Simulation material capable of simultaneously simulating electrical characteristic and thermal characteristic parameters of biological tissue and configuration method thereof

Abstract

The invention relates to a simulation material capable of simultaneously simulating electric heating characteristics of biological tissues and a configuration method thereof, and the configuration method comprises the following steps: calculating component concentration of the simulation material according to electric characteristics, thermal characteristics and temperature conditions of target tissues and a fitting model, and preparing a target simulation material; the fitting model is the relationship between the electrothermal performance parameters of the simulation material and the concentrations of ethylene glycol, aluminum oxide and sodium chloride forming the simulation material at different temperatures; the construction of the fitting model comprises the steps of preparing a plurality of simulation materials as samples, performing electric heating performance measurement on the prepared samples, and constructing the fitting model based on the component concentration of the simulation materials and the material characteristics according to the samples and the electric heating performance parameters of the samples. The component concentration can be accurately calculated according to the target tissue characteristics and temperature conditions, and a customized simulation material is prepared.

Description

Technical Field

[0001] This invention relates to the field of materials for simulating the unique properties of biological tissues, and in particular to a method for preparing a simulated material that can simultaneously simulate the electrothermal properties of biological tissues, as well as the prepared simulated material. Background Technology

[0002] Radiofrequency ablation is a minimally invasive treatment technique whose core mechanism utilizes the Joule heating effect generated by high-frequency alternating current to selectively destroy target tissue. Temperature is a key factor determining the degree of tissue damage during this process; effective ablation of tumor tissue can be achieved when the local tissue temperature reaches a critical threshold of (65±5)℃. The generation and transfer of heat during radiofrequency ablation are constrained by both the electrical and thermal properties of the tissue. Electrical properties include electrical conductivity and relative electrical constant, while thermal properties include thermal conductivity, thermal diffusivity, and volumetric heat capacity. Therefore, in clinical applications, to achieve accurate assessment of the internal thermal field distribution during radiofrequency ablation, it is essential to comprehensively acquire and analyze the aforementioned electrical and thermal parameters.

[0003] Tissue-simulating materials (phalanxes) have been widely used in medical research as alternatives to biological tissues. Their core design philosophy lies in accurately simulating the key physical properties of the target tissue. Compared to natural biological tissues, these artificial materials offer numerous advantages, including longer shelf life and better uniformity. They can accurately assess the thermal field distribution and energy deposition characteristics during radiofrequency ablation, and ensure strict controllability of experimental conditions and high reproducibility of results. In the field of radiofrequency ablation research, various tissue-simulating materials have been developed over the past few decades to evaluate key parameters such as temperature distribution and ablation volume. One typical material uses a gel matrix composite with thermochromic components; when the temperature reaches a preset threshold, a significant color change occurs, visually indicating the ablation area. However, these materials fail to replicate the electrical and thermal properties of biological tissues, resulting in significant differences in heat generation and transfer mechanisms during radiofrequency ablation compared to real tissues. Several other studies have made significant progress in the development of tissue-simulating materials, but all have limitations to varying degrees. 1. The reusable thermal phantom developed by Dabbagh A et al. has achieved accurate simulation of thermal properties (specific heat capacity, thermal conductivity, etc.) and acoustic properties, but it does not have the ability to simulate electrical properties. 2. Although the thermochromic model developed by Negussie AH et al. successfully matched thermal characteristic parameters such as thermal conductivity, it also failed to solve the problem of simulating electrical characteristic parameters. 3. Kuroda M et al. achieved frequency correlation simulation of electrical conductivity based on a carrageenan-gellan gum mixed gel system by adjusting the potassium chloride concentration, but this material has no ability to simulate thermal properties. 4. Although the liver phantom based on polyacrylamide gel by Bu-Lin Z et al. simulated tissue conductivity and volumetric heat capacity, it only simulated a single electrical characteristic parameter and a single thermal characteristic parameter, and could not fully characterize the heat generation and heat transfer effects during radiofrequency ablation.

[0004] Comprehensive analysis shows that although radiofrequency ablation research requires simulation materials that can comprehensively characterize the electrothermal properties of biological tissues, no tissue simulation material has yet been developed that can simultaneously and accurately simulate the electrical and thermal properties of multiple biological tissues for use in radiofrequency ablation technology. This technological bottleneck has become a key scientific problem restricting the development of radiofrequency ablation research.

[0005] To address this technical challenge, it is essential to provide a method for preparing a simulation material that can simultaneously mimic the electrothermal properties of biological tissues, as well as the simulation material itself, to overcome the shortcomings of existing technologies. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a method for preparing tissue simulation materials for evaluating the temperature control performance of radiofrequency ablation devices. The simulation materials prepared by this method can simultaneously and accurately simulate the electrical and thermal properties of specific biological tissues, providing a reliable solution for performance testing of medical devices such as electrosurgery and radiofrequency ablation, and have significant clinical application value.

[0007] The objective of this invention is achieved through the following technical measures.

[0008] A method for preparing a simulated material that can simultaneously simulate the electrothermal properties of biological tissue is provided. Based on the electrical properties, thermal properties, and temperature conditions of the target tissue, the component concentration of the simulated material is calculated according to a fitting model, and the target simulated material is prepared based on the calculated component concentration. The fitting model represents the relationship between the electrothermal performance parameters of the simulated material and the contents of ethylene glycol, alumina, and sodium chloride that constitute the simulated material at different temperatures. The construction of the fitting model includes preparing multiple simulated materials as samples, measuring the electrothermal properties of the prepared samples, and constructing the fitting model based on the sample composition concentration and material properties according to the sample and the sample electrothermal property parameters.

[0009] Preferably, in the above-mentioned method for preparing the simulation material that can simultaneously simulate the electrothermal properties of biological tissue, the simulation material uses polyacrylamide gel, agar gel, or gellan gel as the base matrix, and is composed of ethylene glycol, alumina, and sodium chloride. The content of ethylene glycol is not higher than 80% (v / v), the content of alumina is not higher than 6% (v / v), and the content of sodium chloride is 0.1-0.8% (w / v).

[0010] Preferably, in the above-mentioned method for preparing a simulated material that can simultaneously simulate the electrothermal properties of biological tissue, the simulated material is a polyacrylamide gel-based simulated material, and the preparation process is as follows: S1, Preparation of the precursor solution, specifically: For samples with an ethylene glycol volume fraction of less than 65%: Dissolve 40% (w / v) acrylamide / N,N'-methylenebisacrylamide in deionized water at a mass ratio of 19:1; or For samples with an ethylene glycol volume fraction of less than 65%, dissolve 16% (w / v) acrylamide / N,N'-methylenebisacrylamide in EG solvent at a mass ratio of 19:1. S2, the preparation of the mixed solution is as follows: Measure the precursor solution and transfer it to the reaction vessel, ensuring that the total mass of acrylamide / N,N'-methylenebisacrylamide accounts for 10% of the total volume of the final simulated material; add ethylene glycol and deionized water; S3, stirring and dispersing, specifically: Place the mixture on a magnetic stirrer and add sodium chloride and aluminum oxide in sequence. Add the aluminum oxide in steps. After the aluminum oxide has been completely added, continue to stir magnetically for 10-30 minutes, and then disperse ultrasonically at 40-60 kHz for 10-15 minutes. S4 initiates the polymerization reaction, specifically: 0.1% (w / v) of ammonium persulfate was added as a polymerization initiator, and 0.1% (v / v) of tetramethylethylenediamine was added to accelerate polymerization; S5, stop stirring and sealing: immediately stop magnetic stirring after adding TEMED, and seal the container with plastic wrap to prevent dehydration; S6, Polymerization and Preservation: Specifically, the sample is left to stand at room temperature until polymerization is complete, forming a simulated material sample with the final structure; the sample is then preserved at 4°C.

[0011] Another preferred embodiment of the above-mentioned method for preparing a simulated material capable of simultaneously simulating the electrothermal properties of biological tissues, wherein the simulated material is an agar gel-based simulated material, and the preparation process is as follows: S1, solvent premixing and addition, specifically: premix deionized water and ethylene glycol, heat in a water bath to 85-90℃ and keep the temperature constant; S2, add agar powder at a concentration of 1.5-2.0% (w / v) after addition, stir magnetically, and maintain the temperature at 85-90℃ to dissolve for 10-20 minutes until completely transparent; S3, stirring and dispersing, specifically: Cool to 60℃-62℃, then add sodium chloride and aluminum oxide in sequence, adding the aluminum oxide in steps. After the aluminum oxide is completely added, continue magnetic stirring for 10-30 minutes, then ultrasonically disperse at 40-60kHz for 10-15 minutes to ensure uniform particle dispersion. S4, gel molding, specifically: physical cross-linking is completed by standing at room temperature for 30 minutes, then refrigerated at 4℃ for 2 hours to enhance mechanical strength, and sealed and stored in a 0.9% physiological saline moist environment; Another preferred embodiment is the method for preparing the above-mentioned simulation material capable of simultaneously simulating the electrothermal properties of biological tissues. The simulation material is based on gellan gel, and the preparation process is as follows: S1, solvent premixing system, specifically: premixed deionized water and ethylene glycol, with 0.1% sodium citrate added as a chelating agent; S2, the high-temperature dissolution stage, specifically involves heating in a water bath to a constant temperature of 85-90℃, adding 0.5% (w / v) low-acyl gellan gum, and magnetically stirring until completely transparent; S3, stirring and dispersion, specifically: cool to 60℃-62℃, add sodium chloride, aluminum oxide and 0.8 mmol / L CaCl2 solution in sequence, then stir magnetically for 10-30 minutes, and disperse ultrasonically at 40-60 kHz for 10-15 minutes to ensure that the particles are fully dispersed and the cross-linking is uniform. S4, gel molding, specifically: cool and solidify at room temperature for 2 hours, then immerse in physiological saline containing 0.1% CaCl2 for storage. Avoid freezing during storage.

[0012] Furthermore, the above-mentioned method for preparing simulated materials that can simultaneously simulate the electrothermal properties of biological tissues involves measuring the electrical properties of the prepared samples, including measuring one or more parameters such as conductivity, electrical performance constant, and relative electrical performance constant. The thermal properties of the prepared sample are measured, including one or more parameters such as thermal conductivity, thermal diffusivity, and volumetric heat capacity. The electrical and thermal properties of the samples were tested at multiple radiofrequency ablation temperatures.

[0013] Preferably, the above-mentioned method for configuring the simulation material capable of simultaneously simulating the electrothermal properties of biological tissues, by establishing the fitting model through variance analysis combined with least squares fitting, specifically includes: 1) Construction of the fitting algorithm and data acquisition, specifically: Experimental data on the corresponding electrothermal characteristic parameters were collected at multiple temperature points for a series of prepared simulated material samples. The least squares regression fitting algorithm was used to establish a quantitative relationship between the electrical and thermal characteristic parameters and the experimental factors, namely, the content of ethylene glycol, the content of alumina, and the content of sodium chloride. 2) Analysis of variance, specifically: For each electrothermal performance parameter, an analysis of variance was performed independently under each temperature condition. The independent variables of the analysis of variance model included three main factors: ethylene glycol content, alumina content, and sodium chloride content, as well as their pairwise interactions: ethylene glycol × sodium chloride, ethylene glycol × alumina, and sodium chloride × alumina. The significance determination scheme is as follows: using p≤0.05 as the threshold, the significant influencing factors of each parameter under each temperature condition are identified, and the independent and synergistic contributions of each factor to the electrical and thermal properties of the tissue simulation material sample are revealed. 3). Regression model development and validation, specifically: Model construction: Based on the results of the analysis of variance, only significant factors and interactions are included to establish a series of least squares regression models corresponding to different parameter-temperature combinations. The established regression models are the fitted models.

[0014] Another preferred embodiment of the above-mentioned method for configuring the simulated material capable of simultaneously simulating the electrothermal properties of biological tissues, wherein the fitting model is established by fitting an ANN neural network, specifically includes: 8.1 Data Preparation and Preprocessing To establish an effective fitting model, experimental data must first be prepared and preprocessed. The experimental data includes the electrical conductivity, thermal conductivity, volumetric heat capacity, and corresponding temperature values ​​of the simulated material. The specific steps are as follows: (1) Data collection and organization In this experiment, 27 sample phantoms were prepared. The input characteristics of each sample included: electrical conductivity (measured at 480 kHz), thermal conductivity, volumetric heat capacity, and temperature (including four temperature points: 45°C, 60°C, 75°C, and 90°C). The data for each sample includes test results at the four temperatures mentioned above, with the corresponding target output being the concentrations of three components: the volume fraction of ethylene glycol, the volume fraction of alumina, and the mass concentration of sodium chloride. (2) Data normalization To prevent poor model training performance due to differences in the dimensions of different features, a Min-Max normalization method is used to process the data. Specifically, the input features are normalized to map the data to the [0,1] interval, and the output targets (concentrations of ethylene glycol, alumina, and sodium chloride) are also normalized.

[0015] (3) Dataset partitioning The experimental dataset was divided into three parts: training set (70% of the data) for training the neural network; validation set (15% of the data) for adjusting hyperparameters and preventing overfitting during training; and test set (15% of the data) for evaluating the model's generalization ability.

[0016] 8.2 Neural Network Model Construction and Training Based on the prepared data, construct and train the neural network model. The specific steps are as follows: (1) Model structure design This invention uses a five-layer feedforward neural network for modeling, and the network structure is as follows: Input layer: 4 nodes, corresponding to electrical conductivity, thermal conductivity, volumetric heat capacity and temperature respectively.

[0017] Hidden layer: The first hidden layer uses the Logsig activation function, which can handle non-linear relationships in the data.

[0018] The second hidden layer employs the Poslin activation function, which ensures the non-negativity of the output and helps maintain the interpretability of the physics.

[0019] The third hidden layer employs the Purelin activation function, allowing the network to model linear relationships and ensuring the linear interpretability of the output results.

[0020] Output layer: 3 nodes, corresponding to the concentration prediction of ethylene glycol, alumina and sodium chloride respectively.

[0021] (2) Optimization algorithms and regularization Optimization algorithm: The Levenberg-Marquardt algorithm is selected to optimize network parameters, combined with gradient clipping to prevent gradient explosion and ensure the stability of the optimization process.

[0022] Loss function: Optimizes model performance by minimizing mean squared error (MSE), thereby ensuring that the model can accurately fit the data.

[0023] Regularization measures: Introduce L2 weight decay and Dropout layers to reduce overfitting and improve the model's generalization ability.

[0024] (3) Training process Normalized training data is input into the model for multiple training rounds. After each round, the model is evaluated using validation set data, and network parameters (learning rate, number of hidden layer nodes, etc.) are adjusted. During training, early stopping is used to monitor the loss value on the validation set to prevent overfitting.

[0025] 8.3 Model Validation and Evaluation After training the neural network model, it needs to be validated and evaluated. The specific steps are as follows: (1) Performance evaluation The model's performance was evaluated using a test set, and the root mean square error and coefficient of determination were calculated.

[0026] (2) Sensitivity analysis By using SHAP value analysis or weighted analysis, the influence of each input feature on the output concentration can be assessed. This method clearly identifies which features have a greater impact on model predictions.

[0027] (3) Experimental verification Based on the optimal ratio predicted by the ANN model, corresponding phantom samples were prepared. The electrothermal parameters of the samples were measured, and the experimental values ​​were compared with the model predictions to verify the practical application effect of the model.

[0028] Specifically, the above-mentioned method for preparing simulated materials capable of simultaneously mimicking the electrothermal properties of biological tissues involves preparing simulated material samples based on polyacrylamide gel, and preparing samples with EG contents of 0%, 40%, and 80% v / v. A total of 27 simulated material samples were obtained, with contents of 0%, 3%, and 6% v / v and NaCl contents of 0.2%, 0.4%, and 0.6% w / v. The electrothermal performance parameters, namely electrical conductivity, thermal conductivity, and volumetric heat capacity, were measured for each sample at temperature points of 45℃, 60℃, 75℃, and 90℃. Based on the composition content of the samples and the corresponding electrothermal performance parameters, a fitting model was established by combining analysis of variance with least squares fitting. The regression equations for the three parameters at the four temperature points are shown in Table 1.

[0029] This invention also provides a simulation material that can simultaneously simulate the electrical and thermal properties of biological tissues. It uses polyacrylamide gel, agar gel, or gellan gel as the base matrix and is composed of ethylene glycol, alumina, and sodium chloride. The content of ethylene glycol is greater than 0 and not more than 80% (v / v), the content of alumina is greater than 0 and not more than 6% (v / v), and the content of sodium chloride is 0.1-0.8% (w / v).

[0030] This invention discloses a method for preparing a simulation material capable of simultaneously simulating the electrothermal properties of biological tissues. The simulation material uses polyacrylamide gel / agar gel / gellan gel as the base material and introduces three components with adjustable concentrations. By precisely controlling their concentrations, both electrical and thermal properties can be accurately simulated simultaneously. This method overcomes the limitations of existing technologies that typically employ fixed formulations or single-variable adjustments. It innovatively achieves more comprehensive tissue simulation through multi-component concentration control, thus solving the problem of incomplete simulation in existing technologies.

[0031] The present invention selects four representative temperature points within a specific temperature range of radiofrequency ablation to establish material formulations at different temperatures, covering a wide temperature range application scenario for radiofrequency ablation. Through multi-temperature point design, high-precision tissue simulation materials can be prepared within the actual working temperature range of radiofrequency ablation.

[0032] This invention establishes a fitting model relating the electrothermal performance parameters of the simulated material to the concentrations of ethylene glycol, alumina, and sodium chloride constituting the simulated material at different temperatures. Based on the target microstructure and temperature conditions, the component concentrations can be accurately calculated to prepare customized simulated materials.

[0033] This invention makes material preparation more flexible by establishing an adjustable formulation system, and combines wide-temperature-range data modeling with multi-method verification to make up for the shortcomings of poor temperature adaptability or weak interpretability in the prior art. Attached Figure Description

[0034] The invention will be further described with reference to the accompanying drawings, but the contents of the drawings do not constitute any limitation on the invention.

[0035] Figure 1 This is a schematic diagram of a method for preparing a simulated material that can simultaneously simulate the electrothermal properties of biological tissue according to the present invention.

[0036] Figure 2 This is a schematic diagram of the simulated material prepared in one embodiment 2 of the present invention.

[0037] Figure 3 This is a schematic diagram of the equipment used for conductivity measurement in Example 2.

[0038] Figure 4 This is a schematic diagram of the dual-needle sensor.

[0039] Figure 5 This is a schematic diagram showing the sensor being fully embedded in the PAG sample for detection. Detailed Implementation

[0040] The present invention will be further described in conjunction with the following embodiments.

[0041] Example 1 A method for preparing a simulated material that can simultaneously simulate the electrothermal properties of biological tissues involves calculating the component concentration of the simulated material based on the electrical properties, thermal properties, and temperature conditions of the target tissue, according to a fitted model, and then preparing the target simulated material based on the calculated component concentration.

[0042] The fitting model represents the relationship between the electrothermal performance parameters of the simulated material and the contents of ethylene glycol, alumina, and sodium chloride that make up the simulated material at different temperatures.

[0043] The construction of the fitting model includes preparing multiple simulated materials as samples, measuring the electrothermal properties of the prepared samples, and constructing the fitting model based on the sample composition concentration and material properties according to the sample and the sample electrothermal property parameters.

[0044] The target tissue is biological tissue. To study its electrical and thermal properties under different temperature conditions, it is necessary to simulate the target tissue using a simulation material. By analyzing the electrothermal properties of the simulation material under different conditions, the electrothermal properties of the target tissue under those conditions can be estimated. Therefore, the closer the simulation material is to the target tissue, the more accurate the prediction results will be.

[0045] When the method of the present invention is performed for the first time, the process is as follows: Figure 1 As shown, multiple simulated materials were first prepared as samples. The electrothermal properties of the prepared samples were measured, and a fitting model based on the sample composition concentration and material properties was constructed according to the samples and their electrothermal property parameters. After obtaining the fitting model, the component concentration of the simulated material was calculated based on the electrical properties, thermal properties, and temperature conditions of the target tissue, using the fitting model. The target simulated material was then prepared based on the calculated component concentration.

[0046] Once the fitting model has been established, when using it, only the fourth step is needed: calculate the component concentration of the simulated material based on the electrical properties, thermal properties, and temperature conditions of the target tissue, and then prepare the target simulated material based on the calculated component concentration.

[0047] It should be noted that electrical property measurements include one or more parameters such as electrical conductivity, electrical performance constant, and relative electrical performance constant, while thermal property measurements include one or more parameters such as thermal conductivity, thermal diffusivity, and volumetric heat capacity.

[0048] The simulated material uses polyacrylamide gel, agarose gel, or gellan gel as the base matrix, and is composed of ethylene glycol, alumina, and sodium chloride. The content of ethylene glycol is no higher than 80% (v / v), the content of alumina is no higher than 6% (v / v), and the content of sodium chloride is 0.1-0.8% (w / v). This invention uses polyacrylamide gel / agarose gel / gellan gel as the base matrix to construct a ternary control system composed of ethylene glycol, alumina, and sodium chloride. Through the synergistic effect of each component, precise control of the material's electrical and thermal properties is achieved.

[0049] This configuration method uses a polyacrylamide gel-based simulated material as the simulation material, and the preparation process is as follows: S1, Preparation of the precursor solution, specifically: For samples with an ethylene glycol volume fraction of less than 65%: Dissolve 40% (w / v) acrylamide / N,N'-methylenebisacrylamide in deionized water at a mass ratio of 19:1; For samples with an ethylene glycol volume fraction of less than 65%, dissolve 16% (w / v) acrylamide / N,N'-methylenebisacrylamide in EG solvent at a mass ratio of 19:1.

[0050] S2, the preparation of the mixed solution is as follows: Measure the precursor solution and transfer it to the reaction vessel, ensuring that the total mass of acrylamide / N,N'-methylenebisacrylamide accounts for 10% of the total volume of the final simulated material; add ethylene glycol and deionized water.

[0051] S3, stirring and dispersing, specifically: Place the mixture on a magnetic stirrer and add sodium chloride and aluminum oxide in sequence. Add the aluminum oxide in steps. After the aluminum oxide is completely added, continue magnetic stirring for 10-30 minutes, and then disperse it ultrasonically at 40-60 kHz for 10-15 minutes.

[0052] S4 initiates the polymerization reaction, specifically: 0.1% (w / v) of ammonium persulfate was added as a polymerization initiator, and 0.1% (v / v) of tetramethylethylenediamine was added to accelerate polymerization; S5, stop stirring and sealing: immediately stop magnetic stirring after adding TEMED, and seal the container with plastic wrap to prevent dehydration; S6, Polymerization and Preservation: Specifically, the sample is left to stand at room temperature until polymerization is complete, forming a simulated material sample with the final structure; the sample is then preserved at 4°C.

[0053] In this configuration method, the simulation material can also be an agar gel-based simulation material, and the preparation process is as follows: S1, solvent premixing and addition, specifically: premix deionized water and ethylene glycol, heat in a water bath to 85-90℃ and keep the temperature constant.

[0054] S1. Add agar powder to a concentration of 1.5-2.0% (w / v) after addition, stir magnetically, and maintain the temperature at 85-90℃ for 10-20 minutes to dissolve until completely transparent.

[0055] S3, stirring and dispersing, specifically: Cool down to about 60℃, then add sodium chloride and aluminum oxide in sequence, adding the aluminum oxide in steps. After the aluminum oxide is completely added, continue to stir magnetically for 10-30 minutes, and then disperse ultrasonically at 40-60kHz for 10-15 minutes.

[0056] S4, gel molding, specifically: physical cross-linking is completed by standing at room temperature for 30 minutes, then refrigerated at 4℃ for 2 hours to enhance mechanical strength, and then sealed and stored in a 0.9% physiological saline moist environment.

[0057] The method for preparing a simulated material that can simultaneously simulate the electrothermal properties of biological tissues is as follows. The simulated material can also be a simulated material based on gellan gel.

[0058] S1, the solvent premix system, specifically: premixed deionized water and ethylene glycol, with 0.1% sodium citrate added as a chelating agent.

[0059] S2, the high-temperature dissolution stage, specifically involves heating in a water bath to a constant temperature of 85-90℃, adding 0.5% (w / v) low-acyl gellan gum, and magnetically stirring until completely transparent.

[0060] S3, stirring and dispersion, specifically: cool to about 60℃, add sodium chloride, aluminum oxide and 0.8 mmol / L CaCl2 solution in sequence, and disperse by ultrasonication at 40-60 kHz for 10-15 minutes.

[0061] S4, gel molding, specifically: cool and solidify at room temperature for 2 hours, then immerse in physiological saline containing 0.1% CaCl2 for storage. Avoid freezing during storage.

[0062] This invention prepares multiple simulated materials as samples, with different contents of ethylene glycol, sodium chloride, and alumina in each sample. Electrical properties of these samples are measured, including one or more parameters such as electrical conductivity, electrical performance constant, and relative electrical performance constant. Thermal properties of the prepared samples are measured, including one or more parameters such as thermal conductivity, thermal diffusivity, and volumetric heat capacity.

[0063] The present invention provides a multi-temperature point sampling test scheme, which is designed to test the electrical and thermal characteristics of the sample at multiple radio frequency ablation temperature points (45℃, 60℃, 75℃ and 90℃).

[0064] Conductivity measurements were performed using a testing system based on standard four-electrode technology. The system consisted of an AC power supply, an oscilloscope, a connecting circuit board, and a four-electrode probe equipped with platinum electrodes. During the experiment, an AC current was applied through the two external electrodes, while the voltage difference between the two internal electrodes was recorded using the oscilloscope. All measurements were performed at a frequency of 480 kHz. The system was calibrated using physiological saline at room temperature before measurement. During the measurement, the temperature of the PAG sample was monitored using a temperature probe. When the sample reached the predetermined temperature, the four-electrode probe was carefully applied to the sample. After allowing the oscilloscope readings to stabilize, the average of the input voltage and output current was recorded to determine the sample's conductivity. Each sample was measured three times at marked points, and the average value was taken; the data difference should be less than 5%.

[0065] Thermal property measurements were performed using a detection system based on the dual-probe thermal pulse method. The system included a power supply, a relay module, an AD sampling module, a main control module, and a dual-needle sensor. The dual-needle sensor consisted of parallel probes spaced 6 mm apart: a heating probe (30 mm long, 2.0 mm in diameter) and a temperature-sensing probe (15 mm long, 1.5 mm in diameter). During measurement, the main control module activated the relay to power the heating probe for 10 seconds, after which the temperature-sensing probe continuously monitored temperature changes. Data acquisition continued for 140 seconds after heating was stopped. Temperature was recorded once per second throughout the entire heating-cooling cycle, yielding 150 data points per measurement. The temperature data was transmitted to a computer in real time via a data cable. The thermal conductivity parameters of the sample were calculated using an empirical model. The measurement steps are as follows.

[0066] (1) Insert the dual-needle sensor vertically into the tissue simulation material sample, ensuring that the two probes remain parallel; (2) Start the main control program, set the heating time to 10 seconds and the total sampling time to 150 seconds; (3) Trigger the relay module to apply constant power to the heating needle; (4) The thermometer needle records temperature changes at a frequency of 1Hz and stores the data synchronously in the computer; (5) After completing a single measurement, let the sample stand for 10 minutes to allow the temperature to return to its initial state; (6) Repeat steps 2-5 to obtain a total of three sets of data; (7) Use an algorithm to fit the temperature-time curve and calculate the thermal conductivity parameters; (8) Verify the system drift using a glycerol standard sample before and after each batch measurement.

[0067] The fitting model can be established by combining analysis of variance with least squares fitting, specifically including: 1) Construction of the fitting algorithm and data acquisition, specifically: Experimental data on the corresponding electrothermal characteristic parameters were collected at multiple temperature points for a series of prepared simulated material samples.

[0068] The least squares regression fitting algorithm was used to establish the quantitative relationship between the electrical and thermal characteristic parameters and the experimental factors, namely the content of ethylene glycol, alumina, and sodium chloride.

[0069] 2) Analysis of variance, specifically: For each electrothermal performance parameter, an analysis of variance was performed independently under each temperature condition. The independent variables of the analysis of variance model included three principal factors: ethylene glycol content, alumina content, and sodium chloride content, as well as their pairwise interactions: ethylene glycol × sodium chloride, ethylene glycol × alumina, and sodium chloride × alumina.

[0070] The significance determination scheme is as follows: using p≤0.05 as the threshold, the significant influencing factors of each parameter under each temperature condition are identified, and the independent and synergistic contributions of each factor to the electrical and thermal properties of the tissue simulation material sample are revealed.

[0071] 3). Regression model development and validation, specifically: Model construction: Based on the results of analysis of variance, only significant factors and interactions were included to establish a series of least squares regression models corresponding to different parameter-temperature combinations.

[0072] The fitting model can also be established using an ANN neural network, specifically including: 8.1 Data Preparation and Preprocessing To establish an effective fitting model, experimental data must first be prepared and preprocessed. The experimental data includes the electrical conductivity, thermal conductivity, volumetric heat capacity, and corresponding temperature values ​​of the simulated material. The specific steps are as follows: (1) Data collection and organization In this experiment, 27 sample phantoms were prepared. The input characteristics of each sample included: electrical conductivity (measured at 480 kHz), thermal conductivity, volumetric heat capacity, and temperature (including four temperature points: 45°C, 60°C, 75°C, and 90°C). The data for each sample includes test results at the four temperatures mentioned above, with the corresponding target output being the concentrations of three components: the volume fraction of ethylene glycol, the volume fraction of alumina, and the mass concentration of sodium chloride. (2) Data normalization To prevent poor model training performance due to differences in the dimensions of different features, a Min-Max normalization method is used to process the data. Specifically, the input features are normalized to map the data to the [0,1] interval, and the output targets (concentrations of ethylene glycol, alumina, and sodium chloride) are also normalized.

[0073] (3) Dataset partitioning The experimental dataset was divided into three parts: training set (70% of the data) for training the neural network; validation set (15% of the data) for adjusting hyperparameters and preventing overfitting during training; and test set (15% of the data) for evaluating the model's generalization ability.

[0074] 8.2 Neural Network Model Construction and Training Based on the prepared data, construct and train the neural network model. The specific steps are as follows: (1) Model structure design This invention uses a five-layer feedforward neural network for modeling, and the network structure is as follows: Input layer: 4 nodes, corresponding to electrical conductivity, thermal conductivity, volumetric heat capacity and temperature respectively.

[0075] Hidden layer: The first hidden layer uses the Logsig activation function, which can handle non-linear relationships in the data.

[0076] The second hidden layer employs the Poslin activation function, which ensures the non-negativity of the output and helps maintain the interpretability of the physics.

[0077] The third hidden layer employs the Purelin activation function, allowing the network to model linear relationships and ensuring the linear interpretability of the output results.

[0078] Output layer: 3 nodes, corresponding to the concentration prediction of ethylene glycol, alumina and sodium chloride respectively.

[0079] (2) Optimization algorithms and regularization Optimization algorithm: The Levenberg-Marquardt algorithm is selected to optimize network parameters, combined with gradient clipping to prevent gradient explosion and ensure the stability of the optimization process.

[0080] Loss function: Optimizes model performance by minimizing mean squared error (MSE), thereby ensuring that the model can accurately fit the data.

[0081] Regularization measures: Introduce L2 weight decay and Dropout layers to reduce overfitting and improve the model's generalization ability.

[0082] (3) Training process Normalized training data is input into the model for multiple training rounds. After each round, the model is evaluated using validation set data, and network parameters (learning rate, number of hidden layer nodes, etc.) are adjusted. During training, early stopping is used to monitor the loss value on the validation set to prevent overfitting.

[0083] 8.3 Model Validation and Evaluation After training the neural network model, it needs to be validated and evaluated. The specific steps are as follows: (1) Performance evaluation The model's performance was evaluated using a test set, and the root mean square error and coefficient of determination were calculated.

[0084] (2) Sensitivity analysis By using SHAP value analysis or weighted analysis, the influence of each input feature on the output concentration can be assessed. This method clearly identifies which features have a greater impact on model predictions.

[0085] (3) Experimental verification Based on the optimal ratio predicted by the ANN model, corresponding phantom samples were prepared. The electrothermal parameters of the samples were measured, and the experimental values ​​were compared with the model predictions to verify the practical application effect of the model.

[0086] This invention discloses a method for preparing a simulation material capable of simultaneously simulating the electrothermal properties of biological tissues. The simulation material uses polyacrylamide gel / agar gel / gellan gel as the base material and introduces three components with adjustable concentrations. By precisely controlling their concentrations, it achieves simultaneous and accurate simulation of both electrical and thermal properties. This method overcomes the limitations of existing technologies that typically employ fixed formulations or single-variable adjustments. It innovatively achieves more comprehensive tissue simulation through multi-component concentration control, thus solving the problem of incomplete simulation in existing technologies.

[0087] The present invention selects four representative temperature points within a specific temperature range of radiofrequency ablation to establish material formulations at different temperatures, covering a wide temperature range application scenario for radiofrequency ablation. Through multi-temperature point design, high-precision tissue simulation materials can be prepared within the actual working temperature range of radiofrequency ablation.

[0088] This invention establishes a fitting model relating the electrothermal performance parameters of the simulated material to the concentrations of ethylene glycol, alumina, and sodium chloride constituting the simulated material at different temperatures. Based on the target microstructure and temperature conditions, the component concentrations can be accurately calculated to prepare customized simulated materials.

[0089] This invention makes material preparation more flexible by establishing an adjustable formulation system, and combines wide-temperature-range data modeling with multi-method verification to make up for the shortcomings of poor temperature adaptability or weak interpretability in the prior art.

[0090] This invention also provides a tissue simulation material for RFA scenarios. This material is capable of simulating biological tissues with well-defined electrical conductivity and thermal parameters. Using the method of this invention, materials simulating different types of tissues can be prepared within the frequency and temperature range commonly used in RFA, and these materials can then be used to evaluate the intra-tissue temperature distribution and electrical properties during the RFA process.

[0091] Example 2 This invention illustrates a method for preparing a simulated material capable of simultaneously simulating the electrothermal properties of biological tissue, using polyacrylamide gel matrix as a simulation material. In this embodiment, simulated material samples based on polyacrylamide gel (PGA samples) are specifically prepared, with EG contents of 0%, 40%, and 80% v / v. A total of 27 simulated material samples were collected, with contents of 0%, 3%, and 6% v / v and NaCl contents of 0.2%, 0.4%, and 0.6% w / v. The electrothermal performance parameters (conductivity, thermal conductivity, and volumetric heat capacity) of each sample were measured at temperature points of 45℃, 60℃, 75℃, and 90℃. Based on the composition content of the samples and the corresponding electrothermal performance parameters, a fitting model was established by combining analysis of variance with least squares fitting.

[0092] It should be noted that, in this embodiment, the specific composition of the PGA sample may not be limited to the content distribution in this example. Those skilled in the art can select or adjust the composition according to the method of this embodiment, but this will not affect the method of this embodiment.

[0093] The following is a detailed description of the solution for this example.

[0094] 2.1 Composition of tissue simulation materials In this embodiment, the composition of the tissue simulation material is shown in Table 2.

[0095] Table 2. Composition and Applications of Tissue Simulation Materials

[0096] Table 2 lists the components used to prepare the tissue-simulating material. All chemical reagents were purchased from Macklin. Polyacrylamide gel (PAG) was chosen as the gel matrix due to its many excellent properties, such as good stability, high temperature resistance, and thermal properties similar to soft tissue. In this embodiment, the total concentration of acrylamide AA and N,N'-methylenebisacrylamide MBAA was maintained at 10% (w / v), with the crosslinking concentration of MBAA being 5% of its total mass. This formulation ensures that the material has sufficient rigidity to remain stable and withstand the insertion of a measuring needle without cracking or splitting.

[0097] Ethylene glycol (EG) is a liquid that is completely miscible with water. Its volumetric heat capacity is... Water has a higher volumetric heat capacity, which is Therefore, adding EG to a matrix that is primarily composed of water can effectively reduce the volumetric heat capacity, bringing it closer to the values ​​typically observed in biological tissues (usually within a certain range). and The results are consistent with the above. However, it should be noted that the addition of EG also leads to a decrease in the thermal conductivity and thermal diffusivity of the material. Therefore, it is necessary to introduce other components to quantitatively adjust these thermal properties.

[0098] Alumina ( Alumina is a key additive for regulating the thermal properties of materials. Its thermal conductivity is 30-40. By adding a small amount of alumina to the solution and ensuring that it is evenly dispersed, the thermal conductivity of the material can be effectively enhanced.

[0099] Sodium chloride (NaCl) is added to adjust the conductivity of the material because dissociated ions can move freely in aqueous solution. Conductivity measurements of NaCl solutions with different concentrations show a linear relationship between conductivity and NaCl concentration; the higher the concentration, the greater the conductivity. However, the dissociation of NaCl in solution is affected by the concentration of EG (extracellular ether). As the EG concentration increases, the dissociation of NaCl is gradually suppressed, leading to a decrease in the concentration of free ions. Therefore, conductivity is affected by both EG and NaCl concentrations.

[0100] Ammonium sulfate (APS) and N,N,N′,N′-tetramethylethylenediamine (TEMED) were used as initiators and catalysts for PAG polymerization at concentrations of 0.1% w / v and 0.1% v / v, respectively. The remainder of the tissue-simulating material consisted of deionized water. In this embodiment, a series of materials with different EG values ​​were prepared. Polyacrylamide gel-based samples (PAG samples) with NaCl concentrations were used as simulated material samples. The changes in the electrical and thermal properties of the material were investigated by adjusting the concentrations of these components.

[0101] 2.2 Experimental Design This embodiment creates a full factorial experimental design (DOE) to study the volume fraction of EG, The effects of volume fraction and mass concentration of NaCl on the electrical and thermal properties of the material.

[0102] Table 3 lists the experimental design, with three concentration levels for each factor: EG (0%, 40%, 80% v / v). (0%, 3%, 6% v / v) and NaCl (0.2%, 0.4%, 0.6% w / v). The selected EG levels cover the range of volumetric heat capacities observed in various tissues. For Notably, preliminary experiments observed that when the volume fraction exceeded 6%, the nanoparticles tended to aggregate into larger clusters. The selected NaCl concentration was chosen to ensure that the conductivity covered the typical range for most biological tissues. By combining all possible factor levels, a total of 27 PAG samples were prepared for comprehensive analysis in this embodiment.

[0103] The formula for calculating the volume fraction is as follows: , Where φ represents volume fraction, Indicates added to the sample quality express The density of the PAG sample is 3.97, and V represents the volume of the PAG sample. The temperature during the RFA process was systematically controlled within the range of 45–90 °C. Because the electrical and thermal properties of tissues vary with temperature, measuring the PAG sample at only a single temperature point is insufficient, as it cannot fully cover the diverse conditions encountered in RFA applications. To improve the applicability of the method over a wide temperature range, the electrical and thermal properties of each PAG sample were measured at four different temperature points: 45 °C, 60 °C, 75 °C, and 90 °C. A precisely controlled water bath was used to maintain the temperature of the PAG samples within an accuracy range of ±2 °C. To ensure the reliability of the data, three repeated measurements were performed at each temperature point, and the mean and standard deviation were calculated as the final results.

[0104] Table 3. Concentration of each factor at three different levels

[0105] 2.3 Experimental Design Based on the calculation of the mass or volume of each component in a specific PAG sample according to Table 3 and Section 2.2, PAG samples with corresponding compositions were prepared. The preparation process for these PAG samples was as follows: For samples containing 0% and 40% EG, a precursor solution was prepared by dissolving 40% (w / v) AA / MBAA in deionized water at a mass ratio of 19:1. In contrast, for samples containing 80% EG (over 65%), the precursor solution was prepared by dissolving 16% (w / v) AA / MBAA in EG solvent while maintaining the same mass ratio. The precise volume of the precursor solution was then measured and transferred to a reaction vessel, ensuring that the total mass of AA / MBAA accounted for exactly 10% of the final material volume. After adding the precursor solution, a predetermined volume of EG and additional deionized water were added. The mixture was then placed on a magnetic stirrer, and NaCl was added, followed by the gradual addition of... Gradually join It is crucial to ensure complete dispersion before adding each time to prevent clumping and achieve a uniform distribution. After fully adding... The mixture was then stirred for an additional 30 minutes on a magnetic stirrer. Subsequently, APS was added as a polymerization initiator at a concentration of 0.1% (w / v), followed by TEMED at a concentration of 0.1% (v / v) to accelerate the polymerization process. Magnetic stirring was stopped immediately after the addition of TEMED, and the container was sealed with plastic wrap to prevent dehydration. The mixture was polymerized at room temperature until completion, forming the final PAG sample. For optimal preservation, the PAG sample was stored at 4°C until further application. The prepared PAG sample is shown in Figure 2.

[0106] 2.4 Material property measurement 2.4.1 Electrical performance measurement Conductivity measurements were performed using a test system based on standard four-electrode technology. The measurement system consisted of an AC power supply, an oscilloscope, a connecting circuit board, and a four-electrode probe equipped with platinum electrodes, such as... Figure 3 As shown. The platinum electrode spacing was 2 mm and the length was 5 mm. During the experiment, an alternating current was applied through the two external electrodes, while the voltage difference between the two internal electrodes was recorded using an oscilloscope. All measurements were performed at a frequency of 480 kHz, which is relevant to the RFA application. The system was calibrated using physiological saline at room temperature before measurement. During the measurement, the temperature of the PAG sample was monitored using a temperature probe. When the sample reached the predetermined temperature, the four-electrode probe was carefully applied to the sample. After allowing the readings on the oscilloscope to stabilize, the average of the input voltage and output current was recorded to determine the conductivity of the sample. Measurements were performed at three different locations for each PAG sample. After each measurement, the probe was thoroughly cleaned with deionized water to prevent residue accumulation that could lead to errors in subsequent measurements.

[0107] 2.4.2 Thermal performance measurement Thermal performance measurements were performed using a measurement system based on a dual-probe thermal pulse method. This method is widely used due to its high accuracy and reliability in determining the thermal parameters of biological tissues. The measurement system consists of a power supply, a relay module, an AD sampling module, a main control module, and a dual-needle sensor. The dual-needle sensor consists of two parallel needles spaced 6 mm apart: a heating needle (30 mm long, 2.0 mm in diameter) and a temperature recording needle (15 mm long, 1.5 mm in diameter), as shown in Figure 4. Subsequently, the sensor was completely embedded in the PAG sample, as shown in Figure 4. Figure 5As shown in the figure. During the measurement, the main control module activates the relay switch to power the heating needle for 10 seconds. Afterward, the temperature recording needle monitors temperature fluctuations, and data acquisition continues for 140 seconds after heating stops. Throughout the heating and cooling cycle, a temperature reading is recorded once per second, resulting in 150 data points per measurement. The temperature data is transmitted in real-time to the computing unit via a data cable, which outputs the thermal parameters of the PAG samples. The algorithm for deriving these parameters from temperature-time data is common knowledge and will not be elaborated here. Before measurement, the system was calibrated at room temperature using glycerol and 0.2% gelatin to ensure accuracy and reliability. Each measurement was repeated three times to confirm consistency. The electrical and thermal performance test results of the 27 samples are shown in Table 4.

[0108] Table 4

[0109] σ represents electrical conductivity, k represents thermal conductivity, and ρc represents volumetric heat capacity. The corresponding units are as follows: the unit for electrical conductivity is... The unit of thermal conductivity is The unit of volumetric heat capacity is EG (ethylene glycol) and The unit for (alumina) is volume fraction (v / v), while the unit for NaCl (sodium chloride) is mass concentration (w / v).

[0110] 2.5 Fitting Algorithm To establish a quantitative relationship between measured parameters and experimental factors, a systematic fitting algorithm was implemented. For all 27 PAG samples, experimental data for electrical conductivity, thermal conductivity, and volumetric heat capacity were collected at four different temperature points (45℃, 60℃, 75℃, and 90℃). The analysis began with individual analysis of variance (ANOVA) for each parameter at each temperature condition. The ANOVA model included three principal factors (EG, ...). (and NaCl) and their second-order interactions (EG×NaCl, EG× and NaCl× () was used as the independent variable. Statistical significance was assessed using a conventional threshold p ≤ 0.05 to identify factors significant for each parameter under the four temperature conditions. This methodological framework provides important insights into the individual and combined contributions of each factor to the variations in the electrical and thermal properties of the PAG samples.

[0111] Tables 5 and 6 summarize the results of the analysis of variance on the statistical significance of the evaluation factors for electrical conductivity, thermal conductivity, and volumetric heat capacity at 45℃, 60℃, 75℃, and 90℃.

[0112] Table 5 shows the results of the analysis of variance for the three parameters at 45℃ and 60℃.

[0113] α indicates statistical significance, p ≤ 0.05. The symbols in the table are defined as follows: σ represents electrical conductivity, k represents thermal conductivity, and ρc represents volumetric heat capacity. The corresponding units are: electrical conductivity in S / m, thermal conductivity in W / (m·K), and volumetric heat capacity in J / (m³·K).

[0114] Table 6. Results of ANOVA on the three parameters at 75℃ and 90℃ for the evaluation factors

[0115] α indicates statistical significance, p ≤ 0.05. The symbols in the table are defined as follows: σ represents electrical conductivity, k represents thermal conductivity, and ρc represents volumetric heat capacity. The corresponding units are: electrical conductivity in S / m, thermal conductivity in W / (m·K), and volumetric heat capacity in J / (m³·K).

[0116] Tables 5 and 6 present the results of the analysis of variance, identifying the factors that significantly affected the three parameters under four temperature conditions. Regarding conductivity, the analysis shows that EG, NaCl and its interaction term (EG×NaCl) were statistically significant at all four temperature conditions. Regarding thermal conductivity, although EG× The p-value at 45℃ (0.066) was slightly higher than the conventional significance level of 0.05, but because it was close to the threshold and potentially meaningful, it was retained in the regression model. Therefore, the results indicate that EG, and its interaction items (EG×) ) is a significant factor affecting the thermal conductivity at the four temperature points. Furthermore, analysis of volumetric heat capacity determined EG and its relationship with... Interactive items (EG×) ) is the main factor affecting this performance at all four temperature levels.

[0117] Based on the ANOVA results, least squares regression models were developed for each parameter, including only statistically significant factors and their interactions. This process yielded 12 different models, each corresponding to a specific parameter-temperature combination. The predictive accuracy of these models was evaluated using the coefficient of determination (R²) and adjusted R-squared values ​​to ensure robustness.

[0118] Based on the ANOVA results, the regression equations developed using the least squares regression method, along with their corresponding coefficients of determination (R²) and adjusted R-squared values, are listed in Table 7. These equations model the relationship between electrical conductivity, thermal conductivity, and volumetric heat capacity at four temperature points.

[0119] Table 7. Regression equations for the three parameters at four temperature points

[0120] α indicates statistical significance, p ≤ 0.05. The symbols in the table are defined as follows: σ represents electrical conductivity, k represents thermal conductivity, and ρc represents volumetric heat capacity. The corresponding units are: electrical conductivity in S / m, thermal conductivity in W / (m·K), and volumetric heat capacity in J / (m³·K). ck is a constant with a value of 1000.

[0121] The regression equations listed in Table 7 were developed based on ANOVA results. To develop PAG samples capable of simultaneously simulating the electrical and thermal properties of specific biological tissues, three regression equations need to be solved at the corresponding temperatures to determine the component concentrations. Given that EG and [other components] can be calculated from the regression equations of thermal conductivity and volumetric heat capacity... The concentration of EG, and considering the eighth-order difference due to the unit difference between these two parameters, should be taken into account when simulating these thermal parameters. Therefore, ensuring the accuracy of thermal conductivity should be prioritized. The thermal conductivity simulation range of the proposed method was initially defined. When the volume fraction of EG is 0% and The thermal conductivity of the PAG sample reaches its maximum when the volume fraction of EG is 6%, while it reaches its maximum when the volume fraction of EG is 100%. When the volume fraction is 0%, the thermal conductivity reaches its minimum. When the target thermal conductivity is within this range, an iterative optimization algorithm can be applied to determine the optimal concentration combination. This algorithm will... The volume fraction of [agent] was increased from 0% to 6% in increments of 0.1%, while the volume fraction of EG was increased from 0% to 100% in increments of 1%. For each [agent]... In combination with EG, the algorithm calculates thermal conductivity and identifies candidate solutions with an absolute error of less than 0.001 between the calculated and target values. Among these candidate solutions, volumetric heat capacity is calculated, and the combination with the smallest error is selected as the optimal solution. Then, based on the determined... The NaCl concentration is calculated using EG values ​​to ensure the final solution precisely matches the required performance. This iterative optimization method guarantees the identification of the global optimum through exhaustive search, significantly enhancing the applicability and flexibility of the method.

[0122] These validated models were then used to predict the optimal combination of factors to achieve the target material properties. The predictions were subsequently validated experimentally to confirm the reliability and practical applicability of the developed models.

[0123] 2.6 Verification To verify the reliability and accuracy of the regression model established in Section 2.5, a systematic validation method was employed. The validation process began by obtaining theoretical parameter values ​​for liver tissue at four specific temperatures (45°C, 60°C, 75°C, and 90°C), as liver tissue is the primary target of RFA. These theoretical values ​​were substituted into the corresponding regression model to determine the component concentrations required for preparing PAG samples. Based on these calculations, several PAG samples were prepared, each specifically designed to match one of the four temperature points. The prepared samples were measured at their respective temperatures to obtain experimental parameter values, and the error between the experimental and theoretical values ​​was calculated.

[0124] To verify the reliability of the regression model established in Table 7, liver tissue was selected for validation. The electrical conductivity (480 kHz), thermal conductivity, and volumetric heat capacity of liver tissue at four specific temperatures were obtained from the literature. PAG samples were prepared to simulate these three parameters of liver tissue at four different temperature points, with each sample corresponding to one temperature. The component concentrations of each PAG sample were calculated according to the corresponding regression model, as shown in Table 8, which details the specific component concentrations used to prepare the PAG samples. For clarity, Table 9 lists the theoretical values ​​of liver tissue at the four temperature points and the corresponding experimental comparisons.

[0125] Table 8. Component Concentrations of PAG Samples

[0126] In Table 8, the component concentration is defined as follows: the volume fraction of EG. The volume fraction and the mass concentration of NaCl.

[0127] Table 9 compares the three parameters of the prepared PAG samples with the target values.

[0128] a The symbols in the table are defined as follows: σ represents electrical conductivity, k represents thermal conductivity, and ρc represents volumetric heat capacity. The corresponding units are: electrical conductivity is expressed in S / m, thermal conductivity in W / (m·K), and volumetric heat capacity in J / (m³·K).

[0129] According to Tables 8 and 9, the measured parameter values ​​of the PAG samples prepared using this method showed excellent agreement with the theoretical values ​​of liver tissue at 45℃, 60℃, and 75℃, with all measurement errors remaining within 5%. Among the measured parameters, conductivity showed a slightly larger deviation, which can be attributed to the iterative algorithm used in the calculation prioritizing the optimization of thermal parameters. Furthermore, the conductivity of the PAG samples is temperature-sensitive; even small temperature fluctuations during the measurement process can affect the measured values. At 90℃, the thermal conductivity and volumetric heat capacity of liver tissue exceeded the maximum simulated range achievable by the PAG samples. Therefore, EG and [other parameters] were used to maximize thermal conductivity during the preparation of the PAG samples. The combined results show that the error between the measured and theoretical values ​​at 90°C is greater than that at the other three temperatures. This is consistent with the fact that the thermal conductivity of biological tissues at 90°C typically exceeds 0.7, exceeding the maximum simulation capability of the PAG sample, indicating potential limitations of this method in high-temperature applications. Overall, these results demonstrate that this method can accurately prepare tissue simulation materials that simultaneously simulate the electrical and thermal properties of specific biological tissues.

[0130] This embodiment presents a method for preparing polyacrylamide-based tissue simulation materials using a fitted model for radiofrequency ablation (RFA), capable of simultaneously simulating specified electrical conductivity and any two of three thermal parameters. Considering the temperature-dependent changes in the electrical and thermal properties of biological tissues, and the fact that RFA operates within a specific temperature range, four representative temperature points were selected within this range. Using the proposed method, tissue simulation materials accurately simulating the electrical and thermal properties of biological tissues at the four temperatures were successfully prepared, demonstrating the applicability of the method over a wide temperature range. The developed materials offer significant advantages, including ease of implementation, high reliability, and the ability to closely simulate target tissue properties, making them suitable for research and clinical applications.

[0131] Previous research on RFA tissue simulation materials has primarily faced the challenge of incomplete simulation of electrical and thermal properties. This deficiency often leads to inaccuracies in evaluating key results during the RFA process, such as temperature distribution and ablation volume. In this embodiment, this gap is addressed by utilizing polyacrylamide gel (a material widely used in previous studies) and introducing three components with variable concentrations. By precisely controlling the concentrations of these components, simultaneous, comprehensive, and accurate simulation of the electrical and thermal properties of biological tissues is achieved, thus filling a significant gap in this field. Furthermore, the proposed method is highly adaptable, allowing for the preparation of customized tissue simulation materials by referencing appropriate formulas under specific temperature conditions.

[0132] To assess the long-term stability of the material's properties, the electrical and thermal properties of the PAG samples were measured two weeks after verification. During this period, the PAG samples were stored in sealed containers at 4°C to prevent water evaporation and stabilize their properties. Each PAG sample was heated to its specific target temperature prior to measurement. The results showed that both electrical and thermal properties remained stable with variations within 5%, demonstrating the material's robustness and suitability for long-term use.

[0133] In summary, this embodiment proposes and verifies a novel method for preparing tissue simulation materials. This method can simulate a specified electrical conductivity and any two of the three thermal parameters, thereby enabling an accurate and comprehensive assessment of heat generation, transfer, and distribution using tissue simulation materials.

[0134] It should be noted that this example uses the electrothermal properties of a polyacrylamide gel-based simulated material as an example to illustrate the specific steps of the method in this embodiment. Those skilled in the art can also apply this method to simulated materials based on agar gel or gellan gel, and establish corresponding fitting models based on the relationship between the material content and electrothermal properties of the corresponding simulated material samples. The component composition of the desired target fitting model can then be obtained based on the corresponding fitting model. Further details will not be elaborated here.

[0135] Example 3 Other features of this example are the same as those of Example 1 or 2. This example illustrates the establishment of the fitting model through ANN neural network fitting, specifically including...

[0136] 1. Data Preparation and Preprocessing To establish an effective fitting model, experimental data must first be prepared and preprocessed. The experimental data includes the electrical conductivity, thermal conductivity, volumetric heat capacity, and corresponding temperature values ​​of the simulated material. The specific steps are as follows: (1) Data collection and organization In this experiment, 27 sample phantoms were prepared. The input characteristics of each sample included: electrical conductivity (measured at 480 kHz), thermal conductivity, volumetric heat capacity, and temperature (including four temperature points: 45°C, 60°C, 75°C, and 90°C). The data for each sample includes test results at the four temperatures mentioned above, with the corresponding target output being the concentrations of three components: the volume fraction of ethylene glycol, the volume fraction of alumina, and the mass concentration of sodium chloride. (2) Data normalization To prevent poor model training performance due to differences in the dimensions of different features, a Min-Max normalization method is used to process the data. Specifically, the input features are normalized to map the data to the [0,1] interval, and the output targets (concentrations of ethylene glycol, alumina, and sodium chloride) are also normalized.

[0137] (3) Dataset partitioning The experimental dataset was divided into three parts: Training set: 70% of the data, used to train the neural network.

[0138] Validation set: Comprising 15% of the data, used to adjust hyperparameters and prevent overfitting during training.

[0139] Test set: 15% of the data, used to evaluate the model's generalization ability.

[0140] 2. Neural Network Model Construction and Training Based on the prepared data, construct and train the neural network model. The specific steps are as follows: (1) Model structure design This invention uses a five-layer feedforward neural network for modeling, and the network structure is as follows: Input layer: 4 nodes, corresponding to electrical conductivity, thermal conductivity, volumetric heat capacity and temperature respectively.

[0141] Hidden layer: The first hidden layer uses the Logsig activation function, which can handle non-linear relationships in the data.

[0142] The second hidden layer employs the Poslin activation function, which ensures the non-negativity of the output and helps maintain the interpretability of the physics.

[0143] The third hidden layer employs the Purelin activation function, allowing the network to model linear relationships and ensuring the linear interpretability of the output results.

[0144] Output layer: 3 nodes, corresponding to the concentration prediction of ethylene glycol, alumina and sodium chloride respectively.

[0145] (2) Optimization algorithms and regularization Optimization algorithm: The Levenberg-Marquardt algorithm is selected to optimize network parameters, combined with gradient clipping to prevent gradient explosion and ensure the stability of the optimization process.

[0146] Loss function: Optimizes model performance by minimizing mean squared error (MSE), thereby ensuring that the model can accurately fit the data.

[0147] Regularization measures: Introduce L2 weight decay and Dropout layers to reduce overfitting and improve the model's generalization ability.

[0148] (3) Training process Normalized training data is input into the model for multiple training rounds. After each round, the model is evaluated using validation set data, and network parameters (learning rate, number of hidden layer nodes, etc.) are adjusted. During training, early stopping is used to monitor the loss value on the validation set to prevent overfitting.

[0149] 3. Model Validation and Evaluation After training the neural network model, it needs to be validated and evaluated. The specific steps are as follows: (1) Performance evaluation The model's performance was evaluated using a test set, and the root mean square error and coefficient of determination were calculated.

[0150] The model's evaluation results on the test set are as follows: RMSE: 0.04, R²: 0.97. This indicates that the model has high accuracy and good fit in predicting the concentrations of ethylene glycol, alumina, and sodium chloride based on the target parameters.

[0151] (2) Sensitivity analysis By using SHAP value analysis or weighted analysis, the influence of each input feature on the output concentration can be assessed. This method clearly identifies which features have a greater impact on model predictions.

[0152] (3) Experimental verification Based on the optimal ratio predicted by the ANN model, corresponding phantom samples were prepared. The electrothermal parameters of the samples were measured, and the experimental values ​​were compared with the model predictions to verify the practical application effect of the model.

[0153] Table 10 compares the three parameters of the prepared PAG samples with the target values ​​(ANN model).

[0154] Furthermore, experimental results demonstrate that agarose gel and gellan gel matrices, together with polyacrylamide, constitute three optional gel matrices for the simulated material. The electrothermal properties of the entire simulated material are mainly affected by the concentrations of ethylene glycol, sodium chloride, and alumina. Therefore, the measurement data obtained based on the polyacrylamide gel phantom are basically consistent with the results based on the agarose gel and gellan gel phantoms. The relationship between the material content and electrothermal properties of the agarose gel-based and gellan gel-based simulated materials will not be elaborated upon again.

[0155] Example 4 A simulated material capable of simultaneously mimicking the electrothermal properties of biological tissues was prepared using the method of Example 1 or 2. The simulated material uses polyacrylamide gel, agarose gel, or gellan gel as a base matrix and is composed of ethylene glycol, alumina, and sodium chloride. The ethylene glycol content is greater than 0 and not higher than 80% (v / v), the alumina content is greater than 0 and not higher than 6% (v / v), and the sodium chloride content is 0.1-0.8% (w / v).

[0156] Based on the electrical and thermal properties and temperature conditions of the target tissue, the component concentration of the simulated material is calculated according to the fitted model, and the target simulated material is prepared.

[0157] The simulated material in this embodiment can simulate biological tissues with specific electrothermal properties, and customized simulated materials can be prepared.

[0158] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the scope of protection of the present invention. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the essence and scope of the technical solutions of the present invention.

Claims

1. A method for configuring a simulation material capable of simultaneously simulating the electrical and thermal properties of biological tissue, the method comprising: According to the electrical properties, thermal properties and temperature conditions of the target tissue, the component concentration of the simulation material is calculated according to the fitting model, and the target simulation material is prepared; ​ The fitting model is the relationship between the electrical and thermal performance parameters of the simulation material and the contents of ethylene glycol, alumina and sodium chloride constituting the simulation material at different temperatures; The construction of the fitting model includes preparing a plurality of simulation materials as samples, measuring the electrical and thermal performance of the prepared samples, and constructing the fitting model based on the component concentration of the simulation material and the material properties according to the samples and the electrical and thermal performance parameters of the samples.

2. The method of claim 1, wherein the method is characterized by: The simulation material uses polyacrylamide gel or agar gel or gellan gum as the basic matrix, is composed of ethylene glycol, alumina and sodium chloride, the content of ethylene glycol is not higher than 80% (v / v), the content of alumina is not higher than 6% (v / v), and the content of sodium chloride is 0.1-0.8% (w / v).

3. The method of claim 2, wherein the method further comprises: providing a plurality of materials having different electrical and thermal properties; and selecting a material from the plurality of materials based on the electrical and thermal properties of the material. The simulation material is a simulation material based on polyacrylamide gel, and the preparation process is: S1, preparation of precursor solution, specifically: For samples with ethylene glycol volume fraction below 65%: dissolve 40% (w / v) acrylamide / N,N'-methylene bisacrylamide in deionized water at a mass feed ratio of 19:1; Or For samples with ethylene glycol volume fraction below 65%: dissolve 16% (w / v) acrylamide / N,N'-methylene bisacrylamide in EG solvent at a mass feed ratio of 19:1; S2, preparation of mixed solution, specifically: Measure the precursor solution and transfer it to the reaction container, ensure that the total mass of acrylamide / N,N'-methylene bisacrylamide accounts for 10% of the total volume of the final simulation material; add ethylene glycol and deionized water; S3, stirring and dispersion, specifically: Place the mixture on a magnetic stirrer, add sodium chloride and alumina in sequence; alumina is added in a stepwise manner; after the alumina is completely added, continue magnetic stirring for 10-45 minutes, and then ultrasonic dispersion for 10-15 minutes at 40-60 kHz; S4, initiation of polymerization reaction, specifically: Add 0.1% (w / v) ammonium persulfate as a polymerization initiator, and add 0.1% (v / v) tetramethyl ethylenediamine to accelerate polymerization; S5, stop stirring and seal, specifically: stop magnetic stirring immediately after adding TEMED, and seal the container with plastic wrap to prevent dehydration; S6, polymerization and preservation, specifically: stand at room temperature until the polymerization is completed, and form the final tissue simulation material sample; store the sample in a 4℃ environment.

4. The method of claim 2, wherein the method further comprises: providing a plurality of materials having different electrical and thermal properties; and selecting a material from the plurality of materials based on the electrical and thermal properties of the material. The simulation material is a simulation material based on agar gel, and the preparation process is: S1, solvent premixing and adding, specifically: premix deionized water and ethylene glycol, heat to 85-90℃ in a water bath and maintain the temperature; S2, add agar powder at a concentration of 1.5-2.0% (w / v) after adding, magnetic stirring, maintain the temperature at 85-90℃ until completely transparent; S3, stirring and dispersion, specifically: Cooling to 60-62℃, adding sodium chloride, alumina in turn, alumina is added by stepwise addition; After the alumina is completely added, continue magnetic stirring for 10-30 minutes, and then ultrasonic dispersion for 10-15 minutes at 40-60 kHz to ensure uniform dispersion of particles; S4, gel forming, specifically: room temperature standing for 30 minutes to complete physical crosslinking, then cold storage at 4℃ for 2 hours to enhance mechanical strength, and sealed storage in 0.9% physiological saline moisturizing environment.

5. The method of claim 2, wherein the method further comprises: The simulation material is a simulation material based on a gelatin gum gel, and the preparation process is as follows: ​ S1, solvent premixing system, specifically: premixing deionized water and ethylene glycol, adding 0.1% sodium citrate as a chelating agent; S2, high-temperature dissolution stage, specifically: heating to 85-90℃ in a water bath, adding 0.5% (w / v) low acyl gelatin gum, and stirring until completely transparent; S3, stirring and dispersion, specifically: cooling to 60-62℃, adding sodium chloride, alumina, and 0.8mmol / L CaCl2 solution in turn, and then stirring for 10-30 minutes and ultrasonic dispersion for 10-15 minutes at 40-60 kHz to ensure sufficient dispersion and uniform crosslinking of particles; S4, gel forming, specifically: cooling and solidification at room temperature for 2 hours, and then immersion in 0.1% CaCl2-containing physiological saline for storage, avoiding freezing storage.

6. The configuration method of the simulation material capable of simultaneously simulating the electrical and thermal properties of biological tissues according to any one of claims 1 to 5, characterized in that: The electrical property measurement of the prepared sample includes measuring one or more parameters of conductivity, electrical property constant, and relative electrical property constant; The thermal property measurement of the prepared sample includes measuring one or more parameters of thermal conductivity, thermal diffusivity, and volumetric heat capacity; The electrical and thermal properties of the sample are tested at multiple radiofrequency ablation temperature points.

7. The configuration method of the simulation material capable of simultaneously simulating the electrical and thermal properties of biological tissues according to claim 6, characterized in that: The fitting model is established by variance analysis combined with least squares fitting, specifically including: 1) Construction of fitting algorithm and data collection, specifically: For a series of prepared simulation material samples, experimental data of corresponding electrical and thermal property parameters are collected at multiple temperature points; A least squares regression fitting algorithm is used to establish a quantitative relationship between electrical and thermal property parameters and experimental factors, i.e., ethylene glycol content, alumina content, and sodium chloride content; 2) Variance analysis, specifically: For each electrical and thermal property parameter, variance analysis is independently performed under each temperature condition, and the independent variable of the variance analysis model includes three main factors, i.e., ethylene glycol content, alumina content, and sodium chloride content, and their two-way interactions, i.e., ethylene glycol x sodium chloride, ethylene glycol x alumina, and sodium chloride x alumina; The significance determination scheme is: taking p≤0.05 as the threshold to identify the significant influencing factors of each parameter under each temperature condition, and revealing the independent and synergistic contributions of each factor to the electrical and thermal properties of the tissue simulation material sample; 3) Regression model development and verification, specifically: Model construction: Based on the results of ANOVA, only significant factors and interactions were included to establish a series of least squares regression models corresponding to different parameter-temperature combinations. The established regression model is the fitting model.

8. The method according to claim 6, wherein the method is characterized in that: The fitting model is established by ANN neural network fitting, specifically including: 8.1 Data preparation and preprocessing In order to establish an effective fitting model, first prepare the experimental data and perform preprocessing. The experimental data includes the electrical conductivity, thermal conductivity, volumetric heat capacity of the simulation material and the corresponding temperature values. The specific steps are as follows: (1) Data collection and organization A plurality of sample bodies are prepared in advance. The input features of each sample include: electrical conductivity, thermal conductivity, volumetric heat capacity, and temperature. The data of each sample includes the test results at each temperature, and the target output is the concentration of three components: the volume fraction of ethylene glycol, the volume fraction of aluminum oxide, and the mass concentration of sodium chloride. (2) Data normalization The Min-Max normalization method is used to process the data. Specific operation: Normalize the input features to map the data to the [0, 1] interval, and normalize the output target ethylene glycol, aluminum oxide, and sodium chloride concentration. (3) Data set division The experimental data set is divided into three parts: training set: 70% of the data, used to train the neural network; validation set: 15% of the data, used to adjust hyperparameters and prevent overfitting during training; test set: 15% of the data, used to evaluate the generalization ability of the model. 8.2 Neural network model construction and training According to the data preparation, construct and train the neural network model, the specific steps are as follows: (1) Model structure design A five-layer feedforward neural network is used for modeling, and the network structure is as follows: Input layer: 4 nodes, corresponding to electrical conductivity, thermal conductivity, volumetric heat capacity, and temperature; Hidden layer: First hidden layer: uses Logsig activation function to handle nonlinear relationships in data; Second hidden layer: uses Poslin activation function to ensure non-negativity of output, which helps to maintain physical interpretability; Third hidden layer: uses Purelin activation function to allow the network to model linear relationships and ensure linear interpretability of the output results; Output layer: 3 nodes, corresponding to the concentration prediction of ethylene glycol, aluminum oxide, and sodium chloride; (2) Optimization algorithm and regularization Optimization algorithm: Levenberg-Marquardt algorithm is selected to optimize network parameters, combined with gradient clipping to prevent gradient explosion phenomenon, ensuring the stability of the optimization process; Loss function: minimize mean square error (MSE) to optimize the performance of the model, so as to ensure that the model can accurately fit the data; Regularization measures: introduce L2 weight decay and Dropout layer to reduce overfitting and improve the generalization ability of the model; (3) Training process The normalized training data is input into the model for multiple rounds of training; after each round of training, the model is evaluated using the validation set data to adjust the network parameters; during the training process, the early stopping method is used to monitor the loss value of the validation set to prevent overfitting; 3. Model verification and evaluation After training the neural network model, model verification and evaluation are performed; the specific steps are as follows: (1) Performance evaluation The test set is used to evaluate the performance of the model, and the root mean square error and the determination coefficient are calculated; (2) Sensitivity analysis Through SHAP value analysis or weight analysis, the influence of each input feature on the output concentration is evaluated. In this way, it can be clearly identified which features have a greater impact on model prediction; (3) Experimental verification According to the optimal proportion predicted by the ANN model, the corresponding phantom sample is prepared; the electrical and thermal parameters of the sample are measured, and the experimental values are compared with the model predicted values to verify the actual application effect of the model.

9. The configuration method of the simulation material capable of simultaneously simulating the electrical and thermal characteristics of biological tissues according to claim 7, characterized in that: Preparation of polyacrylamide gel-based simulation material samples, preparation of EG content of 0%, 40%, 80% v / v, A total of 27 simulation material samples with EG content of 0%, 3%, 6% v / v and NaCl content of 0.2%, 0.4%, 0.6% w / v; measure the electro-thermal performance parameters of each sample at temperature points of 45℃, 60℃, 75℃, 90℃, corresponding to conductivity, thermal conductivity and volumetric heat capacity; according to the composition content of the sample and the corresponding electro-thermal performance parameters, a fitting model is established by variance analysis combined with least squares fitting, and the regression equations of the three parameters at the four temperature points are shown in Table 1: Table I. Regression equations of three parameters at four temperature points The symbols in Table 1 are defined as follows: σ represents electrical conductivity, k represents thermal conductivity, ρc represents volumetric heat capacity; the corresponding units are: the unit of electrical conductivity is S / m, the unit of thermal conductivity is W / (m-K), the unit of volumetric heat capacity is J / (m3-K), k is a constant with a value of 1000; in the regression equation in Table 1, EG represents the content of ethylene glycol, ·NaCl represents the content of NaCl, and ·Al2O3 represents the content of Al2O3.

10. A simulation material capable of simultaneously simulating parameters of electrical and thermal properties of biological tissue, characterized by: The polyacrylamide gel or agar gel or gellan gum is used as the basic matrix, and the basic matrix is composed of ethylene glycol, aluminum oxide and sodium chloride, the content of ethylene glycol is greater than 0 and not higher than 80% (v / v), the content of aluminum oxide is greater than 0 and not higher than 6% (v / v), and the content of sodium chloride is 0.1-0.8% (w / v).

Images