Necrotic fascitis process simulation method and system based on animal model

Abstract

The invention discloses a necrotizing fascitis process simulation method and system based on an animal model. The method comprises the steps of animal model selection, infection experiment design, experiment data collection, simulation system construction and necrotizing fascitis process simulation. The invention relates to the technical field of necrotizing fascitis process simulation, in particular to a necrotizing fascitis process simulation method and system based on an animal model, and according to the scheme, a necrotizing fascitis experimental analysis method based on the animal model is adopted to carry out pre-experiment preparation of interference process simulation; carrying out data preprocessing by adopting a method of carrying out multi-modal fusion in combination with artificial features and slice image features; a deep neural network method combined with improved infection process modeling is adopted to construct a simulation system, and certain support and exploration are provided for the research direction of necrotizing fasciitis.

Description

Technical Field

[0001] The present invention relates to the technical field of necrotizing fasciitis process simulation, and in particular to a necrotizing fasciitis process simulation method and system based on an animal model. Background Art

[0002] The necrotizing fasciitis process simulation method based on animal models is a technology that uses experimental animals to simulate and study necrotizing fasciitis, a serious infectious disease. By inducing similar human necrotizing fasciitis pathological processes in animals, researchers can observe the progression of the disease, immune response, pathological changes and treatment effects in detail. The main role of this method is to help scientists understand the pathological mechanism of the disease, evaluate the effectiveness of new drugs and treatment strategies, and provide theoretical support and experimental basis for clinical application.

[0003] However, in the existing necrotizing fasciitis process simulation methods, there is a technical problem that the traditional necrotizing fasciitis research and analysis process relies on human pathogen analysis, which leads to the scarcity of experimental data and increases the difficulty of analysis; in the existing necrotizing fasciitis process simulation methods, there is a technical problem that the data characteristics of the process simulation are complex and diverse, and thus rely on a good data preprocessing process; in the existing necrotizing fasciitis process simulation methods, there is a technical problem that the process simulation of necrotizing fasciitis is highly complex and the prediction indicators involved are difficult to set, which increases the technical difficulty and makes it difficult to effectively combine and experiment with animal models. Summary of the invention

[0004] In view of the above situation, in order to overcome the defects of the prior art, the present invention provides a method and system for simulating the process of necrotizing fasciitis based on an animal model. In view of the technical problem that in the existing method for simulating the process of necrotizing fasciitis, the traditional research and analysis process of necrotizing fasciitis relies on human pathogen analysis, which leads to the scarcity of experimental data and increases the difficulty of analysis, this scheme creatively adopts a necrotizing fasciitis experimental analysis method based on an animal model to conduct preliminary experimental preparation for interference process simulation, and by selecting mice as modeling animals, the quantity and quality of experimental data of necrotizing fasciitis are improved, providing good data support for subsequent research and practice; in view of the technical problem that in the existing method for simulating the process of necrotizing fasciitis, the data characteristics of process simulation are complex and diverse, and thus rely on a good data preprocessing process, this scheme creatively adopts Data preprocessing is performed by combining artificial features and slice image features for multimodal fusion, which provides a better data source for subsequent model prediction. In the existing necrotizing fasciitis process simulation methods, there are technical problems such as the high complexity of necrotizing fasciitis process simulation and the difficulty in setting the prediction indicators involved, which increases the technical difficulty and makes it difficult to effectively combine and experiment with animal models. This scheme creatively uses a deep neural network method combined with improved infection process modeling to construct a simulation system, screens and sets the parameter combination of the prediction target through infection process modeling, and then combines the specific experiments and parameter combinations of the animal model for subsequent model improvement and loss function design, which is helpful to study the pathogenesis and treatment of necrotizing fasciitis through animal experiments, and provides certain support and exploration for the research direction of necrotizing fasciitis.

[0005] The technical solution adopted by the present invention is as follows: The present invention provides a method for simulating the process of necrotizing fasciitis based on an animal model, the method comprising the following steps:

[0006] Step S1: animal model selection;

[0007] Step S2: infection experiment design;

[0008] Step S3: experimental data collection;

[0009] Step S4: simulation system construction;

[0010] Step S5: Necrotizing fasciitis process simulation.

[0011] Furthermore, in step S1, the animal model selection is used to select an appropriate animal model for the study of necrotizing fasciitis process simulation, specifically, a mouse animal model is selected for the process simulation experiment, and strain extraction and cultivation are performed to perform animal grouping and animal model construction to obtain a necrotizing fasciitis process simulation animal model;

[0012] The necrotizing fasciitis process simulation animal model specifically includes experimental animal grouping data, bacterial liquid preparation data and basic model data.

[0013] Further, in step S2, the infection experiment design is used to select the pathogen of necrotizing fasciitis for infection experiment, specifically to construct an infection experiment environment and conduct a necrotizing fasciitis infection process experiment, and obtain infection experiment process data through pathological analysis, resonance examination and post-processing, specifically including the following steps:

[0014] Step S21: process experiment, specifically, conducting a necrotizing fasciitis infection process experiment based on the necrotizing fasciitis process simulation animal model to obtain experimental raw data;

[0015] Step S22: infection tissue pathology analysis, specifically, modeling and infection tissue pathology analysis of mice in the most obvious infection period to obtain a pathology analysis data set;

[0016] Step S23: MRI examination of the infected foci of mice, specifically, anesthetizing the infected mice and performing MRI examination, performing imaging analysis of key parts, and obtaining MRI detection data;

[0017] Step S24: infection experiment process data collection, specifically combining the original experimental data, the scanning analysis experimental data and the nuclear magnetic resonance detection data to obtain the infection experiment process data.

[0018] Further, in step S3, the experimental data collection is used to collect basic data required by the necrotizing fasciitis process simulation system, and optimize and enhance the data, specifically, based on the data in the necrotizing fasciitis process simulation animal model and the infection experiment process data, obtain process simulation raw data through experimental data collection, and perform data preprocessing on the process simulation raw data to obtain process simulation optimization data;

[0019] The steps of experimental data collection and data preprocessing include:

[0020] Step S31: experimental data collection, specifically collecting the original data set required for necrotizing fasciitis process simulation from the infection experimental process data to obtain process simulation original data;

[0021] Step S32: data integrity check, specifically, performing missing, duplicate and outlier detection on the process simulation raw data to obtain an integrity check data set;

[0022] Step S33: outlier processing, specifically, filling the outliers in the integrity check data set with a Z-score filling method to obtain an outlier processing data set;

[0023] Step S34: extracting slice features, specifically using a standard convolutional neural network to extract slice image features from the scanning analysis experimental data in the process simulation raw data to obtain slice feature data;

[0024] Step S35: extracting numerical data features, specifically, using an artificial feature extraction method to extract numerical data features from the abnormal processing data set to obtain numerical feature data;

[0025] Step S36: multimodal feature fusion, specifically, obtaining integrated necrotizing fasciitis process simulation feature data by performing feature selection and feature fusion on the image feature data and the numerical feature data;

[0026] Step S37: data integration, specifically, integrating the necrotizing fasciitis process simulation feature data to obtain process simulation optimization data.

[0027] Further, in step S4, the simulation system is constructed for designing a necrotizing fasciitis process simulation system, specifically, based on the process simulation optimization data, a deep neural network method combined with improved infection process modeling is used to construct the simulation system to obtain a necrotizing fasciitis process simulation model;

[0028] The deep neural network combined with improved infection process modeling specifically includes an infection process module, a neural ordinary differential solver, a deep neural network basic subnet and a process combination optimization loss function;

[0029] The infection process module is used to generate relevant parameters for necrotizing fasciitis process simulation by segmenting the infection process, and to set a target combination of prediction parameters for process simulation;

[0030] The neural constant differential solver is used to enhance the fitting performance of the deep neural network model for necrotizing fasciitis process simulation;

[0031] The deep neural network basic subnet is used as a basic prediction subnet for necrotizing fasciitis process simulation;

[0032] The process combination optimizes the loss function to improve the model performance by minimizing the loss;

[0033] The steps of constructing a simulation system by using a deep neural network method combined with improved infection process modeling to obtain a necrotizing fasciitis process simulation model include:

[0034] Step S41: constructing an infection process module, specifically, modeling the infection process, obtaining an infection process module parameter combination, and using the infection process module parameter combination as the infection process module, and the calculation formula is:

[0035] θ=[β,W,σ,α,γ MR ,γ CR ,γ D ];

[0036] Where θ is the combination of infection process module parameters, which is used to represent the infection process module; β is the infection transmission rate parameter, which is used to represent the pathogen transmission rate of necrotizing fasciitis; W is the weight parameter of the basic subnet of the deep neural network; σ is the incubation rate parameter, which is used to represent the time rate of contact with pathogens and the beginning of infection; α is the tissue destruction rate parameter, which is used to represent the rate at which infection develops to tissue destruction; γ MR is the treatment success rate parameter, which is used to indicate the recovery rate of the infected person after receiving treatment, γ CR is the disease progression rate parameter, which represents the rate at which the infection changes from an early stage to a more severe stage, γ D is a mortality parameter used to indicate the mortality rate caused by failure to promptly treat necrotizing fasciitis;

[0037] Step S42: constructing a neural constant differential solver, specifically, using the neural constant differential solver to calculate the necrotizing fasciitis process simulation vector representation according to the time step, and the calculation formula is:

[0038] Z(t+1)=f(Z(t),θ t ,W);

[0039] where Z(t+1) is the vector representation of the necrotizing fasciitis process at time t+1, t is the time step index, f(·) is the neural network identification function, specifically the deep neural network, Z(t) is the vector representation of the necrotizing fasciitis process at time t, θ t is the combination of infection process module parameters at time t, W is the weight parameter of the basic subnet of the deep neural network;

[0040] Step S43: constructing a basic subnet of a deep neural network, specifically constructing a basic subnet of a deep neural network including an input layer, a hidden layer and an output layer, and improving the hidden layer by combining the infection process module and the neural constant differential solver to construct the basic subnet of the deep neural network;

[0041] Step S44: construct a process combination optimization loss function, which is used to optimize the overall loss of the model in combination with the infected process module. The calculation formula is:

[0042]

[0043] Where L is the process combination optimization loss function, T is the total number of time steps, t is the time step index, min is the minimum value function, θ is the infection process module parameter combination, X tis real data, used to represent the real data obtained from the experiment before the simulation of the necrotizing fasciitis process. is the predicted data, which is used to represent the predicted data obtained by the necrotizing fasciitis process simulation model. t is the total number of actual treatments, is the predicted total number of treatments, CR t is the total number of true deteriorations, is the total number of predicted deteriorations, D t is the true total number of deaths. is the total number of deaths predicted;

[0044] Step S45: training the necrotizing fasciitis process simulation model, specifically, by constructing the infection process module, the neural constant differential solver, the deep neural network basic subnet and the process combination optimization loss function, performing model training to obtain the necrotizing fasciitis process simulation model Model NF .

[0045] Further, in step S5, the necrotizing fasciitis process simulation is used to combine the animal model experimental data and the deep learning model to perform necrotizing fasciitis process simulation, specifically using the necrotizing fasciitis process simulation model to perform necrotizing fasciitis process simulation based on the animal model to obtain necrotizing fasciitis process simulation prediction data;

[0046] The necrotizing fasciitis process simulation prediction data specifically refers to the predicted infection process model parameter combination change data with time steps.

[0047] The necrotizing fasciitis process simulation system based on an animal model provided by the present invention comprises an infection experiment module, a data collection module, a simulation construction module and a process simulation module;

[0048] The infection experiment module is used to select an animal model and design an infection experiment, obtain a necrotizing fasciitis process simulation animal model and infection experiment process data through animal model selection and infection experiment design, and send the necrotizing fasciitis process simulation animal model and infection experiment process data to the data collection module;

[0049] The data collection module is used to collect experimental data, obtain process simulation optimization data through experimental data collection, and send the process simulation optimization data to the simulation construction module;

[0050] The simulation construction module is used to construct a process simulation system for necrotizing fasciitis, obtain a process simulation model for necrotizing fasciitis through simulation system construction, and send the process simulation model for necrotizing fasciitis to the process simulation module;

[0051] The process simulation module is used to simulate the necrotizing fasciitis process, and obtain necrotizing fasciitis process simulation prediction data through necrotizing fasciitis process simulation.

[0052] The beneficial effects achieved by the present invention using the above scheme are as follows:

[0053] (1) In view of the technical problem that the traditional necrotizing fasciitis research and analysis process relies on human pathogen analysis in the existing necrotizing fasciitis process simulation methods, which leads to the scarcity of experimental data and increases the difficulty of analysis, this scheme creatively adopts the necrotizing fasciitis experimental analysis method based on animal models to conduct pre-experimental preparation for interference process simulation. By selecting mice as modeling animals, the quantity and quality of necrotizing fasciitis experimental data are improved, providing good data support for subsequent research and practice;

[0054] (2) In view of the technical problem that the existing necrotizing fasciitis process simulation methods have complex and diverse data features and thus rely on a good data preprocessing process, this scheme creatively adopts a multimodal fusion method combining artificial features and slice image features for data preprocessing, providing a better data source for subsequent model prediction;

[0055] (3) In view of the technical problems in the existing necrotizing fasciitis process simulation methods, the process simulation of necrotizing fasciitis is relatively complex and the prediction indicators involved are difficult to set, which increases the technical difficulty and makes it difficult to effectively combine and experiment with animal models. This scheme creatively adopts a deep neural network method combined with improved infection process modeling to construct a simulation system. Through infection process modeling, the parameter combination of the prediction target is screened and set, and then the specific experiments and parameter combinations of the animal model are combined for subsequent model improvement and loss function design, which is helpful to study the pathogenesis and treatment of necrotizing fasciitis through animal experiments, and provides certain support and exploration for the research direction of necrotizing fasciitis. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 A schematic diagram of a process of simulating necrotizing fasciitis based on an animal model provided by the present invention;

[0057] Figure 2 A schematic diagram of a necrotizing fasciitis process simulation system based on an animal model provided by the present invention;

[0058] Figure 3 Schematic diagram of the process designed for step S2 infection experiment;

[0059] Figure 4 A schematic diagram of the process of collecting experimental data in step S3;

[0060] Figure 5 A schematic diagram of the process of simulating the system construction for step S4;

[0061] Figure 6 This is a schematic diagram of the body temperature measurement result in step S2;

[0062] Figure 7 Schematic diagram of the infection range measurement results in step S2.

[0063] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. DETAILED DESCRIPTION

[0064] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0065] In the description of the present invention, it should be understood that terms such as “upper”, “lower”, “front”, “back”, “left”, “right”, “top”, “bottom”, “inside” and “outside” indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore should not be understood as limiting the present invention.

[0066] Example 1, see Figure 1 The present invention provides a method for simulating the process of necrotizing fasciitis based on an animal model, the method comprising the following steps:

[0067] Step S1: animal model selection;

[0068] Step S2: infection experiment design;

[0069] Step S3: experimental data collection;

[0070] Step S4: simulation system construction;

[0071] Step S5: Necrotizing fasciitis process simulation.

[0072] Example 2, see Figure 1 and Figure 2In step S1, the animal model is selected to select an appropriate animal model for the study of necrotizing fasciitis process simulation, specifically, a mouse animal model is selected for the process simulation experiment, and strain extraction and cultivation are performed to group animals and construct animal models to obtain a necrotizing fasciitis process simulation animal model;

[0073] The necrotizing fasciitis process simulation animal model specifically includes experimental animal grouping data, bacterial liquid preparation data and basic model data;

[0074] The experimental animal grouping data specifically includes the first experimental group data, the second experimental group data, the third experimental group data and the control group data;

[0075] The first experimental group data specifically refers to the streptococcus injection experimental data;

[0076] The second experimental group data specifically refers to the experimental data of injection of Staphylococcus aureus;

[0077] The third experimental group data specifically refers to the experimental data of injection of a mixture of streptococci and Staphylococcus aureus;

[0078] The control group data specifically refers to the data of the experiment with injection of normal saline;

[0079] The bacterial solution preparation data specifically refers to the experimental bacterial solution data obtained by preparing the bacterial solution using 4 sterile test tubes injected with an appropriate amount of physiological saline; the bacterial solution preparation specifically includes streptococcal lysate, Staphylococcus aureus strain lysate and a mixed bacterial solution prepared by mixing equal amounts; the experimental bacterial solution data specifically refers to the strain concentration data obtained by measuring the turbidity meter;

[0080] The basic model data specifically refers to four groups of ten C57BL / 6J mice in each group, of which females and males are equal. Among the four groups of mice, the first three groups of mice are experimental groups, and the fourth group of mice is a control group. The basic animal model experimental data are recorded by measuring body temperature and infection range, extracting and detecting mouse peripheral blood cells, detecting mouse C-reactive protein and interleukin-6, and extracting mouse fascia tissue RNA and performing RNA reverse transcription.

[0081] The body temperature and infection range measurement specifically refers to measuring the body temperature of mice in the experimental group and the control group on days 1, 3, 5, 7, 9 and 11 after the experimental infection, and observing and measuring the infection range of mice in the experimental group and the control group on days 0, 1, 3, 5, 7, 9 and 11 after the experiment. The infection range is specifically measured by measuring the longest distance, the shortest distance and the average distance of the infection focus of each mouse with a soft ruler, and then using the circle area formula to calculate the average infection range of each group of mice;

[0082] The extraction and detection of mouse peripheral blood cells specifically refers to collecting blood from the mouse orbits using a 1 ml heparinized centrifuge tube;

[0083] The mouse C-reactive protein and interleukin-6 detection specifically refers to the absorbance value data of each well measured at a wavelength of 450nm using an ELISA plate strip detected by extracting mouse serum and detecting C-reactive protein and interleukin-6;

[0084] The mouse fascia tissue RNA extraction and RNA reverse transcription specifically refer to the concentration data of the sample target gene and the internal reference gene obtained by mouse tissue extraction, RNA extraction and RNA reverse transcription.

[0085] Example 3, see Figure 1 , Figure 2 , Figure 3 , Figure 6 and Figure 7 This embodiment is based on the above embodiment. In step S2, the infection experiment design is used to select the pathogen of necrotizing fasciitis for infection experiment, specifically to construct an infection experiment environment and conduct a necrotizing fasciitis infection process experiment, and obtain infection experiment process data through pathological analysis, resonance inspection and post-processing, specifically including the following steps:

[0086] Step S21: process experiment, specifically, conducting a necrotizing fasciitis infection process experiment based on the necrotizing fasciitis process simulation animal model to obtain experimental raw data;

[0087] Step S22: infection tissue pathology analysis, specifically, modeling and infection tissue pathology analysis of mice in the most obvious infection period to obtain a pathology analysis data set;

[0088] The steps of infection tissue pathological analysis include:

[0089] Step S221: treating mice, specifically treating the mice with successful modeling by cervical dislocation;

[0090] Step S222: tissue collection and fixation, specifically, fixing the treated mice on a clean bench, cutting out infected tissue samples and placing them in a 4% paraformaldehyde solution, and then placing the samples in a 4°C refrigerator for fixation for 36 hours to obtain fixed tissue samples;

[0091] Step S223: tissue preparation, specifically, taking out the fixed tissue sample and cutting the tissue of appropriate size, placing the tissue in 70%, 80%, 90%, 95% ethanol and anhydrous ethanol for 1 hour each for gradient dehydration, and placing the dehydrated tissue in xylene for transparent treatment to obtain tissue preparation;

[0092] Step S224: paraffin embedding, specifically placing the tissue preparation in paraffin, embedding at 60° C. for 1 hour, to obtain a paraffin-embedded tissue sample;

[0093] Step S225: Slice making, specifically using a microtome to cut the paraffin-embedded tissue sample into 4 mm thick slices, flattening them on a 45°C water surface, and then placing the flattened slices on a glass slide coated with an anti-shedding agent, and baking them in an oven at 60°C for 2 hours to obtain original tissue slice samples;

[0094] Step S226: Slice staining, specifically, dewaxing the original tissue slice sample with xylene and hydrating it with gradient anhydrous ethanol, then staining it with HE, and then staining it with hematoxylin solution, then differentiating it with hydrochloric acid alcohol and anti-blueing it with ammonia water, then staining it with eosin and rinsing it with running water to obtain a stained slice sample;

[0095] Step S227: sealing analysis, specifically, after the stained slice sample is dehydrated with gradient alcohol and transparentized with xylene, a drop of neutral gum is added on the slide, and a cover glass is covered, and the slice result is analyzed by electron microscope scanning to obtain scanning analysis experimental data;

[0096] Step S23: MRI examination of the infected foci of mice, specifically, anesthetizing the infected mice and performing MRI examination, performing imaging analysis of key parts, and obtaining MRI detection data;

[0097] The imaging analysis of the key parts specifically includes sagittal, coronal and transverse imaging analysis;

[0098] Step S24: infection experiment process data collection, specifically combining the original experimental data, the scanning analysis experimental data and the nuclear magnetic resonance detection data to obtain the infection experiment process data.

[0099] See also Figure 6 , where the circular legend is the temperature measurement results of the first experimental group data, the square legend is the temperature measurement results of the second experimental group data, the equilateral triangle legend is the temperature measurement results of the third experimental group data, and the inverted triangle legend is the temperature measurement results of the control group data. The horizontal direction is the experimental days, and the vertical direction is the temperature value;

[0100] See also Figure 7 , where legend A is the infection range result of the first experimental group data, legend B is the infection range result of the second experimental group data, legend C is the infection range result of the third experimental group data, and legend D is the infection range result of the control group data. The horizontal direction is the experimental days, and the vertical direction is the infection range area value;

[0101] By performing the above operations, in view of the technical problem that in the existing necrotizing fasciitis process simulation methods, the traditional necrotizing fasciitis research and analysis process relies on human pathogen analysis, which leads to the scarcity of experimental data and increases the difficulty of analysis, this scheme creatively adopts the necrotizing fasciitis experimental analysis method based on animal models to conduct preliminary experimental preparations for interference process simulation. By selecting mice as modeling animals, the quantity and quality of experimental data of necrotizing fasciitis are improved, providing good data support for subsequent research and practice.

[0102] Example 4, see Figure 1 , Figure 2 and Figure 4 This embodiment is based on the above embodiment. In step S3, the experimental data collection is used to collect basic data required by the necrotizing fasciitis process simulation system and optimize and enhance the data. Specifically, based on the data in the necrotizing fasciitis process simulation animal model and the infection experiment process data, the process simulation raw data is obtained through experimental data collection, and the process simulation raw data is preprocessed to obtain process simulation optimization data;

[0103] The steps of experimental data collection and data preprocessing include:

[0104] Step S31: experimental data collection, specifically collecting the original data set required for necrotizing fasciitis process simulation from the infection experimental process data to obtain process simulation original data;

[0105] Step S32: data integrity check, specifically, performing missing, duplicate and outlier detection on the process simulation raw data to obtain an integrity check data set;

[0106] Step S33: outlier processing, specifically, filling the outliers in the integrity check data set with a Z-score filling method to obtain an outlier processing data set;

[0107] Step S34: extracting slice features, specifically using a standard convolutional neural network to extract slice image features from the scanning analysis experimental data in the process simulation raw data to obtain slice feature data;

[0108] Step S35: extracting numerical data features, specifically, using an artificial feature extraction method to extract numerical data features from the abnormal processing data set to obtain numerical feature data;

[0109] Step S36: multimodal feature fusion, specifically, obtaining integrated necrotizing fasciitis process simulation feature data by performing feature selection and feature fusion on the image feature data and the numerical feature data;

[0110] Step S37: data integration, specifically, integrating the necrotizing fasciitis process simulation feature data to obtain process simulation optimization data.

[0111] By executing the above operations, in order to address the technical problem that in the existing necrotizing fasciitis process simulation methods, the data features of the process simulation are complex and diverse, and therefore rely on a good data preprocessing process, this solution creatively adopts a multimodal fusion method that combines artificial features and slice image features for data preprocessing, providing a better data source for subsequent model predictions.

[0112] Example 5, see Figure 1 , Figure 2 and Figure 5 This embodiment is based on the above embodiment. In step S4, the simulation system is constructed to design a necrotizing fasciitis process simulation system. Specifically, based on the process simulation optimization data, a deep neural network method combined with improved infection process modeling is used to construct the simulation system to obtain a necrotizing fasciitis process simulation model.

[0113] The deep neural network combined with improved infection process modeling specifically includes an infection process module, a neural ordinary differential solver, a deep neural network basic subnet and a process combination optimization loss function;

[0114] The infection process module is used to generate relevant parameters for necrotizing fasciitis process simulation by segmenting the infection process, and to set a target combination of prediction parameters for process simulation;

[0115] The neural constant differential solver is used to enhance the fitting performance of the deep neural network model for necrotizing fasciitis process simulation;

[0116] The deep neural network basic subnet is used as a basic prediction subnet for necrotizing fasciitis process simulation;

[0117] The process combination optimizes the loss function to improve the model performance by minimizing the loss;

[0118] The steps of constructing a simulation system by using a deep neural network method combined with improved infection process modeling to obtain a necrotizing fasciitis process simulation model include:

[0119] Step S41: constructing an infection process module, specifically, modeling the infection process, obtaining an infection process module parameter combination, and using the infection process module parameter combination as the infection process module, and the calculation formula is:

[0120] θ=[β,W,σ,α,γ MR ,γ CR ,γD ];

[0121] Where θ is the combination of infection process module parameters, which is used to represent the infection process module; β is the infection transmission rate parameter, which is used to represent the pathogen transmission rate of necrotizing fasciitis; W is the weight parameter of the basic subnet of the deep neural network; σ is the incubation rate parameter, which is used to represent the time rate of contact with pathogens and the beginning of infection; α is the tissue destruction rate parameter, which is used to represent the rate at which infection develops to tissue destruction; γ MR is the treatment success rate parameter, which is used to indicate the recovery rate of the infected person after receiving treatment, γ CR is the disease progression rate parameter, which represents the rate at which the infection changes from an early stage to a more severe stage, γ D is a mortality parameter used to indicate the mortality rate caused by failure to promptly treat necrotizing fasciitis;

[0122] Step S42: constructing a neural constant differential solver, specifically, using the neural constant differential solver to calculate the necrotizing fasciitis process simulation vector representation according to the time step, and the calculation formula is:

[0123] Z(t+1)=f(Z(t),θ t ,W);

[0124] where Z(t+1) is the vector representation of the necrotizing fasciitis process at time t+1, t is the time step index, f(·) is the neural network identification function, specifically the deep neural network, Z(t) is the vector representation of the necrotizing fasciitis process at time t, θ t is the combination of infection process module parameters at time t, W is the weight parameter of the basic subnet of the deep neural network;

[0125] Step S43: constructing a basic subnet of a deep neural network, specifically constructing a basic subnet of a deep neural network including an input layer, a hidden layer and an output layer, and improving the hidden layer by combining the infection process module and the neural constant differential solver to construct the basic subnet of the deep neural network;

[0126] The calculation formula for the hidden layer improvement is:

[0127]

[0128] In the formula, is the first hidden layer in the tth i The output at time steps, i is the time step unit index, which is specifically used to represent the number of days of infection of necrotizing fasciitis, ELU(·) is the hidden layer activation function, and input u is the total set of input layer data, u is the total number of input layers, j is the input layer index, is the input data corresponding to the jth input layer and the weight of the first hidden layer, input j is the total data of the jth input layer, is the bias term of the first hidden layer, k1 is the unit number index of the first hidden layer;

[0129] is the second hidden layer in the tth i The output at time steps, ELU(·) is the hidden layer activation function, h1 is the number of hidden units in the first hidden layer, k1 is the unit number index of the first hidden layer, are the weights corresponding to the first and second hidden layers, is the bias term of the second hidden layer, k2 is the unit number index of the second hidden layer;

[0130] is the Nth hidden layer in the tth i The output at the time step, N is the total number of hidden layers, the specific value is 14, h N-1 is the number of hidden units in the N-1th hidden layer, k N-1 is the index of the number of units in the N-1th hidden layer, is the N-1th hidden layer in the tth i The output at time steps is is the bias term of the Nth hidden layer, k N is the index of the number of units in the Nth hidden layer;

[0131] Step S44: construct a process combination optimization loss function, which is used to optimize the overall loss of the model in combination with the infected process module. The calculation formula is:

[0132]

[0133] Where L is the process combination optimization loss function, T is the total number of time steps, t is the time step index, min is the minimum value function, θ is the infection process module parameter combination, X t is real data, used to represent the real data obtained from the experiment before the simulation of the necrotizing fasciitis process. is the predicted data, which is used to represent the predicted data obtained by the necrotizing fasciitis process simulation model. t is the total number of actual treatments, is the predicted total number of treatments, CR t is the total number of true deteriorations, is the total number of predicted deteriorations, D t is the true total number of deaths. is the total number of deaths predicted;

[0134] Step S45: training the necrotizing fasciitis process simulation model, specifically, by constructing the infection process module, the neural constant differential solver, the deep neural network basic subnet and the process combination optimization loss function, performing model training to obtain the necrotizing fasciitis process simulation model Model NF .

[0135] By performing the above operations, in view of the technical problem that in the existing necrotizing fasciitis process simulation methods, the process simulation of necrotizing fasciitis is highly complex and the prediction indicators involved are difficult to set, which increases the technical difficulty and makes it difficult to effectively combine and experiment with animal models, this scheme creatively adopts a deep neural network method combined with improved infection process modeling to construct a simulation system, screens and sets the parameter combination of the prediction target through infection process modeling, and then combines the specific experiments and parameter combinations of the animal model for subsequent model improvement and loss function design, which is helpful to study the pathogenesis and treatment of necrotizing fasciitis through animal experiments, and provides certain support and exploration for the research direction of necrotizing fasciitis.

[0136] Example 6, see Figure 1 and Figure 2 This embodiment is based on the above embodiment. In step S5, the necrotizing fasciitis process simulation is used to combine the animal model experimental data and the deep learning model to simulate the necrotizing fasciitis process. Specifically, the necrotizing fasciitis process simulation model is used to perform a necrotizing fasciitis process simulation based on the animal model to obtain necrotizing fasciitis process simulation prediction data;

[0137] The necrotizing fasciitis process simulation prediction data specifically refers to the predicted infection process model parameter combination change data with time steps.

[0138] Embodiment 7, see Figure 1 and Figure 2 , this embodiment is based on the above embodiment, and the necrotizing fasciitis process simulation system based on the animal model provided by the present invention includes an infection experiment module, a data collection module, a simulation construction module and a process simulation module;

[0139] The infection experiment module is used to select an animal model and design an infection experiment, obtain a necrotizing fasciitis process simulation animal model and infection experiment process data through animal model selection and infection experiment design, and send the necrotizing fasciitis process simulation animal model and infection experiment process data to the data collection module;

[0140] The data collection module is used to collect experimental data, obtain process simulation optimization data through experimental data collection, and send the process simulation optimization data to the simulation construction module;

[0141] The simulation construction module is used to construct a process simulation system for necrotizing fasciitis, obtain a process simulation model for necrotizing fasciitis through simulation system construction, and send the process simulation model for necrotizing fasciitis to the process simulation module;

[0142] The process simulation module is used to simulate the necrotizing fasciitis process, and obtain necrotizing fasciitis process simulation prediction data through necrotizing fasciitis process simulation.

[0143] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0144] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the invention.

[0145] The present invention and its embodiments are described above, which is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it and design structural methods and embodiments similar to the technical solution without creativity without departing from the purpose of the invention, they should all fall within the protection scope of the present invention.

Claims

1. Method for simulating the process of necrotizing fasciitis based on animal models, Features: The method comprises the following steps: Step S1: selecting an animal model, specifically selecting a mouse animal model for a process simulation experiment, and extracting and culturing strains, performing animal grouping and animal model construction to obtain a necrotizing fasciitis process simulation animal model; Step S2: infection experiment design, specifically, constructing an infection experiment environment and conducting a necrotizing fasciitis infection process experiment, and obtaining infection experiment process data through pathological analysis, resonance examination and post-processing; Step S3: experimental data collection, specifically, obtaining process simulation raw data through experimental data collection based on the data in the necrotizing fasciitis process simulation animal model and the infection experiment process data, and performing data preprocessing on the process simulation raw data to obtain process simulation optimization data; Step S4: constructing a simulation system for designing a necrotizing fasciitis process simulation system, specifically, based on the process simulation optimization data, using a deep neural network method combined with improved infection process modeling to construct a simulation system to obtain a necrotizing fasciitis process simulation model, specifically including the following steps: Step S41: constructing an infection process module; Step S42: constructing a neural constant differential solver; Step S43: constructing a deep neural network basic subnet; Step S44: constructing a process combination optimization loss function; Step S45: training a necrotizing fasciitis process simulation model; Step S5: simulating the process of necrotizing fasciitis, specifically using the necrotizing fasciitis process simulation model to perform a necrotizing fasciitis process simulation based on an animal model to obtain necrotizing fasciitis process simulation prediction data; The necrotizing fasciitis process simulation prediction data specifically refers to the predicted infection process model parameter combination change data with time steps.

2. The method for simulating the process of necrotizing fasciitis based on an animal model according to claim 1, Features: In step S2, the infection experiment design is used to select the pathogen of necrotizing fasciitis for infection experiment, specifically to construct an infection experiment environment and conduct a necrotizing fasciitis infection process experiment, and obtain infection experiment process data through pathological analysis, resonance examination and post-processing, specifically including the following steps: Step S21: process experiment, specifically, conducting a necrotizing fasciitis infection process experiment based on the necrotizing fasciitis process simulation animal model to obtain experimental raw data; Step S22: infection tissue pathology analysis, specifically, modeling and infection tissue pathology analysis of mice in the most obvious infection period to obtain a pathology analysis data set; Step S23: MRI examination of the infected foci of mice, specifically, anesthetizing the infected mice and performing MRI examination, performing imaging analysis of key parts, and obtaining MRI detection data; Step S24: infection experiment process data collection, specifically combining the original experimental data, the scanning analysis experimental data and the nuclear magnetic resonance detection data to obtain the infection experiment process data.

3. The method for simulating the process of necrotizing fasciitis based on an animal model according to claim 2, Features: In step S3, the steps of experimental data collection and data preprocessing include: Step S31: experimental data collection, specifically collecting the original data set required for necrotizing fasciitis process simulation from the infection experimental process data to obtain process simulation original data; Step S32: data integrity check, specifically, performing missing, duplicate and outlier detection on the process simulation raw data to obtain an integrity check data set; Step S33: outlier processing, specifically, filling the outliers in the integrity check data set with a Z-score filling method to obtain an outlier processing data set; Step S34: extracting slice features, specifically using a standard convolutional neural network to extract slice image features from the scanning analysis experimental data in the process simulation raw data to obtain slice feature data; Step S35: extracting numerical data features, specifically, using an artificial feature extraction method to extract numerical data features from the abnormal processing data set to obtain numerical feature data; Step S36: multimodal feature fusion, specifically, obtaining integrated necrotizing fasciitis process simulation feature data by performing feature selection and feature fusion on the image feature data and the numerical feature data; Step S37: data integration, specifically, integrating the necrotizing fasciitis process simulation feature data to obtain process simulation optimization data.

4. The method for simulating the process of necrotizing fasciitis based on an animal model according to claim 3, Features: The deep neural network combined with improved infection process modeling specifically includes an infection process module, a neural ordinary differential solver, a deep neural network basic subnet and a process combination optimization loss function; The infection process module is used to generate relevant parameters for necrotizing fasciitis process simulation by segmenting the infection process, and to set a target combination of prediction parameters for process simulation; The neural constant differential solver is used to enhance the fitting performance of the deep neural network model for necrotizing fasciitis process simulation; The deep neural network basic subnet is used as a basic prediction subnet for necrotizing fasciitis process simulation; The process combination optimizes the loss function to improve the model performance by minimizing the loss.

5. The method for simulating the process of necrotizing fasciitis based on an animal model according to claim 4, Features: In step S4, the step of constructing a simulation system by using a deep neural network method combined with improved infection process modeling to obtain a necrotizing fasciitis process simulation model includes: Step S41: constructing an infection process module, specifically, modeling the infection process, obtaining an infection process module parameter combination, and using the infection process module parameter combination as the infection process module, and the calculation formula is: θ=[β,W,σ,α,γ MR ,c CR ,c D ]; Where θ is the combination of infection process module parameters, which is used to represent the infection process module; β is the infection transmission rate parameter, which is used to represent the pathogen transmission rate of necrotizing fasciitis; W is the weight parameter of the basic subnet of the deep neural network; σ is the incubation rate parameter, which is used to represent the time rate of contact with pathogens and the beginning of infection; α is the tissue destruction rate parameter, which is used to represent the rate at which infection develops to tissue destruction; γ MR is the treatment success rate parameter, which is used to indicate the recovery rate of the infected person after receiving treatment, γ CR is the disease progression rate parameter, which represents the rate at which the infection changes from an early stage to a more severe stage, γ D is a mortality parameter used to indicate the mortality rate caused by failure to promptly treat necrotizing fasciitis; Step S42: constructing a neural constant differential solver, specifically, using the neural constant differential solver to calculate the necrotizing fasciitis process simulation vector representation according to the time step, and the calculation formula is: Z(t+1)=f(Z(t),θ t ,W); where Z(t+1) is the vector representation of the necrotizing fasciitis process at time t+1, t is the time step index, f(·) is the neural network identification function, specifically the deep neural network, Z(t) is the vector representation of the necrotizing fasciitis process at time t, θ t is the combination of infection process module parameters at time t, W is the weight parameter of the basic subnet of the deep neural network; Step S43: constructing a basic subnet of a deep neural network, specifically constructing a basic subnet of a deep neural network including an input layer, a hidden layer and an output layer, and improving the hidden layer by combining the infection process module and the neural constant differential solver to construct the basic subnet of the deep neural network; Step S44: construct a process combination optimization loss function, which is used to optimize the overall loss of the model in combination with the infected process module. The calculation formula is: Where L is the process combination optimization loss function, T is the total number of time steps, t is the time step index, min is the minimum value function, θ is the infection process module parameter combination, X t is real data, used to represent the real data obtained from the experiment before the simulation of the necrotizing fasciitis process. is the predicted data, which is used to represent the predicted data obtained by the necrotizing fasciitis process simulation model. t is the total number of actual treatments, is the predicted total number of treatments, CR t is the total number of true deteriorations, is the total number of predicted deteriorations, D t is the true total number of deaths. is the total number of deaths predicted; Step S45: training the necrotizing fasciitis process simulation model, specifically, by constructing the infection process module, the neural constant differential solver, the deep neural network basic subnet and the process combination optimization loss function, performing model training to obtain the necrotizing fasciitis process simulation model Model NF .

6. The method for simulating the process of necrotizing fasciitis based on an animal model according to claim 5, Features: In step S5, the necrotizing fasciitis process simulation is used to combine the animal model experimental data and the deep learning model to simulate the necrotizing fasciitis process; In step S1, the animal model selection is used to select an appropriate animal model for the study of necrotizing fasciitis process simulation; The necrotizing fasciitis process simulation animal model specifically includes experimental animal grouping data, bacterial liquid preparation data and basic model data.

7. The method for simulating the process of necrotizing fasciitis based on an animal model according to claim 6, Features: In step S22, the steps of infection tissue pathological analysis include: step S221: mouse processing; step S222: tissue collection and fixation; step S223: tissue preparation; step S224: paraffin embedding; step S225: section preparation; step S226: section staining; step S227: sealing analysis.

8. A necrotizing fasciitis process simulation system based on an animal model, used to implement the necrotizing fasciitis process simulation method based on an animal model as described in any one of claims 1 to 7, Features: It includes infection experiment module, data collection module, simulation construction module and process simulation module.

9. The necrotizing fasciitis process simulation system based on an animal model according to claim 8, Features: The infection experiment module is used to select an animal model and design an infection experiment, obtain a necrotizing fasciitis process simulation animal model and infection experiment process data through animal model selection and infection experiment design, and send the necrotizing fasciitis process simulation animal model and infection experiment process data to the data collection module; The data collection module is used to collect experimental data, obtain process simulation optimization data through experimental data collection, and send the process simulation optimization data to the simulation construction module; The simulation construction module is used to construct a process simulation system for necrotizing fasciitis, obtain a process simulation model for necrotizing fasciitis through simulation system construction, and send the process simulation model for necrotizing fasciitis to the process simulation module; The process simulation module is used to simulate the necrotizing fasciitis process, and obtain necrotizing fasciitis process simulation prediction data through necrotizing fasciitis process simulation.

Images