohol wipe before each measurement.

Temperature-dependent dielectric properties

In this study, the temperature-dependent dielectric properties of one of the adrenal glands were measured using the Fisher Isotemp Water Baths (Fisher Scientific 154626Q, Hampton, NH, USA). The temperature-dependent dielectric properties were measured only for the adrenal medulla. Due to the shrinking width of the adrenal cortex tissue with increasing temperature, it was challenging to achieve proper contact between the OECL probe and the adrenal cortex tissue. Therefore, the temperature-dependent dielectric properties of the adrenal cortex were not taken into consideration. The water-filled bath was placed on the lift table. The adrenal gland was placed in a plastic container to prevent direct contact with the water. The temperature of the water bath was raised to predetermined values of 37°C, 46°C, 53°C, and 64°C, and the dielectric properties of the adrenal gland were measured at each temperature using the measurement setup illustrated in Fig. 3. These temperatures represent an important spectrum of temperatures commonly used in clinical practice for effective MWA. The time to reach the sample to the desired temperatures was not recorded in this particular experiment. Furthermore, the sample was not covered during the heating process. This may have an impact on the heating rate, as covering or not covering a sample can affect how it interacts with the heating source. Moreover, the experimental conditions were kept consistent to ensure reliable results.

Figure 3. Temperature-dependent dielectric properties measurement setup.

Measurement protocol and system uncertainty analysis

The dielectric properties measurement equipment was calibrated before each measurement session. The slim-form probe was connected directly to the VNA to avoid any errors caused by the movement of the cable. The calibration of the equipment was performed on three standard loads, including open (air), short (shorting block), and load (deionized water) [Reference La Gioia, Porter, Merunka, Shahzad, Salahuddin, Jones and O’Halloran26, Reference Shahzad, Sonja, Jones, Dwyer and O’Halloran27]. The mean value of the temperature of the deionized water used for calibration during the dielectric properties measurement session was 24.45 ± 0.35°C. Once the equipment was calibrated, a validation measurement was performed by verifying the dielectric properties of 0.1 M NaCl (saline) solution at 24.90°C [Reference La Gioia, Porter, Merunka, Shahzad, Salahuddin, Jones and O’Halloran26, Reference Peyman, Gabriel and Grant28]. A total of six validation measurements were performed. An estimation of the uncertainty components related to the accuracy and repeatability of the measurement system was performed for each frequency point and then averaged over the frequency range of the measurement (0.5–8.5 GHz). The uncertainty about the accuracy ($U{C_{ACC}}$) of the measurements is expressed as the average percentage difference between the measured and reference dielectric properties data. The uncertainty in accuracy in terms of percentage is defined as follows:

(1)\begin{equation}U{C_{ACC}}\left(\, f \right) = \,\,\left( {\frac{{D{P_{meas}}\left(\, f \right) - \,D{P_{ref}}\left( \,f \right)}}{{D{P_{ref}}\left(\, f \right)}}} \right)\, \times 100,\end{equation}

where $D{P_{meas}}\left(\, f \right)$ represents the measured dielectric properties of 0.1 M NaCl solution at a specific frequency point $f$ and $D{P_{ref}}\left(\, f \right)$ represents the reference dielectric properties of 0.1 M NaCl solution at the measured temperature as outlined by Gabriel and Peyman [Reference Gabriel and Peyman29]. The uncertainty related to repeatability ($U{C_{REP}}$) provides a measure of the random errors affecting the measurement process. The repeatability is expressed as the standard deviation of the dielectric properties data that are repeatedly acquired under the same measurement condition. The uncertainty in repeatability of measurements in terms of percentage is defined as follows:

(2)\begin{equation}U{C_{REP}}\left(\, f \right) = \,\,\left( {\frac{{D{P_{meas}}\left(\, f \right) - \,D{P_{mean}}\left(\, f \right)}}{{D{P_{mean}}\left(\, f \right)}}} \right)\, \times 100,\end{equation}

where $D{P_{mean}}\left(\, f \right)$ represents the mean of the measured dielectric properties of 0.1 M NaCl solution at a specific frequency point $f$. In this study, the uncertainty due to the cable movement was not considered as the probe was directly connected to the VNA. Moreover, the probe was not moved during the dielectric properties measurement sessions. The combined uncertainty ($U{C_{COM}}$) was computed by using the standard uncertainty ($U{C_{STD}}$) values of each of the uncertainty components ($U{C_{ACC}}$ and $U{C_{REP}}$). The expanded uncertainty ($U{C_{EXP}}$) was calculated by considering a 95% confidence interval, according to the guidelines of the National Institute of Standard and Technology [Reference Taylor and Kuyatt30]. In this study, the expanded uncertainty of the measurement equipment was found to be 0.78% and 5.08% for relative permittivity and conductivity, respectively. All the uncertainty components computed in this study are reported in Table 1. The uncertainty components are computed separately for both the relative permittivity and conductivity of 0.1 M NaCl for a frequency range of 0.5–8.5 GHz.

Table 1. Uncertainty in accuracy and repeatability components for measured relative permittivity and conductivity of 0.1 M NaCl for a frequency range of 0.5–8.5 GHz

λ = multiplying factor for computing the expanded uncertainty, R = rectangular distribution, N = normal distribution.

Fitting model on ex vivo ovine dielectric properties data

The dielectric properties of biological tissues are highly frequency-dependent [Reference Gabriel, Gabriel and Corthout31]. The two most commonly used parametric models of dielectric properties are the Cole–Cole model and the Debye model [Reference Salahuddin, Porter, Krewer and O’ Halloran21, Reference Ištuk, Porter, O’loughlin, McDermott, Santorelli, Abedi, Joachimowicz, Roussel and O’halloran32]. The Cole–Cole model is computationally expensive and cannot be easily expressed in the time domain [Reference Salahuddin, Porter, Krewer and O’ Halloran21]. However, the Debye model can be easily expressed in both the time and frequency domains and is not computationally expensive [Reference Salahuddin, Porter, Krewer and O’ Halloran21, Reference Fujii33]. Therefore, this study has employed the Debye model as a fitting model for the ovine dielectric properties measured in this study. The Debye model is defined as follows:

(3)\begin{equation}\hat \varepsilon \,\, = \,\,{\varepsilon _{\infty \,}} + \,\mathop \sum \limits_{p = 1}^n \frac{{\Delta {\varepsilon _{p\,}}}}{{1 + j\omega {\tau _p}}} + \,\frac{{{\sigma _s}}}{{j\omega {\varepsilon _o}}}\,,\end{equation}

where $\hat \varepsilon $ is the complex permittivity of the material, ${\varepsilon _{\infty \,}}$is the permittivity at the highest frequency, $\Delta {\varepsilon _{p\,}}$is the change in the permittivity at the pth dispersion, ${\tau _p}$ is the relaxation constant at the pth dispersion, ${\sigma _s}$ is the static ionic conductivity, $\omega \,$is the angular frequency, and ${\varepsilon _o}$ is the permittivity of free space. The Debye parameters were optimized by using the algorithm. In this study, a weighted least squares method (W-LSM) was developed in MATLAB (R2017b, The MathWorks, Inc., Natick, MA, USA) to fit the two-pole Debye model [Reference Salahuddin, Porter, Krewer and O’ Halloran21]. The W-LSM was proposed by Fujii [Reference Fujii33] as a numerical fitting technique. The W-LSM assigns weights to the data points being fitted. These weights are used to emphasize or de-emphasize certain points in the fitting process, depending on their relative importance or reliability [Reference Fujii33]. A complex weight factor $\,\gamma $ was introduced to facilitate the control and thus enhance the accuracy of the fit on the measured dielectric properties data. The error was chosen to be dependent on the permittivity itself and is modeled as follows:

(4)\begin{equation}\,{\dot E_i} \approx \,{\left\{ {\hat \varepsilon \left( {{\omega _i}} \right)} \right\}^\gamma },\end{equation}

where $\hat \varepsilon \left( {{\omega _i}} \right)$ is the measured complex permittivity at the frequency ${\omega _i}$. The weight factor in this study was kept at 1 to assign equal weight to all permittivity values. The nonlinear least squared method was used with the Newton iterative method to minimize the total weighted and the squared error as follows [Reference Salahuddin, Porter, Krewer and O’ Halloran21]:

(5)\begin{equation}{E^2} = \,\mathop \sum \limits_{i = 1}^{{N_f}} \left[ {\frac{{{{\left\{ {{C_{r\,}}\left( {{\omega _i}} \right) - \,{M_{r\,}}\left( {{\omega _i}} \right)} \right\}}^2}}}{{\,\,\,\,\,{{\left\{ {{E_r}\left( {{\omega _i}} \right)} \right\}}^{2\,}}}} + \,\frac{{{{\left\{ {{C_{i\,}}\left( {{\omega _i}} \right) - \,{M_{i\,}}\left( {{\omega _i}} \right)} \right\}}^2}}}{{\,\,\,\,\,{{\left\{ {{E_i}\left( {{\omega _i}} \right)} \right\}}^{2\,}}}}\,} \right],\end{equation}

where ${N_f}$ denotes the number of frequency points, ${C_{r\,}}\left( {{\omega _i}} \right)$ and ${C_{i\,}}\left( {{\omega _i}} \right)$ are the real and imaginary parts of the calculated complex permittivity, respectively, ${M_{r\,}}\left( {{\omega _i}} \right)$ and ${M_{i\,}}\left( {{\omega _i}} \right)\,$are the real and imaginary parts of the measured complex permittivity, respectively, and ${E_r}\left( {{\omega _i}} \right)$ and ${E_i}\left( {{\omega _i}} \right)$ are the real and imaginary parts of the allowable error ${\dot E_i}$, respectively. The two-pole Debye model was fitted to the mean dielectric properties of the ovine adrenal medulla and cortex tissues over the measured frequency range (0.5–8.5 GHz). The mean square error (MSE) was calculated between the mean measured dielectric properties and the dielectric properties resulting from the two-pole Debye model. The MSE is a commonly used metric to evaluate the performance of a model by measuring the difference between the predicted and actual values. It represents the average of the squared differences between the predicted and actual values and therefore indicates how well the model fits the data. The MSE can be calculated as follows:

(6)\begin{equation}MSE = \,\frac{1}{n}\, \times \left( {\mathop \sum {{\left( {{\varepsilon _i} - \,{{\hat \varepsilon }_i}} \right)}^2}} \right),\end{equation}

where $n$ denotes the number of data points, ${\varepsilon _i}$ is the ith actual value, ${\hat \varepsilon _i}$ is the ith predicted value, and Σ denotes the sum of the squared differences between predicted and actual values.

Literature review methodology

A systematic review of the literature was conducted to collate the dielectric properties of adrenal glands published in previous studies. The databases including PubMed, Scopus, and Web of Science were included to search for studies that reported the dielectric properties of adrenal glands. The search was conducted using the keywords “adrenal gland”, “dielectric properties”, and “microwave ablation”. After screening the titles and abstracts of the identified studies, the full texts of the studies that met the inclusion criteria were retrieved. Only the studies that reported the dielectric properties of adrenal glands are included, and studies that did not provide sufficient information about the dielectric properties are excluded. All studies that reported the dielectric properties of other tissues or organs were also excluded.

Dielectric properties of adrenal medulla

The relative permittivity and conductivity of the ovine adrenal medulla measured in this study are shown in Fig. 4(a) and (b), respectively. The measured dielectric properties are compared with the dielectric properties of the adrenal medulla from bovine and ovine reported by Shahzad et al. [Reference Shahzad, Clausing, Prakash, Dennedy and O’halloran15] and Bottiglieri et al. [Reference Bottiglieri, Shahzad, Donlon, Dennedy, Lowery, O'Halloran and Farina11], respectively. The solid lines in Fig. 4 show the mean dielectric properties of the adrenal medulla from each study, while the error bars on the solid lines of dielectric properties of the adrenal medulla measured in this study indicate the range of variation in dielectric properties observed among the measured samples. The comparative analysis indicates the following:

Figure 4. Comparison of ex vivo dielectric properties (a) relative permittivity and (b) conductivity for the adrenal medulla from reported studies. The dielectric properties of the adrenal medulla from ovine, bovine, and porcine do not exhibit significant variation in results.

1. 1. Differences in dielectric properties of the adrenal medulla are observed for samples acquired from different species: bovine and ovine.

2. 2. The mean relative permittivity and conductivity of the ovine adrenal medulla measured in this study were compared to those reported by Bottiglieri et al. [Reference Bottiglieri, Shahzad, Donlon, Dennedy, Lowery, O'Halloran and Farina11]. The results show an average percentage difference of 1.60% and 1.55% for relative permittivity and conductivity, respectively.

3. 3. There is a notable difference between the dielectric properties of the bovine and ovine adrenal medulla measured in this study. The results show an average percentage difference of 5.19% and 1.80% for relative permittivity and conductivity, respectively.

Dielectric properties of adrenal cortex

The relative permittivity and conductivity of the ovine adrenal cortex measured in this study are shown in Fig. 5(a) and (b), respectively. The measured dielectric properties are compared with the dielectric properties of the adrenal cortex from porcine, bovine, ovine, and human reported by Peyman and Gabriel [Reference Peyman and Gabriel24], Shahzad et al. [Reference Shahzad, Clausing, Prakash, Dennedy and O’halloran15], and Bottiglieri et al. [Reference Bottiglieri, Shahzad, Donlon, Dennedy, Lowery, O'Halloran and Farina11], respectively. The solid lines in Fig. 5 show the mean dielectric properties of the adrenal cortex from each study, while the error bars on the solid lines of dielectric properties of the adrenal cortex measured in this study indicate the range of variation in dielectric properties observed among the measured samples. The comparative analysis indicates the following:

Figure 5. Comparison of ex vivo dielectric properties (a) relative permittivity and (b) conductivity for the adrenal cortex from reported studies. The dielectric properties of the adrenal cortex in human (Pheochromocytoma), ovine, bovine, and porcine do not exhibit significant variation in results.

1. 1. Differences in dielectric properties of the adrenal cortex are observed for samples acquired from different species: porcine, bovine, ovine, and human.

2. 2. The mean relative permittivity and conductivity of the ovine adrenal cortex measured in this study were compared to those reported by Bottiglieri et al. [Reference Bottiglieri, Shahzad, Donlon, Dennedy, Lowery, O'Halloran and Farina11]. The results show an average percentage difference of 6.12% and 3.35% for relative permittivity and conductivity, respectively.

3. 3. The relative permittivity of the human adrenal cortex is found to be higher than the relative permittivity of ovine, bovine, and porcine samples.

4. 4. There is a notable difference between the dielectric properties of the bovine and ovine adrenal cortex measured in this study. The results show an average percentage difference of 8.76% and 13.83% for relative permittivity and conductivity, respectively.

5. 5. There is a notable difference between the dielectric properties of the porcine and ovine adrenal cortex measured in this study. The results show an average percentage difference of 10.75% and 12.83% for relative permittivity and conductivity, respectively.

6. 6. The dielectric properties of the bovine and porcine adrenal cortex did not exhibit significant differences. The results show an average percentage difference of 2.04% and 4.40% for relative permittivity and conductivity, respectively.

Ovine, bovine, porcine, and human adrenal glands showed interspecies differences in their dielectric properties, which can be attributed to several factors. One major factor in these differences is the difference in tissue composition. Ovine, bovine, and human adrenal glands differ in the relative amounts of various components such as water content, proteins, lipids, and ions [Reference Grossi, Leifsson, Jensen, Vainer and Iburg34, Reference Di Meo, Bonello, Farhat, Farrugia, Pasian, Camilleri Podesta, Suleiman, Calleja-Agius and Sammut35]. These differences can affect the dielectric properties of the tissue, as each component has a unique response to an electric field. Another factor is the difference in tissue structure. The arrangement of cells, extracellular matrix, and blood vessels can vary across species, which can affect how the tissue responds to an electric field [Reference Perysinakis, Aggeli, Kaltsas and Zografos7, Reference Carbone, Borges, Eiden, García and Hernández-Cruz36, Reference Peyman and Gabriel37]. For example, the size and shape of cells influence their orientation in an electric field, and the presence of blood vessels can alter the conductivity of the tissue. Moreover, Meo et al. [Reference Di Meo, Bonello, Farhat, Farrugia, Pasian, Camilleri Podesta, Suleiman, Calleja-Agius and Sammut35] reported interspecies differences between porcine, ovine, bovine, and human for liver and kidney tissues. The authors found that the differences in dielectric properties between the two organs in animals compared to the two organs in the human cadaver are relatively high. Salahuddin et al. [Reference Salahuddin, La Gioia, Elahi, Porter, O’Halloran and Shahzad38] reported that there is more than 25% variation between the dielectric properties of the same tissue for different species [Reference Salahuddin, La Gioia, Elahi, Porter, O’Halloran and Shahzad38]. Finally, the methodology for characterizing the dielectric properties of bovine and ovine by Shahzad et al. [Reference Shahzad, Clausing, Prakash, Dennedy and O’halloran15] and Bottiglieri et al. [Reference Bottiglieri, Shahzad, Donlon, Dennedy, Lowery, O'Halloran and Farina11], respectively, is the same compared to this study, and therefore, all studies would have reported temperature and sample preparation so these factors can be discounted and differences in dielectric properties can be attributed to samples. As we show in the later section, the temperature at which samples are measured plays a significant role in the obtained dielectric properties values. Thus, we recommend always reporting the sample temperature for future analysis and better comparison between studies. Understanding these differences is important for the development of diagnostic and therapeutic tools for adrenal gland-related diseases in different animal models and human patients.

Table 2. Summary of the parameters of the two-pole Cole–Cole model fitted to the measured adrenal data (cortex and medulla) in the frequency range 0.5–8.5 GHz from the literature

Fitting model on ex vivo dielectric properties data of adrenal glands

This study collated the Cole–Cole models of porcine, bovine, and ovine adrenal glands published in previous studies. All the reported studies utilized the two-pole Cole–Cole model to fit their respective ex vivo dielectric properties data. The models generally showed similar trends in the dielectric properties of adrenal glands, such as the frequency dependence of complex permittivity and conductivity. The collation of the Cole–Cole models is important for developing a robust model for analyzing the ex vivo dielectric properties data of adrenal glands and for understanding their physiological and pathological characteristics. The Cole–Cole models from the previous studies are tabulated in Table 2.

In this study, the two-pole Debye model was fitted to the mean dielectric properties of the ovine adrenal medulla and cortex. Figure 6(a) and (b) shows the measured and calculated dielectric properties from the two-pole Debye model of the adrenal medulla and cortex over a frequency range of 0.5–8.5 GHz. The measured dielectric data of the adrenal medulla and cortex are shown as dotted lines, while the calculated data from the two-pole Debye model are shown as solid lines. The comparison between the measured and calculated dielectric properties data provides an assessment of the accuracy of the two-pole Debye model in describing the dielectric properties of the tissue. In this case, the two-pole Debye model appears to provide a good fit to the measured data, with the calculated data closely following the trends of the experimental data. The relative permittivity of the adrenal medulla and cortex both increase with frequency, while the conductivity of the adrenal medulla is higher than that of the adrenal cortex at all frequencies.

Figure 6. Two-pole Debye model parameters of adrenal medulla and cortex over 0.5–8.5 GHz frequency band: (a) relative permittivity and (b) conductivity. The measured dielectric data of the adrenal medulla and cortex (dotted lines) are compared with calculated data from the two-pole Debye model (solid lines).

The parameters of the two-pole Debye model fitted to the mean dielectric properties of the adrenal medulla and cortex are presented in Table 3. The results showed that the two-pole Debye model was able to accurately capture the complex dielectric behavior of adrenal tissues. These parameters will aid the broadband EM simulations of adrenal gland tissues for the development of EM diagnostic tools, such as imaging techniques or other applications.

Table 3. Two-pole Debye model parameters fitted to the measured adrenal data (medulla and cortex) in the frequency range 0.5–8.5 GHz

Table 4. Mean-squared error value of the fit error. The error values for the fit of each adrenal tissue type (medulla and cortex) are calculated as the mean value across all frequency points

In this study, the MSE was used to quantify the difference between the measured dielectric properties and the calculated values by the two-pole Debye model across all frequency points. The MSE values are tabulated in Table 4. The MSE was calculated for both the relative permittivity and conductivity values of the adrenal medulla and cortex. It can be observed from Table 4 that the MSE values vary depending on the tissue type. Moreover, it can be observed from Table 4 that the model was found to have a good fit for the experimental data of the adrenal cortex compared to the medulla, with low MSE values for both the relative permittivity and conductivity.

Temperature-dependent dielectric properties of adrenal medulla

This study has also investigated the temperature-dependent dielectric properties of the ovine adrenal medulla for temperatures ranging from 37°C to 64°C. Due to the shrinking width of the adrenal cortex tissue with increasing temperature, it was challenging to achieve proper contact between the OECL probe and the adrenal cortex tissue. Therefore, the temperature-dependent dielectric properties of the adrenal cortex were not taken into consideration. Fig. 7(a) and (b) shows the temperature-dependent relative permittivity and conductivity of the ovine adrenal medulla over a frequency range of 0.5–8.5 GHz, respectively. The dielectric properties were recorded at 37°C, 46°C, 53°C, and 64°C. The solid lines in Fig. 7 show the mean dielectric properties of the adrenal medulla at each temperature, while the error bars indicate the range of variation in dielectric properties. It can be observed from Fig. 7 that the dielectric properties of the ovine adrenal medulla decrease with increasing temperature. The dielectric properties tend to drop quickly after 46°C and continue to drop as the temperature is increased further and the tissue becomes more dehydrated [Reference Kwilas, Donahue, Tsang and Hodge39, Reference Bottiglieri, Farina, Shahzad, Orloughlin, Orhalloran and Elahi40]. Moreover, at low temperatures, the dielectric properties of the tissue are higher, which means that more energy is absorbed by the tissue when exposed to microwaves. This behavior can lead to faster heating of the tissue and more efficient ablation. The observed temperature-dependent changes in the dielectric properties of the ovine adrenal medulla have important implications for the development of MWA technologies for hypertension treatment. The findings suggest that careful consideration of the temperature-dependent dielectric properties of the tissue is necessary to achieve an efficient and safe MWA.

Figure 7. Temperature-dependent dielectric properties (a) relative permittivity and (b) conductivity for the adrenal medulla over 0.5–8.5 GHz frequency band. The temperature of the adrenal gland sample was raised to 37°C, 46°C, 53°C, and 64°C.

Choice of animal model for preclinical testing of microwave ablation devices for adrenal ablation – a perspective from dielectric properties

The choice of an animal model for preclinical testing of MWA devices for adrenal ablation is crucial as it can greatly affect the accuracy and reliability of the results obtained. When selecting an animal model for this purpose, one important consideration is the dielectric properties of the adrenal gland and surrounding tissues. Different tissues have different dielectric properties, which can affect the way they interact with EM fields, such as those produced by MWA devices [Reference Amin, Elahi, Shahzad, Porter and O’Halloran41, Reference Vidjak, Farina, Ruvio, O’Halloran and Cavagnaro42]. Therefore, selecting an animal model that closely mimics the dielectric properties of human adrenal tissue is important for accurate preclinical testing.

Based on the comparison of dielectric properties of the adrenal medulla and cortex in Figs. 4 and 5, respectively, the dielectric properties of human adrenal glands showed notable differences compared to ovine, bovine, and porcine adrenal glands. The relative permittivity of the human adrenal cortex from a Pheochromocytoma patient was found to be in agreement with the relative permittivity of the bovine adrenal cortex. Therefore, from the perspective of dielectric properties, it can be concluded that no animal model can fully replicate the dielectric properties of human adrenal glands. To this end, factors such as the size and dielectric properties of the adrenal gland in the chosen animal model must be carefully considered to ensure that the results of the testing are relevant to the human application of the device. Moreover, while preclinical testing in animal models can provide valuable insights into the safety and efficacy of MWA devices for adrenal ablation, it is essential to confirm the results in human clinical trials.