Artificial intelligence dynamic evaluation mouse multiple injury model construction method and system

Abstract

The invention discloses an artificial intelligence dynamic evaluation mouse multiple injury model construction method and system, and belongs to the technical field of injury models. The method comprises the steps that controlled impact is applied to the brain of a mouse to cause brain injury, and controlled trauma is introduced into the thoracic cavity and the abdominal cavity to simulate thoracic cavity and abdominal cavity injury; physiological parameters of the injured mouse are monitored in real time through a sensing device, and physiological data reflecting craniocerebral, thoracic and abdominal injury influences are collected; inputting the physiological data into an artificial intelligence algorithm model, evaluating the injury severity of the injured mouse in real time and generating a corresponding evaluation result; the evaluation result is used for grading the severity of the injury, and a corresponding mouse craniocerebral combined thoracic and abdominal multiple injury model is established according to the grading result. According to the technical scheme, accurate simulation and grading evaluation of multi-part injuries of the mouse are achieved, and multi-injury mechanism exploration and intervention strategy effect verification are effectively supported through multi-modal physiological parameter real-time monitoring and artificial intelligence dynamic analysis.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of injury models, in particular to an artificial intelligence dynamic evaluation mouse multiple injury model construction method and system. BACKGROUND

[0002] With the high incidence of multiple injuries in trauma research and the complex physiological cascade reaction, it is an urgent need for basic research and translation research to establish a scientific, controllable and dynamically evaluated animal model. In practical applications, craniocerebral injury combined with chest and abdominal injury is often accompanied by multiple system interaction imbalance such as autonomic nervous balance disorder, circulatory and respiratory dysfunction. The early death risk of individuals with severe craniocerebral injury is highly related to multiple parameters such as lactate level, base excess and injury score, which puts higher requirements on the dynamic classification and scientific determination of key time windows of experimental animal models. The traditional single injury model has been difficult to reflect the complex evolution and cascade mechanism of multiple injuries in the physiological process.

[0003] At present, the existing animal model technology mainly has three outstanding problems. First, the controllability of injury is limited, such as traditional free fall impact, liver and spleen puncture, etc. The injury intensity only depends on the mechanical parameters, and the linear correlation between physiological response and input parameters cannot be realized, resulting in inaccurate model classification and poor repeatability. Craniocerebral and thoracoabdominal injuries are often induced separately, ignoring the essential characteristics of multiple injuries in cross-chamber biomechanical response and physiological coupling. Second, the dynamic monitoring and multi-modal integration mechanism are missing. The existing model relies on end-point behavioral tests (such as water maze, rotarod experiment) or single physiological parameters, making it difficult to realize real-time continuous monitoring of key signals such as intracranial pressure, blood pressure and respiratory rate, thus missing key processes such as shock compensation and secondary brain injury. In addition, behavioral and histological indicators are influenced by subjective interpretation, making it difficult to quantify complex injury dynamics. Third, there are obvious deficiencies in cross-system interaction modeling and individual adaptation. Traditional finite element models and stress wave propagation simulations are mostly applied to large individuals, and the simulation accuracy of small mouse individual structural differences and cross-chamber mechanical coupling is limited, making it difficult to realize the cascade reaction and precise modeling of multiple injuries.

[0004] At the same time, although artificial intelligence technology has achieved auxiliary decision-making in some fields, dynamic and real-time intelligent classification in animal models is still a blank. The existing scoring system or decision-making model based on static data is mostly used for single risk prediction, and lacks the ability to integrate and incrementally learn dynamic physiological signals. The problems of inconsistent data standards in multiple centers and insufficient model generalization ability limit the high universality and consistency of intelligent classification tools in animal experiments. The existing technology needs to make breakthroughs in injury controllability, dynamic monitoring, multi-modal data fusion, cross-system interaction and intelligent classification. How to properly solve the above problems has become a topic that the industry urgently needs to solve. SUMMARY

[0005] The application provides an artificial intelligence dynamic evaluation mouse multiple injury model construction method and system, which improves the controllability and scientificity of the mouse multiple injury model, realizes intelligentization and standardization of injury grading, and is helpful for mechanism research and intervention effect verification.

[0006] According to a first aspect of the application, an artificial intelligence dynamic evaluation mouse multiple injury model construction method is provided, which comprises: Applying a controlled impact to the mouse's cranium to cause brain injury, and introducing a controlled trauma to the chest and abdomen to simulate thoracic and abdominal injuries; Real-time monitoring of physiological parameters of the injured mouse by a sensing device, and collection of physiological data reflecting the influence of craniocerebral and chest and abdominal injuries, the physiological parameters including any one or more of intracranial pressure, blood pressure and respiratory rate; Inputting the physiological data into an artificial intelligence algorithm model, performing time series analysis and cross-system data fusion, real-time evaluation of the injury severity of the injured mouse and generation of a corresponding evaluation result; Using the evaluation result to grade the injury severity, and establishing a corresponding mouse craniocerebral combined with chest and abdominal multiple injury model accordingly.

[0007] In one embodiment, the controlled impact applied to the mouse's cranium to cause brain injury comprises: Implanting a pressure sensor in the mouse's cranial cavity to monitor intracranial pressure in real time; Adjusting the force of the impact device according to the feedback of the intracranial pressure to achieve a predetermined degree of craniocerebral injury.

[0008] In one embodiment, the introduction of a controlled trauma to the chest and abdomen to simulate thoracic and abdominal injuries comprises: Inserting a miniature hydraulic puncture needle into the mouse's thoracic or abdominal cavity to pierce the internal organs and induce bleeding; A micro-flow pump associated with the puncture needle dynamically adjusts the bleeding rate according to the blood pressure changes of the mouse, thereby simulating the process of uncontrolled hemorrhagic shock.

[0009] In one embodiment, the real-time monitoring of the physiological parameters of the injured mouse comprises: Implanting an optical fiber intracranial pressure sensor in the mouse's cranial cavity for monitoring changes in intracranial pressure; Implanting a flexible strain sensor in the mouse's thoracic cavity or thoracic wall for monitoring respiratory rate and judging the status of hemopneumothorax; Implanting a miniature ultrasonic probe in the mouse's abdominal cavity for monitoring the amount of bleeding in the abdominal cavity; Implanting a microneedle blood lactic acid sensor in the mouse's body for monitoring systemic metabolic disorder;

[0010] In one embodiment, further comprising: inputting intracranial pressure, respiratory rate, amount of hemorrhage and blood lactic acid collected by the sensor as multi-modal time series data into an artificial intelligence model; analyzing the time series data by a recurrent neural network to identify key events in the injury process and the correlation between different physiological parameters; outputting a real-time updated multiple injury severity score and / or an early warning signal.

[0011] In one embodiment, further comprising: performing a micro-CT scan on the mouse before injury to obtain an individual anatomical atlas of the mouse including skull thickness, brain volume and thoracic cavity structure; adjusting the impact position and angle of the cranial brain injury based on the anatomical atlas to reduce the injury dispersion caused by individual differences; simulating the propagation path of mechanical stress in the thoracic cavity to the brain by a finite element model, and calibrating the parameters of thoracoabdominal injury to achieve a predetermined degree of secondary brain injury.

[0012] According to a second aspect of the present application, a system for constructing an artificial intelligence dynamic evaluation mouse multiple injury model is provided, comprising: an injury module for applying a controlled impact to the mouse's brain to cause brain injury, and introducing a controlled trauma to the thoracoabdominal region to simulate thoracic and abdominal injuries; a monitoring module for monitoring physiological parameters of the injured mouse in real time through a sensing device, and collecting physiological data reflecting the impact of cranial brain and thoracoabdominal injury, the physiological parameters including any one or more of intracranial pressure, blood pressure and respiratory rate; an analysis module for inputting the physiological data into an artificial intelligence algorithm model, performing time series analysis and cross-system data fusion, and real-time evaluating the injury severity of the injured mouse and generating a corresponding evaluation result; a grading module for using the evaluation result to grade the injury severity, and establishing a corresponding mouse cranial brain combined thoracoabdominal multiple injury model accordingly.

[0013] In one embodiment, the injury module, the monitoring module, the analysis module and the grading module are controlled to perform any one of the above artificial intelligence dynamic evaluation mouse multiple injury model construction methods.

[0014] According to a third aspect of the present application, an electronic device is provided, comprising a communication interface, a processor, a memory; wherein the memory is configured to store program instructions, and the program instructions, when executed by the processor in communication connection with the memory through the communication interface, implement any one of the above artificial intelligence dynamic evaluation mouse multiple injury model construction methods.

[0015] According to a fourth aspect of the present application, a computer readable storage medium is provided, and the computer readable storage medium stores computer program instructions, and the computer program instructions are executed by a computer (for example, a processor in the computer) to implement any one of the artificial intelligence dynamic evaluation mouse multiple injury model construction methods described above.

[0016] To sum up, the present application provides an artificial intelligence dynamic evaluation mouse multiple injury model construction method and system, which comprises: applying a controlled impact to the mouse's brain to cause brain injury, and introducing a controlled trauma to the chest and abdomen to simulate thoracic and abdominal injury; real-time monitoring of the physiological parameters of the injured mouse through a sensing device, and collecting physiological data reflecting the influence of brain and chest and abdominal injury, the physiological parameters including any one or more of intracranial pressure, blood pressure and respiratory rate; inputting the physiological data into an artificial intelligence algorithm model, performing time series analysis and cross-system data fusion, real-time evaluating the injury severity of the injured mouse and generating the corresponding evaluation results; using the evaluation results to grade the injury severity, and accordingly establishing a corresponding mouse brain combined with chest and abdominal multiple injury model. The technical solution of the present application realizes the controllable and high repeatability of the mouse brain combined with chest and abdominal multiple injury model through controlled injury and real-time multi-modal physiological monitoring. Combined with the artificial intelligence algorithm for dynamic evaluation of continuous physiological data, the change of injury severity can be accurately reflected, and the intelligentization and standardization of grading evaluation are realized. The scheme overcomes the limitations of traditional models in injury controllability, dynamic monitoring and single evaluation mode.

[0017] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art from the description, or can be learned by practice of the present application. The objects and other advantages of the present application can be achieved and obtained by means of the structures particularly pointed out in the written description and accompanying drawings.

[0018] The technical solutions of the present application will be further described in detail below with the help of the accompanying drawings and examples. BRIEF DESCRIPTION OF DRAWINGS

[0019] In order to more clearly illustrate the specific embodiments of the present application or the technical solutions in the prior art, the following will briefly introduce the drawings needed to be used in the specific embodiments or prior art description. Obviously, the drawings described below are some embodiments of the present application, and those skilled in the art can also obtain other drawings according to these drawings without creative labor.

[0020] Figure 1 A flow chart of an artificial intelligence dynamic evaluation mouse multiple injury model construction method provided for an embodiment of the present application; Figure 2 A flow chart of another artificial intelligence dynamic evaluation mouse multiple injury model construction method provided for an embodiment of the present application is shown in FIG. 6; Figure 3 A flow chart of another artificial intelligence dynamic evaluation mouse multiple injury model construction method provided for an embodiment of the present application is shown in FIG. 6; Figure 4 A flow chart of another artificial intelligence dynamic evaluation mouse multiple injury model construction method provided for an embodiment of the present application is shown in FIG. 6; Figure 5 A flow chart of another artificial intelligence dynamic evaluation mouse multiple injury model construction method provided for an embodiment of the present application is shown in FIG. 6; Figure 6 A flow chart of another artificial intelligence dynamic evaluation mouse multiple injury model construction method provided for an embodiment of the present application is shown in FIG. 6; Figure 7 A structural diagram of an artificial intelligence dynamic evaluation mouse multiple injury model construction system provided for an embodiment of the present application is shown in FIG. 7; Figure 8 A structural diagram of an electronic device provided for an embodiment of the present application is shown in FIG. 8. DETAILED DESCRIPTION

[0021] The features and exemplary embodiments of various aspects of the present application will be described in detail below with reference to the drawings. The following detailed description is merely intended to explain the present application, and is not intended to limit the present application. The present application can be implemented without some of the specific details, which will be apparent to those skilled in the art. The following description of the embodiments is merely intended to provide a better understanding of the present application by showing examples of the present application.

[0022] It should be noted that, in this document, the terms such as first and second are merely used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between the entities or operations. Moreover, the terms "include", "contain" 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 not only includes those elements, but also includes other elements not explicitly listed or inherent to such a process, method, article or device. Without more limitations, the elements defined by the statement "include" do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0023] As Figure 1As shown, the present application provides a method for constructing an artificial intelligence dynamic assessment mouse multiple injury model, which comprises: In step S11, controlled impact is applied to the mouse's cranium to cause brain injury, and controlled trauma is introduced in the chest and abdomen to simulate thoracic and abdominal injuries; In step S12, physiological parameters of the injured mouse are monitored in real time by a sensing device, and physiological data reflecting the influence of craniocerebral and chest and abdominal injuries are collected, the physiological parameters including any one or more of intracranial pressure, blood pressure and respiratory rate; In step S13, the physiological data is input into an artificial intelligence algorithm model for time series analysis and cross-system data fusion, and the injury severity of the injured mouse is assessed in real time and the corresponding assessment result is generated; In step S14, the assessment result is used to grade the injury severity, and a corresponding mouse craniocerebral combined with chest and abdominal multiple injury model is established.

[0024] In one embodiment, a closed-loop feedback control multiple injury modeling platform is established, the injury level is accurately controlled through mechanical feedback, and a multi-type implantable physiological sensor and an artificial intelligence algorithm are combined to collect, analyze and grade the multiple injury process in real time, providing a high-fidelity and high-universal experimental basis for mechanism research and intervention strategy verification.

[0025] For the construction of a mouse craniocerebral combined with chest and abdominal multiple injury model, closed-loop integration of injury operation and injury assessment is achieved. Unlike traditional single-site injury, controlled impact is applied to the mouse's cranium to cause brain injury, and controlled trauma is introduced in the chest and abdomen to simulate thoracic and abdominal injuries, and at the same time, controllable injuries are introduced in multiple sites on the same animal individual to be closer to the clinical multiple injury situation. Using a controlled mechanical injury method, precise injury energy management can be performed on the cranium and chest and abdomen, avoiding the randomness of injury results due to experimental conditions, animal individual differences and other factors, and improving the grading consistency and repeatability of the experimental model.

[0026] Physiological parameters of the injured mouse are monitored in real time by a sensing device, and physiological data reflecting the influence of craniocerebral and chest and abdominal injuries are collected, the physiological parameters including any one or more of intracranial pressure, blood pressure and respiratory rate, achieving multi-modal and real-time collection of physiological signals. By implanting or attaching high-sensitivity sensors in the mouse, physiological data can be dynamically acquired throughout the whole process after injury induction. These physiological parameters not only reflect the initial effect of injury, but also can track the adaptive response and physiological cascade process of the body to the combined injury in real time.

[0027] The physiological data is input into an artificial intelligence algorithm model for time series analysis and cross-system data fusion to assess the injury severity of the injured mouse in real time and generate a corresponding assessment result. Through deep learning, feature extraction and fusion of multi-modal time series physiological data, the injury degree can be automatically distinguished, trend predicted and early warned beyond the artificial threshold interpretation.

[0028] The assessment result is used to grade the injury severity, and a corresponding mouse craniocerebral combined with thoracic and abdominal multiple injury model is established accordingly. The grading result generated by artificial intelligence can actually distinguish the similarities and differences of multiple injury individuals in terms of injury degree, development trend and prognosis risk, forming a quantifiable and comparable multi-level injury model. Not only can it achieve model customization under different experimental needs, but also realize the transition of animal model from subjective experience interpretation to objective grading.

[0029] A multi-chamber coupled injury module is integrated. In the construction of craniocerebral injury, an electromagnetic driven impactor is used, which can realize real-time monitoring of intracranial pressure (ICP) through pressure sensor and dynamically adjust the impact force (range 0.5-2.5 J), ensuring that the injury degree (mild, moderate, severe) is linearly related to the input energy, solving the technical problem of nonlinear correspondence between mechanical parameters and actual injury degree of brain tissue in traditional gravity falling body model. For thoracic and abdominal injury, the system integrates a miniature hydraulic puncture needle, which dynamically regulates the puncture depth of liver and spleen (0.1-1.0 mm) and blood loss rate (5-20 μL / min) based on real-time monitoring of blood pressure (BP) and blood lactate value, which can accurately simulate the physiological process of uncontrollable hemorrhagic shock, and is superior to the previous fixed parameter mode in terms of injury dynamic controllability and model authenticity.

[0030] In terms of cross-chamber stress conduction and pathological mechanism modeling, the scheme introduces a miniature pressure wave generator implanted in the thoracic cavity, which can release controllable shock waves with a frequency of 10-50 Hz, and based on a finite element simulation model, the propagation path of stress between the brain and thoracic and abdominal cavities is pre-calculated. Through the principle of mechanical coupling, the cascading effect of secondary brain injury caused by thoracic injury can be quantified.

[0031] Through preoperative micro-CT scanning of mouse specimens, three-dimensional anatomical atlas of skull thickness, brain volume, thoracic cavity structure, etc. is obtained, and the impact position and angle are dynamically adjusted accordingly, with an adjustment accuracy of ±2°, significantly reducing the injury dispersion caused by anatomical differences.

[0032] Before model construction, the individual mouse was scanned comprehensively by micro-CT to obtain a three-dimensional anatomical atlas including skull thickness, brain volume, and thoracic cavity structure. Based on the above individualized three-dimensional atlas, the experimenters can dynamically adjust the impact position and angle during injury, with an adjustment accuracy of ±2°, significantly reducing the dispersion of injury distribution caused by individual anatomical variations of mice.

[0033] A dynamic monitoring system was constructed, and an implantable multi-modal sensor array was used to collect key physiological parameters in mice in real time and high frequency, covering 12 dimensions such as intracranial pressure, blood lactate, and respiratory rate, with a sampling frequency of 100 Hz. The implantable multi-modal sensor array is shown in Table 1 below.

[0034] Table 1

[0035] The collected multi-modal time series data are processed by a dynamic model dynamic adaptive artificial intelligence evaluation engine. The short-term memory network (LSTM) is used to identify key time window events such as short-term intracranial pressure mutation and blood lactate inflection point after injury, and the graph neural network (GNN) is used to further analyze the causal relationship between systems to output a dynamic score of multiple injuries in real time. The stage division of the dynamic model construction process is shown in Table 2 below.

[0036]

[0037] In the aspect of cross-system interaction modeling, a multi-scale finite element simulation engine based on individual CT data is introduced to construct a three-dimensional finite element grid of the brain-chest-abdomen, simulate the propagation of shock waves among cerebrospinal fluid, lung parenchyma, and abdominal organs, and calculate the local strain energy density (SED) distribution. Through microdialysis technology, the levels of GFAP in brain tissue and serum Tau protein are monitored to establish a quantitative relationship between SED and biochemical injury markers, realizing dual coupling modeling of biomechanics and biochemistry. In addition, by optogenetic stimulation of the dorsal nucleus of the vagus nerve, controllable induction of autonomic nervous dysfunction is realized to simulate the pathological progression and cascade reaction of multiple systems after multiple injuries.

[0038] A federal learning and adaptive control platform is built, and a multi-center data lake architecture supports local training of artificial intelligence sub-models in different laboratories, and improves the model generalization ability through cloud parameter aggregation. Through multi-modal data standardization, a quantitative description system of neural function and behavior is established, and multiple methods such as gait analysis and cognitive evaluation are used to comprehensively characterize the multiple injury animal model. The system has a built-in dynamic adaptive algorithm that can automatically adjust the injury parameters or implement simulated resuscitation operations based on the MDSS score and blood lactate, etc., to realize closed-loop dynamic optimization of the whole process.

[0039] To verify the controllability of injury, the technical solution designs a linear relationship comparison experiment for craniocerebral injury. The specific method is: taking C57BL / 6 mice as experimental objects (n=20 in each group), respectively using the traditional gravity falling body model (grouped according to the product of weight and falling height gcf) and the electromagnetic driving impactor (combined with pressure feedback real-time calibration) proposed in the scheme to implement craniocerebral injury, and setting the impact energy in the 1.0-2.0J interval gradient increasing. Through systematic evaluation of the relationship between tissue injury degree and input energy under the two injury modes, the results are shown in Table 3 as follows.

[0040] Table 3

[0041] To verify the dynamic modeling capability of the chest and abdominal blood loss process, a comparative experiment is designed. One group uses the traditional liver and spleen puncture method, setting a fixed blood loss amount of 0.5mL; another group uses the hydraulic puncture needle of the technical solution, and adjusts the blood loss rate (5-20μL / min) dynamically, and controls according to the real-time monitored blood pressure drop slope. The results show that in predicting the irreversible node of shock, the survival rate in the traditional model with blood lactic acid >5mmol / L as the boundary is only 40% (the misjudgment rate is as high as 60%), while the survival rate in the dynamic blood loss group of the patent model is increased to 85%, showing that the real-time control mechanism can effectively avoid excessive blood loss.

[0042] To construct an efficient dynamic monitoring system, the technical solution adopts a multi-modal implantable sensor network to realize high-precision and real-time monitoring of key physiological parameters in mice. The implantable sensor has high sensitivity and accuracy, and can continuously capture multi-dimensional physiological signals such as intracranial pressure, blood pressure, and respiratory rate. The accuracy of the implantable sensor is shown in Table 4 as follows.

[0043] Table 4

[0044] In terms of multiple injury grading evaluation, the scheme establishes a MDSS dynamic scoring model. The training data set covers 120 mice from three laboratories (including five strains such as C57BL / 6 and BALB / c), and the test set is 30 mice from an independent center. Through systematic training and independent verification of multi-center large sample data, the efficiency of the model in cross-strain and cross-experimental scene is evaluated comprehensively, and the efficiency comparison is shown in Table 5 as follows.

[0045] Table 5

[0046] The technical solution in the embodiment realizes controllable grading and high repeatability of the mouse craniocerebral combined with thoracic and abdominal multiple injury model through controlled injury and real-time multi-modal physiological monitoring. Combined with artificial intelligence algorithm for dynamic evaluation of continuous physiological data, the change of injury severity can be accurately reflected, and intelligent and standardized grading evaluation is realized. The scheme overcomes the limitations of traditional models in controllability, dynamic monitoring and single evaluation method, and improves the physiological relevance and scientific rigor of the model.

[0047] In one embodiment, as shown in Figure 2 Step S11 includes steps S21-S22 as follows: In step S21, a pressure sensor is implanted in the mouse cranial cavity to monitor intracranial pressure in real time; In step S22, the force of the impact device is adjusted according to the feedback of the intracranial pressure to achieve a predetermined degree of craniocerebral injury.

[0048] In one embodiment, for the construction process of the mouse craniocerebral injury model, a controllable injury and real-time feedback mechanism of physiological parameters is constructed. By implanting a pressure sensor in the mouse cranial cavity, the intracranial pressure can be dynamically and accurately monitored. Compared with traditional animal models relying on mechanical parameters (such as weight, drop height, etc.) or end-point behavioral evaluation, the actual stress and injury degree of brain tissue during the injury process can be reflected in real time. According to the real-time monitoring results of intracranial pressure, the force of the impact device is adjusted to achieve predetermined control of the degree of craniocerebral injury. The mechanism of physiological-mechanical closed-loop control realizes the dynamic matching between injury input parameters and physiological response, and overcomes the problem of nonlinear correlation between mechanical injury parameters and actual tissue injury. Through feedback adjustment, the injury degree can be adjusted in real time during the experiment, improving the consistency of injury grading among multiple animals.

[0049] In one embodiment, as shown in Figure 3 Step S11 also includes steps S31-S32 as follows: In step S31, a micro hydraulic puncture needle is inserted into the thoracic or abdominal cavity of the mouse to pierce the internal organs and induce bleeding; In step S32, a micro flow pump linked to the puncture needle dynamically adjusts the bleeding rate according to the blood pressure changes of the mouse, thereby simulating the uncontrolled hemorrhagic shock process.

[0050] In one embodiment, according to the construction process of mouse thoraco-abdominal injury, a dynamic simulation of pathological changes in the thoraco-abdominal cavity in a multiple injury model is designed. By inserting a miniature hydraulic puncture needle into the thoracic cavity or abdominal cavity of the mouse, mechanical injury and bleeding induction of specific organs (such as liver, spleen, etc.) can be accurately achieved. Compared with traditional mechanical puncture or tissue cutting, it is more controllable and helps to achieve standardization of injury range and site. The accurate implementation of mechanical injury provides a stable basic environment for subsequent real-time acquisition and dynamic regulation of physiological parameters. Based on the feedback closed loop design of bleeding rate dynamic adjustment based on blood pressure changes, through the linkage system of the miniature hydraulic puncture needle and the micro flow pump, the bleeding rate can be adjusted in real time according to the real-time monitoring of the blood pressure level of the mouse, and then the complex pathological process of uncontrolled hemorrhagic shock in clinical practice is accurately simulated. This mechanism effectively overcomes the problems of fixed blood loss, single shock process and lack of dynamic adjustment in previous animal models, achieving high coupling degree of injury-physiological response.

[0051] In one embodiment, as shown in Figure 4 Step S12 includes steps S41-S44 as follows: In step S41, an optical fiber intracranial pressure sensor is implanted in the mouse cranial cavity for monitoring intracranial pressure changes; In step S42, a flexible strain sensor is implanted in the mouse thoracic cavity or thoracic wall for monitoring respiratory rate and judging the status of hemopneumothorax; In step S43, a miniature ultrasonic probe is implanted in the mouse abdominal cavity for monitoring the amount of bleeding in the abdominal cavity; In step S44, a microneedle blood lactic acid sensor is implanted in the mouse for monitoring systemic metabolic disorder.

[0052] In one embodiment, during the construction process of the mouse multiple injury model, multiple types of implantable physiological sensors are introduced to realize real-time, multi-dimensional monitoring of key physiological parameters. By implanting an optical fiber intracranial pressure sensor in the mouse cranial cavity, continuous and dynamic intracranial pressure change information can be obtained, providing objective data for evaluating brain injury degree and subsequent regulation. Compared with the traditional method of relying on endpoint behavior or single-point pathological analysis, the sensitivity to the development process of brain injury is improved, which helps to capture real-time signals of key pathological stages such as brain edema and secondary brain injury.

[0053] Flexible strain sensors on the chest cavity or chest wall are used to monitor the respiratory rate in real time, and can assist in determining important chest pathological conditions such as hemopneumothorax. A miniature ultrasonic probe is deployed in the abdominal cavity, which can non-invasively and dynamically monitor the change in the amount of bleeding in the abdominal cavity, providing high-resolution quantitative basis for evaluating abdominal organ injury and blood loss process. The microneedle blood lactic acid sensor can reflect the whole body metabolism level and oxygenation state in real time, as an important monitoring indicator of multiple injury shock progression. The introduction of these high-integration and multi-modal sensing means enriches the dimension and depth of model data. It breaks through the technical limitations of previous single or intermittent monitoring, and realizes the whole process and closed-loop collection of multi-system physiological information. Through these sensor networks, the complex responses and cascading changes of various tissues and organs of the mouse after injury can be captured in time.

[0054] In one embodiment, as shown in Figure 5 Further comprising steps S51-S53: In step S51, the intracranial pressure, respiratory rate, bleeding volume and blood lactic acid collected by the sensor are input into an artificial intelligence model as multi-modal time series data; In step S52, the time series data is analyzed by a recurrent neural network to identify key events in the injury process and the correlation between different physiological parameters; In step S53, the real-time updated multiple injury severity score and / or early warning signal are output.

[0055] In one embodiment, in the process of constructing and evaluating multiple injury animal models, the intelligent processing and analysis mechanism of multi-modal time series data is integrated, which reflects the scientific grading and early warning ability under the data-driven. By inputting various key physiological parameters such as intracranial pressure, respiratory rate, bleeding volume and blood lactic acid collected by the sensor into the artificial intelligence model in the form of time series data, a solid foundation is laid for continuous monitoring and dynamic analysis of the whole process of multiple injuries. Unlike the traditional evaluation mode which only relies on single or end-point data, this method realizes the full-dimensional information capture of the whole process of injury occurrence, development and recovery, ensuring the real-time of model evaluation.

[0056] At the data processing level, deep learning methods such as recurrent neural networks (RNN) are introduced to analyze multi-modal time series data. Recurrent neural networks are good at processing time series information and can automatically mine the potential relationships and change rules between various physiological parameters at different time points, enabling dynamic identification of key events such as brain hernia, circulatory imbalance, and metabolic deterioration during the injury process. Recurrent neural networks have memory and recursion capabilities, allowing them to integrate historical information from multiple parameters and analyze the interactions between different systems (such as the nervous, circulatory, and metabolic systems), thereby revealing the systemic pathological coupling mechanisms in complex multiple trauma models. The output is a real-time score of the severity of multiple trauma and / or early warning signals, enabling intelligent and standardized dynamic grading management. The scoring system can automatically adjust the model grading criteria based on changes in physiological parameters, reflecting individual differences and progression trends in a timely manner. The early warning mechanism sends signals before critical changes in the injury, combining multi-modal physiological data, deep learning algorithms, and intelligent scoring systems to create a closed-loop platform that spans the entire process from injury to monitoring, assessment, and early warning.

[0057] In one embodiment, as shown in FIG. 6, the method further includes the following steps S61-S63: Figure 6 In step S61, a micro-CT scan is performed on the mouse before injury to obtain individual anatomical atlas of the mouse including skull thickness, brain volume and thoracic structure; In step S62, the impact position and angle of the cranial injury are adjusted based on the anatomical atlas to reduce the injury dispersion caused by individual differences; In step S63, the propagation path of mechanical stress in the thoracic cavity to the brain is simulated by a finite element model, and the parameters of thoracoabdominal injury are calibrated to achieve the predetermined degree of secondary brain injury. In one embodiment, in the process of constructing a mouse multiple trauma model, an integrated mechanism of individualized anatomical information and biomechanical simulation is introduced to further improve the accuracy of model grading and the repeatability of experimental results. By performing a micro-CT scan on the mouse before injury, detailed anatomical data such as skull thickness, brain volume, and thoracic structure are obtained, effectively solving the problem of high dispersion of injury performance caused by large individual anatomical differences in traditional models.

[0058]

[0059] ​Under the support of anatomical atlas, the impact position and angle of brain injury are adjusted based on individual anatomical characteristics. Precise matching not only significantly reduces the injury deviation caused by different anatomical structures between different experimental animals, but also provides a strong guarantee for hierarchical modeling and reproducible experiments. Through finite element modeling, the propagation path of mechanical stress in the chest cavity to the brain is simulated, and the related parameters of chest and abdominal injury are calibrated to achieve precise control of the predetermined degree of secondary brain injury. Finite element simulation can reveal the propagation and energy distribution rules of stress waves under different anatomical structures and stress conditions, thereby providing theoretical support for the mechanical regulation and parameter setting of multiple injury models. The closed-loop process of physical simulation-physiological feedback-parameter calibration makes the multiple injury model have high refinement and controllability in terms of injury control, cross-chamber coupling and mechanism research.

[0060] In one embodiment, Figure 7 is a block diagram of an artificial intelligence dynamic evaluation mouse multiple injury model construction system according to an exemplary embodiment. As Figure 7 shown, the artificial intelligence dynamic evaluation mouse multiple injury model construction system includes an injury module 71, a monitoring module 72, an analysis module 73, and a hierarchical module 74.

[0061] The injury module 71 is used to apply a controlled impact to the mouse brain to cause brain injury, and introduce a controlled trauma to the chest and abdomen to simulate chest and abdominal injuries; The monitoring module 72 is used to monitor the physiological parameters of the injured mouse in real time through a sensing device, and collect physiological data reflecting the influence of brain and chest and abdominal injuries, the physiological parameters including any one or more of intracranial pressure, blood pressure and respiratory rate; The analysis module 73 is used to input the physiological data into an artificial intelligence algorithm model, perform time series analysis and cross-system data fusion, and real-time evaluate the injury severity of the injured mouse and generate the corresponding evaluation results; The hierarchical module 74 is used to grade the injury severity according to the evaluation results, and accordingly establish a corresponding mouse brain combined with chest and abdominal multiple injury model.

[0062] The injury module 71, the monitoring module 72, the analysis module 73 and the hierarchical module 74 included in the block diagram of the artificial intelligence dynamic evaluation mouse multiple injury model construction system are controlled to execute the artificial intelligence dynamic evaluation mouse multiple injury model construction method described in any of the above embodiments.

[0063] As Figure 8 shown, the present application provides an electronic device 800, which includes a communication interface, a processor 801, and a memory 802. The memory 802 is configured to store program instructions, and the program instructions, when executed by the processor 801 in communication connection with the memory 802, apply a controlled impact to the mouse cranium to cause brain injury, and introduce a controlled trauma to the chest and abdomen to simulate thoracic and abdominal injuries; real-time monitoring of physiological parameters of the injured mouse is performed by a sensing device, and physiological data reflecting the influence of craniocerebral and chest and abdominal injuries is collected, the physiological parameters including any one or more of intracranial pressure, blood pressure and respiratory rate; the physiological data is input into an artificial intelligence algorithm model, time series analysis and cross-system data fusion are performed, the injury severity of the injured mouse is evaluated in real time, and a corresponding evaluation result is generated; the evaluation result is used to grade the injury severity, and a corresponding mouse craniocerebral combined with chest and abdominal multiple injury model is established.

[0064] The application provides a computer readable storage medium, and computer program instructions are stored on the computer readable storage medium, and the computer program instructions are executed by a processor to apply a controlled impact to the mouse cranium to cause brain injury, and introduce a controlled trauma to the chest and abdomen to simulate thoracic and abdominal injuries; real-time monitoring of physiological parameters of the injured mouse is performed by a sensing device, and physiological data reflecting the influence of craniocerebral and chest and abdominal injuries is collected, the physiological parameters including any one or more of intracranial pressure, blood pressure and respiratory rate; the physiological data is input into an artificial intelligence algorithm model, time series analysis and cross-system data fusion are performed, the injury severity of the injured mouse is evaluated in real time, and a corresponding evaluation result is generated; the evaluation result is used to grade the injury severity, and a corresponding mouse craniocerebral combined with chest and abdominal multiple injury model is established.

[0065] It should be understood that the specific features, operations and details described above with respect to the method of the application can be similarly applied to the device and system of the application, or vice versa. In addition, each step of the method of the application described above can be performed by the corresponding component or unit of the device or system of the application.

[0066] It should be understood that each module / unit of the device of the application can be realized by software, hardware, firmware or a combination thereof, in whole or in part. Each module / unit can be embedded in a processor of a computer device in hardware or firmware form, or independent of the processor, or in software form stored in a memory of the computer device for calling by the processor to perform the operation of each module / unit. Each module / unit can be realized as an independent component or module, or two or more modules / units can be realized as a single component or module.

[0067] In one embodiment, a computer device is provided, which includes a memory and a processor, the memory having stored thereon computer instructions executable by the processor, the computer instructions, when executed by the processor, instructing the processor to perform steps of the method of an embodiment of the present application. The computer device can be broadly a server, a terminal, or any other electronic device with necessary computing and / or processing capability. In one embodiment, the computer device can include a processor, a memory, a network interface, a communication interface, etc. connected by a system bus. The processor of the computer device can be configured to provide necessary computing, processing and / or control capability. The memory of the computer device can include a non-volatile storage medium and an internal memory. The non-volatile storage medium can have stored therein or thereon an operating system, a computer program, etc. The internal memory can provide an environment for running of the operating system and the computer program in the non-volatile storage medium. The network interface and the communication interface of the computer device can be configured to connect and communicate with external devices through a network. The computer program, when executed by the processor, performs steps of the method of the present application.

[0068] The present application can be implemented as a computer readable storage medium having stored thereon a computer program which, when executed by a processor, causes steps of the method of an embodiment of the present application to be performed. In one embodiment, the computer program is distributed over a plurality of computer devices or processors coupled to a network, such that the computer program is stored, accessed and executed in a distributed manner by one or more computer devices or processors. A single method step / operation, or two or more method steps / operations, can be performed by a single computer device or processor, or by two or more computer devices or processors. One or more method steps / operations can be performed by one or more computer devices or processors, and one or more other method steps / operations can be performed by one or more other computer devices or processors. One or more computer devices or processors can perform a single method step / operation, or perform two or more method steps / operations.

[0069] As will be appreciated by one of ordinary skill in the art, the steps of the methods of the present application can be directed to relevant hardware, such as computer devices or processors, by way of computer program instructions. Computer program instructions can be stored in non-transitory computer-readable storage media that cause steps of the present application to be performed when executed by a computer device or processor. Any reference to memory, storage, databases, or other media in this regard, shall be interpreted to include a non-transitory computer-readable storage medium. Examples of non-transitory computer-readable storage media include RAM, ROM, programmable ROM (PROM), erasable programmable ROM (EPROM), electronically erasable programmable ROM (EEPROM), flash memory, magnetic tapes, floppy disks, magnetic disk drives, optical data storage drives, hard drives, solid-state drives, etc. Examples of volatile memory include RAM, external cache memory, etc.

[0070] Any of the technical features described above can be combined. Although not all possible combinations of the technical features are described, any combination of the technical features should be considered as being covered by the present specification, as long as such a combination does not result in a contradiction.

[0071] Finally, it should be noted that the above-described embodiments are merely intended to illustrate the technical solutions of the present application, but not to limit the present application; even though the present application has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that they can still make modifications to the technical solutions recorded in the above embodiments, or make equivalent replacements to some or all of the technical features; and such modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for constructing an artificial intelligence-based dynamic assessment model of multiple injuries in mice, characterized in that, include: Controlled impacts were applied to the cranium of mice to induce brain damage, and controlled trauma was introduced into the chest and abdomen to simulate thoracic and abdominal injuries. The physiological parameters of the injured mice are monitored in real time by a sensing device, and physiological data reflecting the effects of brain and chest and abdominal injuries are collected. The physiological parameters include one or more of intracranial pressure, blood pressure and respiratory rate. The physiological data is input into an artificial intelligence algorithm model for time-series analysis and cross-system data fusion to assess the severity of injury in the injured mouse in real time and generate corresponding assessment results. The assessment results were used to classify the severity of the injury, and a corresponding mouse model of multiple injuries to the brain and chest and abdomen was established accordingly.

2. The method for constructing a mouse model for dynamic evaluation of multiple injuries using artificial intelligence as described in claim 1, characterized in that, The controlled impact to the brain of mice to cause brain injury includes: A pressure sensor was implanted in the cranial cavity of a mouse to monitor intracranial pressure in real time. The impact force is adjusted based on the feedback of intracranial pressure to achieve a predetermined degree of traumatic brain injury.

3. The method for constructing a mouse model for dynamic evaluation of multiple injuries using artificial intelligence as described in claim 1, characterized in that, The controlled trauma inducing in the chest and abdomen to simulate thoracic and abdominal injuries includes: A miniature hydraulic puncture needle was inserted into the thoracic or abdominal cavity of a mouse to puncture internal organs and induce bleeding. The micro-flow pump linked to the puncture needle dynamically adjusts the bleeding rate according to the mouse's blood pressure changes, thereby simulating an uncontrolled hemorrhagic shock process.

4. The method for constructing a mouse model for dynamic evaluation of multiple injuries using artificial intelligence as described in claim 1, characterized in that, The real-time monitoring of the physiological parameters of the injured mouse includes: A fiber optic intracranial pressure sensor was implanted in the cranial cavity of a mouse to monitor changes in intracranial pressure. Flexible strain sensors were implanted in the thoracic cavity or chest wall of mice to monitor respiratory rate and determine the status of hemopneumothorax. A miniature ultrasound probe was implanted in the peritoneal cavity of a mouse to monitor the amount of bleeding in the peritoneal cavity; Microneedle-type blood lactate sensors were implanted in mice to monitor systemic metabolic disorders.

5. The method for constructing a mouse model for dynamic evaluation of multiple injuries using artificial intelligence as described in claim 1, characterized in that, Also includes: The intracranial pressure, respiratory rate, bleeding volume, and blood lactate collected by the sensors are input into the artificial intelligence model as multimodal time-series data; The time-series data were analyzed using a recurrent neural network to identify the correlations between key events and different physiological parameters during the injury process; Output real-time updated severity scores for multiple injuries and / or early warning signals.

6. The method for constructing a mouse model for dynamic evaluation of multiple injuries using artificial intelligence as described in claim 1, characterized in that, Also includes: Before injury, mice were subjected to miniature CT scans to obtain individual anatomical atlases of mice, including skull thickness, brain volume, and thoracic structure. The impact location and angle of the traumatic brain injury are adjusted based on the anatomical atlas to reduce the injury dispersion caused by individual differences. The propagation path of mechanical stress from the thoracic cavity to the brain is simulated using a finite element model, and the parameters of thoracic and abdominal injuries are calibrated accordingly to achieve the predetermined degree of secondary brain injury.

7. An artificial intelligence-based dynamic assessment system for constructing a mouse model of multiple injuries, characterized in that, include: The injury module is used to apply controlled impacts to the brain of mice to cause brain damage and to introduce controlled trauma to the chest and abdomen to simulate thoracic and abdominal injuries. The monitoring module is used to monitor the physiological parameters of the injured mouse in real time through a sensing device and to collect physiological data reflecting the effects of brain and chest and abdominal injuries. The physiological parameters include one or more of intracranial pressure, blood pressure and respiratory rate. The analysis module is used to input the physiological data into an artificial intelligence algorithm model, perform time-series analysis and cross-system data fusion, assess the severity of the injury in the injured mouse in real time, and generate corresponding assessment results. The grading module is used to grade the severity of the injury based on the assessment results, and to establish a corresponding mouse model of multiple injuries to the brain and chest and abdomen.

8. The artificial intelligence dynamic evaluation mouse multiple injury model construction system as described in claim 7, characterized in that: The damage module, the monitoring module, the analysis module, and the grading module are controlled to execute the artificial intelligence dynamic assessment method for constructing a mouse multiple injury model according to any one of claims 1 to 6.

9. An electronic device, characterized in that, include: Communication interface, processor, memory; The memory is used to store program instructions, which, when executed by the processor connected to the memory via the communication interface, enable the electronic device to implement the artificial intelligence dynamic assessment mouse multiple injury model construction method according to any one of claims 1 to 6.

10. A computer-readable storage medium having program instructions stored thereon, characterized in that, When the program instructions are executed by a computer, the computer enables the computer to implement the artificial intelligence dynamic assessment method for constructing a mouse multiple injury model as described in any one of claims 1 to 6.

Images