Osteotomy Spacing Stabilizer Spacing Angle Calibration System Based on Medical Big Data

Abstract

This invention relates to the field of medical device technology, specifically to a calibration system for the opening angle of an osteotomy expansion stabilizer based on medical big data. The system uses a general model generation module to analyze the bone creep patterns of the current patient based on current patient data, generating a general digital twin model. Then, a practical model generation module calibrates the general digital twin model by assimilating the data, generating a personalized practical digital twin model that can simulate the current patient's bone creep behavior. This model can accurately reproduce the unique biodynamic characteristics of the patient's bones, thus achieving precise comparison between the current patient and historical patient cases. Finally, this precise comparison is used to assess the deviation between the current patient's opening angle and the optimal opening angle range, dynamically calculating the correction angle, thereby achieving adaptive closed-loop control of the opening angle and maintaining system stability.

Description

Technical Field

[0001] This invention relates to the field of medical device technology, and more specifically to an osteotomy expansion stabilizer expansion angle calibration system based on medical big data. Background Technology

[0002] Osteotomy is a surgical procedure used to treat functional impairments caused by skeletal deformities or abnormal growth. During osteotomy, surgeons cut bone to reposition or adjust its angle to correct deformities, improve skeletal function, or restore normal range of motion in joints. An osteotomy stabilizer is a medical device used to fix and support the bone after osteotomy, helping to maintain the correct bone position and angle post-surgery. If the opening angle of the osteotomy stabilizer is inaccurate, it can lead to bone misalignment, poor healing, and even affect the patient's motor function. Therefore, calibrating the opening angle of the osteotomy stabilizer is crucial.

[0003] The typical workflow of a medical big data-based osteotomy spreader stabilizer angle calibration system is as follows: First, the intelligent osteotomy spreader records parameters such as angle and force during daily adjustments and uploads them to the system to complete real-time data acquisition. Then, the system performs a rapid similarity comparison analysis between this data and historical successful osteotomy case big data. Next, the system uses an algorithm model to determine whether the current angle is within the preset "optimal spread range" and predicts the risks that may result from deviation. Finally, the system generates specific spread angle calibration suggestions to guide adjustments to return to the optimal alignment.

[0004] In existing technologies, systems typically define "similar patients" based on static characteristics such as age, sex, bone density, osteotomy location, initial deformity angle, surgical plan, and daily parameters of the retractor. However, the creep behavior of bone under continuous traction is complex and dynamic, including individual biological characteristics such as blood supply, metabolic level, gene expression, and inflammatory response, as well as the continuous changes in the biodynamic properties of the bone and surrounding soft tissues at different stages of the traction process. Relying solely on static characteristic matching may lead to serious misjudgments: for example, two patients with similar static characteristics may have different blood supply and bone elasticity. The system may mistakenly apply a conservative treatment plan suitable for patients with slower creep to the patient with better elasticity, resulting in treatment delays; conversely, excessive traction may cause complications such as microfractures or vascular obstruction, missing the opportunity for truly personalized treatment. Summary of the Invention

[0005] To address the above technical issues, this invention provides a calibration system for the opening angle of an osteotomy stabilizer based on medical big data.

[0006] The osteotomy expansion stabilizer expansion angle calibration system based on medical big data provided in the embodiments of this application specifically includes:

[0007] The data acquisition module is used to collect multidimensional static feature data of the current patient, and at the same time, to build a historical patient database based on medical big data, as well as dynamic adjustment data of the current patient during osteotomy treatment;

[0008] The general model generation module is used to analyze the bone creep pattern of the current patient based on the current patient's multidimensional static feature data and historical patient data, and generate a general digital twin model.

[0009] A practical model generation module is used to calibrate a general digital twin model based on the dynamically adjusted data and perform data assimilation, thereby generating a practical digital twin model.

[0010] The patient comparison module is used to simulate the stress change curve of the current patient based on the digital twin practical model, and combine the historical patient data to evaluate the accurate comparability between the historical patients and the current patients at different historical expansion angles.

[0011] The correction angle acquisition module is used to obtain the optimal expansion angle range for the current patient based on the accurate comparability and the corresponding historical expansion angle, and to determine the correction angle of the current patient's expansion stabilizer.

[0012] In some embodiments of the present invention, the general model generation module includes:

[0013] The bone creep similarity analysis unit is used to analyze the similarity between the bone creep of the current patient and the corresponding multidimensional static feature data in the historical patient data.

[0014] The expansion stage availability analysis unit is used to analyze the degree of change in bone creep rate at each expansion stage based on stress values ​​in historical patient data, and to obtain the availability of each historical patient at each expansion stage.

[0015] A general model generation unit is used to generate a general digital twin model based on the degree of bone creep similarity and the availability, combined with biodynamic parameters translated from the mechanical behavior of historical patients at each stage of expansion.

[0016] In some embodiments of the present invention, the bone creep similarity analysis unit is configured as follows:

[0017] Based on the multidimensional static feature data of the current patient and the corresponding multidimensional static feature data of the historical patient data, the general similarity between the static feature data of the current patient and the historical patient in each dimension is obtained;

[0018] For each dimension of static feature data in historical patient data, patients with the same value are grouped together to form several static feature groups;

[0019] The consistency of bone creep among all patients within the static feature group was analyzed to obtain the degree of influence of static feature data of each dimension on bone creep;

[0020] By weighting the degree of similarity with the degree of influence, the degree of skeletal creep similarity between the current patient and historical patients is obtained.

[0021] In some embodiments of the present invention, the expansion phase availability analysis unit is configured as follows:

[0022] Based on the stress values ​​in historical patient data, the ratio between the cumulative stress change value in the forward period and the cumulative stress change value in the backward period is calculated at each moment in each expansion stage to obtain the degree of change in bone creep rate at each moment in each expansion stage.

[0023] Obtain the maximum change in bone creep rate at all times during each expansion phase;

[0024] Based on the maximum degree of change, the availability of each expansion stage for each historical patient is obtained.

[0025] In some embodiments of the present invention, the practical model generation module includes:

[0026] The optimal stress change curve estimation unit is used to analyze the Kalman gain of each dynamic adjustment when the current patient is undergoing osteotomy treatment based on the dynamic adjustment data, and combine it with the initial predicted stress change curve obtained based on the digital twin general model to obtain the optimal estimated stress change curve for each dynamic adjustment when the current patient is undergoing osteotomy treatment.

[0027] A strong prior probability calculation unit is used to analyze the combination availability of each biodynamic parameter in the biodynamic parameter group of the digital twin general model with other biodynamic parameters based on the degree of bone creep similarity, and to determine the strong prior probability of each biodynamic parameter group of the digital twin general model by combining the general prior probability of the biodynamic parameter group.

[0028] The practical model generation unit is used to generate a practical digital twin model based on the optimal estimated stress change curve and the strong prior probability.

[0029] In some embodiments of the present invention, the optimal stress variation curve estimation unit is configured as follows:

[0030] Based on the measured stress change curve in the dynamic adjustment data, the stress value change at adjacent moments during each dynamic adjustment when the current patient is undergoing osteotomy is analyzed, and the Kalman gain of each dynamic adjustment during the current patient is obtained.

[0031] Based on the aforementioned digital twin general model, the stress change curve of the current patient is simulated to obtain the initial predicted stress change curve;

[0032] Based on the Kalman gain, the weights of the initial predicted stress change curve and the measured stress change curve are adjusted to obtain the optimal estimated stress change curve for each dynamic adjustment during osteotomy treatment of the current patient.

[0033] In some embodiments of the present invention, the utility model generation unit is configured as follows:

[0034] Analyze the difference between the measured stress change curve and the predicted stress change curve, and construct a likelihood function;

[0035] Based on Bayes' theorem, and combining the strong prior probability and the likelihood function, the kernel of the posterior probability is obtained;

[0036] Based on the kernel of the posterior probability, a numerical optimization algorithm is used to drive the general digital twin model to infinitely approximate the optimal estimated stress change curve, thereby determining the maximum posterior probability estimate.

[0037] A practical digital twin model is generated based on the set of biodynamic parameters corresponding to the maximum a posteriori probability estimate.

[0038] In some embodiments of the present invention, the patient comparison module is configured as follows:

[0039] Based on the historical patient data, extract the historical expansion angle and the corresponding historical stress change curve of the historical patients;

[0040] Based on the aforementioned digital twin practical model, the stress change curve of the current patient is simulated under the historical expansion angle to obtain the predicted stress change curve of the current patient again.

[0041] By comparing the degree of overlap between the current patient's stress change curve and the historical stress change curve of the corresponding historical patient at the historical expansion angle, the accurate comparability between the historical patient and the current patient at different historical expansion angles can be assessed.

[0042] In some embodiments of the present invention, the correction angle acquisition module is configured as follows:

[0043] Based on the precise comparability and the corresponding historical expansion angle, the optimal expansion angle for the current patient is obtained, and the optimal expansion angle is expanded to obtain a floating range of the optimal expansion angle;

[0044] Monitor the current patient's real-time spread angle;

[0045] Determine whether the real-time opening angle is within the floating range;

[0046] If not, calculate the minimum angle difference between the real-time opening angle and the floating range to determine the corrected angle of the current patient's opening stabilizer.

[0047] In some embodiments of the present invention, the data acquisition module is configured as follows:

[0048] Collect multidimensional static feature data of the current patient, including age, gender, bone density, etiological diagnosis, osteotomy location and initial deformity angle;

[0049] The system is connected to medical big data, and the data within the medical big data is cleaned, standardized, and processed in a unified format to build a historical patient database.

[0050] During osteotomy treatment of the current patient, dynamic adjustment data of the current patient is collected, including stress change curves.

[0051] Compared with existing technologies, the osteotomy expansion stabilizer expansion angle calibration system based on medical big data provided by this invention has the following beneficial effects:

[0052] This invention collects multidimensional static feature data of the current patient through a data acquisition module, while simultaneously constructing a historical patient database based on medical big data, and dynamically adjusting data during the current patient's osteotomy treatment, providing foundational data support for subsequent accurate comparison and modeling analysis. A general model generation module analyzes the bone creep patterns of the current patient based on current and historical patient data, generating a general digital twin model that can simulate the current patient's bone creep behavior. This general digital twin model obtains the best prior estimate of biodynamics based on historical patient data with similar characteristics to the current patient; therefore, this general digital twin model is no longer a guesswork model but a model "pre-trained" with historical big data and possessing a good starting point. A practical model generation module calibrates the general digital twin model through data assimilation, generating a model that can simulate the current patient's bone creep behavior. A personalized digital twin model of the patient's skeletal creep behavior can serve as a virtual agent capable of accurately predicting the stress changes of the current patient at different expansion angles, replicating the unique biodynamic characteristics of the patient's bones. Based on this digital twin model, truly accurate similarity comparison can be achieved. Through the patient comparison module and the correction angle acquisition module, the precise comparability between the current patient and the historical patient at different historical expansion angles is evaluated, achieving a precise comparison of the individual situation of the current patient and the historical patient. Finally, through this precise comparison, the deviation between the expansion angle of the current patient's osteotomy expansion stabilizer and the optimal expansion angle range is evaluated, and the correction angle is dynamically calculated. This correction angle can effectively correct the deviation, avoid excessive movement, maintain system stability, and thus achieve adaptive closed-loop control of the expansion angle. Attached Figure Description

[0053] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0054] Figure 1 A schematic diagram of the basic components of an osteotomy expansion stabilizer expansion angle calibration system based on medical big data, provided in an embodiment of the present invention;

[0055] Figure 2 This is a schematic diagram of the basic components of another osteotomy expansion stabilizer expansion angle calibration system based on medical big data, provided as an embodiment of the present invention. Detailed Implementation

[0056] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the osteotomy distraction stabilizer distraction angle calibration system based on medical big data proposed in this invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0057] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. Terms such as “comprising,” “including,” or any other variations thereof are intended to cover a non-exclusive inclusion, such that a circuit structure, article, or device comprising a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such article or device. Without further limitation, an element defined by the phrase “comprising one…” does not exclude the presence of additional identical elements in the article or device that includes the element.

[0058] The following description, in conjunction with the accompanying drawings, details the specific solution of the osteotomy expansion stabilizer expansion angle calibration system based on medical big data provided by this invention.

[0059] Please see Figure 1 This illustrates the basic components of an osteotomy expansion stabilizer expansion angle calibration system based on medical big data, provided by an embodiment of the present invention.

[0060] like Figure 1 As shown in the figure, an embodiment of the present invention provides an osteotomy stabilizer opening angle calibration system based on medical big data. The system mainly includes a data acquisition module 10, a general model generation module 20, a practical model generation module 30, a patient comparison module 40, and a correction angle acquisition module 50. Wherein:

[0061] The data acquisition module 10 is used to collect multidimensional static feature data of the current patient, and simultaneously construct a historical patient database based on medical big data, as well as dynamic adjustment data during the current patient's osteotomy treatment. Specifically, the data acquisition module 10 is configured as follows:

[0062] First, comprehensive multidimensional data of the patient at the initial stage of osteotomy surgery is collected. This multidimensional data includes multidimensional static feature data, medical imaging data, and surgical parameter data. Multidimensional static feature data includes age, gender, bone density, etiological diagnosis, osteotomy location, and initial deformity angle. Medical imaging data includes preoperative and postoperative X-rays, CT scans, and other imaging data. Surgical parameter data includes, for example, the type of retractor and fixation method. To facilitate subsequent use, text data within the multidimensional data, such as gender and etiological diagnosis, needs to be converted from text to word vectors. The specific conversion method utilizes existing technologies, such as the Word2Vec model.

[0063] Simultaneously, the system will securely access archived historical osteotomy surgery data that has undergone anonymization. Due to the heterogeneous data sources (e.g., from different hospital information systems or equipment manufacturers), the data within the medical big data requires rigorous data cleaning, standardization, and unified format processing to ensure data quality and consistency. For example, the completion time of all mechanical data will be aligned and units will be standardized; image data will undergo standardized preprocessing and key feature extraction to construct a structured historical patient database, providing foundational data support for subsequent accurate comparison and modeling analysis. It should be noted that all historical patients in this database are those who ultimately underwent successful osteotomy surgeries.

[0064] In addition, during osteotomy treatment of the current patient, dynamic adjustment data of the current patient from the intelligent spreader is collected. The dynamic adjustment data includes the stress change curve (the complete curve of stress change over time) generated after each adjustment of the spreader angle.

[0065] The general model generation module 20 is used to analyze the bone creep patterns of the current patient based on multidimensional static feature data and historical patient data, and generate a general digital twin model. Further, such as... Figure 2 As shown, the general model generation module 20 includes a bone creep similarity analysis unit 21, a spreading stage usability analysis unit 22, and a general model generation unit 23, wherein:

[0066] The bone creep similarity analysis unit 21 is used to analyze the degree of similarity between the current patient's and historical patients' bone creep based on the multidimensional static feature data of the current patient and the corresponding multidimensional static feature data of historical patients. Specifically, the bone creep similarity analysis unit 21 is configured as follows:

[0067] First, based on the multidimensional static feature data of the current patient and the corresponding multidimensional static feature data of historical patients, the general similarity between the current patient and historical patients in each dimension of static feature data is obtained. Specifically, the calculation of general similarity uses existing techniques: if the data is an image, existing image similarity methods (such as histogram comparison) are used; if it is a curve, existing curve similarity methods (such as curve overlap rate) are used; if it is a point, existing point similarity methods (such as point distance) are used; and if it is text, existing text similarity methods (such as cosine similarity between text vectors) are used. Further details are omitted here.

[0068] Then, for the multidimensional static feature data in the historical patient data, the predictive ability of each dimension of static feature for creep behavior is first evaluated. Therefore, for each dimension of static feature data in the historical patient data, patients with the same value are grouped together, such as grouping patients who are all 42 years old into one group, forming several static feature groups.

[0069] Furthermore, the consistency of bone creep among all patients within a static feature group is analyzed. This involves calculating the variance of the creep behavior (represented by the difference in stress values ​​between the initial two moments in historical patient data) within that static feature group to determine the degree of influence of each dimension of static feature data on bone creep. If the variance of the creep behavior within a static feature group is small, it indicates that patients with the same static feature data exhibit highly consistent creep behavior. This means that the static feature (e.g., age) has a strong ability to distinguish and explain creep behavior, and therefore should be assigned a higher weight. Conversely, if the variance is large, it indicates a weak correlation between the static feature and creep behavior, and therefore should be assigned a lower weight. The following formula can then be constructed to represent the... The influence of static feature data in each dimension on bone creep:

[0070] ;

[0071] In the formula, Indicates the first The degree of influence of static feature data in each dimension on bone creep; Indicates the first There are a total of [number] dimensions of static feature data. A number; Indicates the first The static feature data of the dimension The variance of all creep behaviors under a static feature set formed by a set of numerical values; Represented by natural constant An exponential function with base 0.

[0072] Through the first The mean of the variances of creep behavior corresponding to all values ​​in the static feature data of each dimension is used to obtain the first... The degree of influence of static feature data on bone creep; if the variance of creep behavior corresponding to all values ​​is small, it means that patients with the same static feature data have highly consistent creep behavior. This means that the static feature (such as age) has a strong ability to distinguish and explain creep behavior. Therefore, the greater the influence of the static feature on bone creep, the higher the weight should be. Conversely, if the variance is large, it means that the static feature is weakly associated with creep behavior. Therefore, the smaller the influence of the static feature on bone creep, the lower the weight should be.

[0073] Similarly, the influence of other dimensions of static feature data on skeletal creep behavior can be determined as the weight for similarity measurement.

[0074] Finally, since static feature data from different dimensions have varying degrees of influence on skeletal creep behavior, similarity analysis cannot be a simple, equal comparison. Instead, it needs to highlight the role of important static features (those with a high degree of influence). The similarity of static features with strong predictive power for creep behavior (those with a high degree of influence) will have a greater impact on the final similarity score. This ensures that the "similar patients" found by the system are truly similar in biological response (creep behavior), and not merely in superficial static feature similarity. This provides a more reliable and physiologically meaningful reference case database for subsequent digital twin model calibration. Therefore, by weighting the similarity score based on the degree of influence, the skeletal creep similarity between the current patient and historical patients is obtained.

[0075] ;

[0076] In the formula, Indicates the current patient With the The degree of similarity in bone creep among historical patients; Indicates the first The degree of influence of static feature data in each dimension on bone creep; Indicates the current patient With the The first historical patient The degree of similarity between static feature data in each dimension; Indicates the current patient With the The number of common static feature data dimensions of each historical patient.

[0077] Similarly, the degree of similarity between the current patient's skeletal creep and that of each historical patient can be determined.

[0078] The expansion stage availability analysis unit 22 is used to analyze the degree of change in bone creep rate at each expansion stage based on stress values ​​in historical patient data, and to obtain the availability of each expansion stage for each historical patient.

[0079] The main time-varying characteristics of bone creep can be divided into three typical stages. The first stage is the decaying creep stage, where the creep rate gradually decreases after a rapid initial increase. Under tensile force, the internal microstructures of the bone (such as trabeculae and Haversian canals) rapidly adjust and adapt, and the deformation rate quickly reaches its peak before declining. The second stage is the steady-state creep stage, where the creep rate remains relatively constant. At this stage, the bone's repair mechanisms (such as bone formation) and damage mechanisms (such as microcrack formation) reach a dynamic equilibrium. The third stage is the accelerated creep stage, where the creep rate increases dramatically, ultimately leading to fracture. This indicates that the rate of damage accumulation (microcrack propagation and fusion) within the bone far exceeds its repair capacity, and the structure is about to become unstable.

[0080] The calibration goal of digital twin models is to find a set of parameters that can reproduce successful, healthy bone regeneration behavior. Ideally, under controlled traction, the bone stabilizes in the first two stages, achieving effective tissue regeneration. The accelerated creep phase signifies the collapse of biological repair mechanisms and instability of the mechanical structure. If accelerated creep phase data is used for calibration, the optimization algorithm will be forced to find a set of biodynamic parameters that can reproduce this "uncontrolled acceleration" behavior. This may cause the digital twin model to incorrectly represent the characteristics of extremely poor bone viscoelasticity and an extremely low damage threshold, thus becoming an incorrect model simulating "pathological failure" rather than "physiological regeneration." Therefore, data including the third stage is considered an unusable expansion phase from historical patients. In the expansion phase of the osteotomy expansion stabilizer, the main parameters involved are the magnitude, direction, and speed of the expansion force, which directly affect the healing and correction effects of the bone.

[0081] Based on the above analysis, in some embodiments of the present invention, the availability analysis unit 22 for the expansion stage analyzes the degree of change in bone creep rate at each expansion stage based on stress values ​​in historical patient data, thereby obtaining the availability of each historical patient at each expansion stage. Specifically, the availability analysis unit 22 for the expansion stage is configured as follows:

[0082] First, based on the stress values ​​in historical patient data, the ratio between the cumulative stress change value in the forward period and the cumulative stress change value in the backward period is calculated at each moment in each expansion stage to obtain the degree of change in bone creep rate at each moment in each expansion stage; then, the maximum degree of change in bone creep rate at all moments in each expansion stage is obtained.

[0083] Then, based on the maximum degree of change, the availability of each expansion stage for each historical patient is obtained. Specifically, the [missing information] is constructed. The first historical patient The availability calculation formula for each expansion phase is as follows:

[0084] ;

[0085] In the formula, Indicates the first The first historical patient Availability of each expansion phase; Indicates the first The first historical patient The first stage of expansion The stress value at each moment; Indicates the first The first historical patient The first stage of expansion The stress value at each moment; Indicates the first The first historical patient The first stage of expansion The stress value at each moment; Indicates the first The first historical patient The first stage of expansion The stress value at each moment; Represents the maximum value function; This represents the maximum and minimum value normalization function (compared to all expansion stages of all other historical patients); Represents the minimum value greater than 0, and its dimensions are the same as... same, This is to prevent the denominator from being 0, so that it is set The value is .

[0086] Similarly, the availability of each stage of expansion for each historical patient can be determined.

[0087] The general model generation unit 23 is used to generate a digital twin general model based on the degree of similarity and availability of bone creep, combined with the biodynamic parameters translated from the mechanical behavior of each expansion stage of historical patients.

[0088] Now, for every successful case in the medical database, each adjustment involves inverting and fitting a physical model, translating the original stress variation curves into sets of biodynamic parameters with clear physical meaning (such as elastic modulus and viscosity coefficient). For example, the first... The first historical patient Each expansion phase can be translated into a set of biodynamic parameters. ,in, All of these represent biokinetic parameters.

[0089] When constructing a general digital twin model for the current patient, instead of simply averaging the biodynamic parameters of all historical patients, we use the similarity of static feature data between the current patient and historical patients and the availability of each expansion stage as weights to calculate the average parameters of the model, so as to respect the systematic differences in the static features of different populations and the biodynamic properties of bones.

[0090] Based on the above analysis, in some embodiments of the present invention, a general digital twin model is generated by the general model generation unit 23 according to the degree of similarity and availability of bone creep, combined with the biodynamic parameters translated from the mechanical behavior of historical patients at each stage of expansion. Specifically, the general model generation is configured as follows:

[0091] First, after determining the degree of skeletal creep similarity between the current patient and each historical patient, and determining the availability of each expansion stage for each historical patient, the biodynamic parameters of the current patient's digital twin general model are obtained based on the degree of skeletal creep similarity and availability, combined with the translated biodynamic parameter values ​​for each expansion stage of historical patients. The current patient's model is then constructed. The first general model of digital twins The formulas for calculating each biodynamic parameter are as follows:

[0092] ;

[0093] In the formula, Indicates the current patient The first general model of digital twins One biodynamic parameter; Indicates the first The first historical patient Availability of each expansion phase; Indicates the current patient With the The degree of similarity in bone creep among historical patients; Indicates the first The first historical patient The first stage of translation One biodynamic parameter; This indicates the number of historical patients in the historical patient database; Indicates the first The number of patients in each historical stage.

[0094] Similarly, the current patient can be identified. All biodynamic parameters of the general digital twin model.

[0095] Then, all biodynamic parameters are substituted into the physical model framework to generate a model that can simulate the current patient. The existing technology for digital twin general models of bone creep behavior will not be elaborated here. Since this biodynamic parameter represents the best prior estimate of the bone biodynamic characteristics of a population with similar characteristics to the current patient, based on historical experience, substituting this set of biodynamic parameters into a physical model framework (such as the finite element analysis framework) will result in an initial digital twin general model that is no longer a guesswork model, but a model "pre-trained" with historical big data and has a good starting point.

[0096] The practical model generation module 30 is used for calibrating a general digital twin model based on dynamically adjusted data and data assimilation, thereby generating a practical digital twin model. Further, such as... Figure 2 As shown, the practical model generation module 30 includes an optimal stress variation curve estimation unit 31, a strong prior probability calculation unit 32, and a practical model generation unit 33, wherein:

[0097] The optimal stress change curve estimation unit 31 is used to analyze the Kalman gain of each dynamic adjustment when the current patient undergoes osteotomy treatment based on the dynamic adjustment data, and combine it with the initial predicted stress change curve obtained based on the digital twin general model to obtain the optimal estimated stress change curve for each dynamic adjustment when the current patient undergoes osteotomy treatment.

[0098] In data assimilation, the core role of Kalman gain is to balance the confidence level between model predictions and measured data. Its magnitude is primarily determined by the inherent uncertainties of both. The system quantifies the uncertainty of observed data by analyzing the "noise performance" (such as fluctuation amplitude and random error) of the actual stress curve. If the measured curve is highly volatile, noisy, and has low confidence, the algorithm will assign a smaller Kalman gain, placing greater trust in the model's predictions; conversely, if the measured data is smooth, stable, and has low noise, a larger gain will be assigned, placing greater trust in the observed values.

[0099] However, determining the gain solely based on statistical noise may still produce physically unreasonable fusion results. Under constant tension, the internal structure of a skeleton relaxes to adapt to the new length. This creeping behavior mechanically manifests as a steady-state decrease in stress over time. This established physical law provides strong prior knowledge for gain calculation. The system can then determine that if the measured curve shows fluctuations that contradict this trend (such as an increase in stress instead of a decrease), it can be highly likely that this is measurement noise or abnormal interference, thus dynamically reducing the gain weight for these unreliable data points in the calculation.

[0100] Based on the above analysis, in some embodiments of the present invention, the optimal stress change curve estimation unit 31 analyzes the Kalman gain of each dynamic adjustment during the current patient's osteotomy treatment based on dynamic adjustment data, and combines it with the initial predicted stress change curve obtained based on the digital twin general model to obtain the optimal estimated stress change curve for each dynamic adjustment during the current patient's osteotomy treatment. Specifically, the optimal stress change curve estimation unit 31 is configured as follows:

[0101] First, based on the measured stress change curve in the dynamically adjusted data, the stress value changes between adjacent time points during each dynamic adjustment when the current patient undergoes osteotomy treatment are analyzed, and the Kalman gain of each dynamic adjustment during the current patient's osteotomy treatment is obtained. The current patient's... When performing osteotomy The formula for calculating the dynamically adjusted Kalman gain is:

[0102] ;

[0103] In the formula, Indicates the current patient When performing osteotomy The Kalman gain is dynamically adjusted (to adjust the opening angle of the osteotomy spreader stabilizer). Indicates the current patient When performing osteotomy The duration (number of moments) of each dynamic adjustment; Indicates the current patient When performing osteotomy The first dynamic adjustment a moment and The difference in stress values ​​at each time point ( - ); Represented by natural constant An exponential function with base 0.5 is used to... Normalize and remove the influence of dimensions; Indicates the sign function (1 for greater than or equal to 0, 0 for less than 0); This indicates taking the absolute value.

[0104] Furthermore, based on a general digital twin model, the stress change curve of the current patient is simulated to obtain an initial predicted stress change curve. Specifically, the input expansion angle is the current patient's... The opening angle is dynamically adjusted each time during osteotomy treatment, thereby obtaining an initial predicted stress change curve that corresponds one-to-one with the measured stress change curve.

[0105] Then, based on the Kalman gain, the weights of the initial predicted stress change curve and the measured stress change curve are adjusted to obtain the optimal estimated stress change curve for each dynamic adjustment during osteotomy treatment of the current patient. That is, the optimal estimated stress change curve = initial predicted stress change curve + Kalman gain × (measured stress change curve - initial predicted stress change curve). If the Kalman gain is equal to 1, the optimal estimated stress change curve is the measured stress change curve. If the Kalman gain is equal to 0, the optimal estimated stress change curve is the initial predicted stress change curve. If the Kalman gain is between 0 and 1, the optimal estimated stress change curve is the weighted average of the measured stress change curve and the initial predicted stress change curve. The posterior state obtained after this fusion is the estimated curve that is statistically closest to the true and optimal one.

[0106] It should be noted that the optimal estimated stress change curve is dynamically changing. Since each dynamic adjustment of the osteotomy spreader stabilizer opening angle corresponds to a different opening angle, different opening angles correspond to different measured stress change curves and different initial predicted stress change curves. Therefore, different opening angles correspond to different optimal estimated stress change curves.

[0107] The strong prior probability calculation unit 32 is used to analyze the combination availability of each biodynamic parameter in the biodynamic parameter set of the digital twin general model with other biodynamic parameters based on the degree of skeletal creep similarity, and to determine the strong prior probability of each biodynamic parameter set of the digital twin general model by combining the general prior probability of the biodynamic parameter set.

[0108] The rationality of each biodynamic parameter value in a digital twin model depends not only on its historical prevalence but also on whether the associated parameter combinations match the current patient. Traditional priors only determine the probability of the value "high stiffness" occurring (generally using a Gaussian distribution to determine the parameter probability). However, the model may contain multiple different parameter combinations that can produce identical or extremely similar output curves. For example, a specific force-time decay curve could be caused by "high stiffness, high viscosity" tissue or "low stiffness, low viscosity" tissue. This could lead to the system calibrating a seemingly reasonable set of parameters that incorrectly reflects the underlying physiological state, thus failing to predict responses to different future stimuli.

[0109] For each parameter in each group of biodynamic parameters, if the similarity between the historical patients in which it occurs and the current patient is low, then even if the parameter combination itself is common, it is a "mismatch" guess for the current patient and should therefore be assigned a low effective prior probability.

[0110] Based on the above analysis, in some embodiments of the present invention, the strong prior probability calculation unit 32 analyzes the combination availability of each biodynamic parameter in the biodynamic parameter set of the digital twin general model with other biodynamic parameters based on the degree of skeletal creep similarity. Combined with the general prior probability of the biodynamic parameter set (obtained through existing technology, generally using the normal curve method), the strong prior probability of each biodynamic parameter set of the digital twin general model is determined. Specifically, the biodynamic parameter set of the digital twin general model is constructed. The formula for calculating strong prior probability is:

[0111] ;

[0112] in:

[0113] ;

[0114] In the formula: Biodynamic parameter set representing a general model of digital twins Strong prior probability; Represents the biodynamic parameter set The number of biodynamic parameters in biodynamics; Biodynamic parameter set representing a general model of digital twins The Middle The availability of combinations of individual biodynamic parameters with other biodynamic parameters; Biodynamic parameter set representing a general model of digital twins The general prior probability; Indicates containing the first The and the first Biodynamic parameters The historical number of patients; Indicates the first The one containing the first The and the first Biokinetic parameters Historical patients and current patients The degree of similarity in skeletal creep.

[0115] Similarly, the strong prior probabilities of all biodynamic parameter sets of the general digital twin model can be determined (each dynamic adjustment corresponds to a set of biodynamic parameters).

[0116] The practical model generation unit 33 is used to generate a practical digital twin model based on the optimal estimated stress change curve and strong prior probability.

[0117] Through the general model generation module 20, the system has generated a digital twin general model that can simulate the bone creep behavior of the current patient based on the static feature data of historical and current patients. At the start of osteotomy treatment for the current patient, when the patient makes the first adjustment of the spreader angle, the intelligent spreader records a complete stress-time response curve (stress change curve) in real time. Subsequently, the optimal stress change curve estimation unit 31 of the practical model generation module 30 calculates the Kalman gain for each dynamic adjustment during osteotomy treatment to construct the optimal estimated stress curve. Furthermore, the strong prior probability calculation unit 32 of the practical model generation module 30, guided by reflecting the intrinsic physiological condition, determines the strong prior probability of each set of biodynamic parameters in the digital twin general model. Next, by driving the practical digital twin model, the simulated predicted stress change curve is made to infinitely approach the optimal estimated stress curve (a set of biodynamic parameters can generate a predicted stress change curve by inputting into the practical digital twin model; changing different sets of biodynamic parameters makes the predicted stress change curve infinitely approach the optimal estimated stress curve).

[0118] Specifically, the practical model generation unit 33 is configured as follows:

[0119] First, the difference between the measured stress change curve and the initial predicted stress change curve is analyzed, and a likelihood function is constructed. That is, assuming the observation error follows a Gaussian distribution, the likelihood function has a negative exponential relationship with the sum of the squares of the differences between the measured and initial predicted stress change curves. It should be noted that since each dynamic adjustment of the opening angle corresponds to different measured stress change curves and different initial predicted stress change curves, each dynamic adjustment corresponds to a different likelihood function.

[0120] Then, according to Bayes' theorem, combining the strong prior probability and the likelihood function, that is, multiplying the strong prior probability by the likelihood function (multiplying the strong prior probability corresponding to each dynamic adjustment by the likelihood function), we obtain the kernel of the posterior probability after each dynamic adjustment.

[0121] Then, since the kernel of the posterior probability is a complex function in a high-dimensional parameter space (viscosity coefficient, elastic modulus, etc.), it is not possible to directly see where the maximum value is. Therefore, based on the kernel of the posterior probability, a numerical optimization algorithm (such as gradient descent) is used to drive the general digital twin model to infinitely approximate the optimal estimated stress change curve, that is, to "climb a mountain" in the high-dimensional space formed by biodynamic parameters, so as to maximize the posterior probability value. Specifically, starting from an initial parameter (biodynamic parameter) guess (such as the prior mean), the "slope" (gradient) at the current biodynamic parameter point is calculated. Then, a small step is taken along the steepest uphill direction, and this process is repeated until the "mountain top" is reached (the posterior probability no longer increases significantly). The maximum a posteriori estimation (MAP) is determined. The MAP (the "mountain top" position) corresponds to a set of biodynamic parameters, and this set of biodynamic parameters is the optimal set of biodynamic parameters. This process can also be understood as follows: in the general model of digital twins, biodynamic parameters are the input values ​​and stress change curves are the output results. By adjusting the biodynamic parameters, the stress change curves are made to infinitely approach the optimal estimated stress change curves until they can no longer get closer. At this point, the stress change curve is the maximum a posteriori probability estimate, and the set of biodynamic parameters at this point is the optimal set of biodynamic parameters.

[0122] Finally, based on the biodynamic parameter set corresponding to the maximum a posteriori probability estimate, all biodynamic parameters in the biodynamic parameter set are substituted into the physical model framework (such as the finite element analysis framework) to generate a practical digital twin model that can highly simulate the current bone creep behavior of the patient (one practical digital twin model is generated for each dynamic adjustment). Existing technologies will not be elaborated here.

[0123] The patient comparison module 40 is used to simulate the stress change curve of the current patient based on the digital twin practical model, and combine it with historical patient data to evaluate the accurate comparability between the historical patients and the current patients at different historical expansion angles.

[0124] The personalized digital twin model generated by the practical model generation unit 33, which simulates the current patient's bone creep behavior, can serve as a virtual proxy capable of accurately predicting the stress changes of the current patient at different spread angles, thus reproducing the unique biodynamic characteristics of the patient's bones. Based on this practical digital twin model, truly accurate similarity comparison can be achieved. That is, by inputting the spread angles of other historical patients (historical spread angles) into the practical digital twin model, the predicted stress change curve of the current patient at this historical spread angle can be obtained. This curve is then compared with the actual stress change curves of other historical patients at the same historical spread angle. If the two stress change curves coincide, it can be mechanically proven that the creep patterns of the two patients' bones are similar. Therefore, only historical patients who meet this condition can be considered comparable to the current patient in terms of bone creep behavior, and their historical treatment data and prognostic results can provide truly valuable decision support for the current patient's personalized treatment plan.

[0125] Based on the above analysis, in some embodiments of the present invention, the patient comparison module 40, based on a digital twin practical model, simulates the stress change curve of the current patient and, combined with historical patient data, evaluates the accurate comparability between historical patients and the current patient at different historical expansion angles. Specifically, the patient comparison module 40 is configured as follows:

[0126] First, based on historical patient data, extract the historical expansion angle of historical patients and the corresponding historical stress change curve at that historical expansion angle, that is, the actual stress change curve of historical patients at that historical expansion angle.

[0127] Furthermore, based on the practical digital twin model (the practical digital twin model obtained from the last dynamic adjustment), the stress change curve of the current patient is simulated under the historical expansion angle. That is, the historical expansion angle and the corresponding expansion time are brought into the practical digital twin model that can simulate the bone creep behavior of the current patient to obtain the re-predicted stress change curve of the current patient corresponding to the historical expansion angle (the re-predicted stress change curve here is for the convenience of distinguishing it from the aforementioned initial stress change curve and has no other special meaning).

[0128] Finally, by comparing the degree of overlap between the current patient's re-stress change curve and the historical stress change curve of historical patients at different historical expansion angles, the precise comparability of historical and current patients at different historical expansion angles is assessed. (Constructing historical expansion angles...) Patients with a history of time With current patients The formula for calculating precise comparability is:

[0129] ;

[0130] In the formula, This indicates the perspective of history. Patients with a history of time With current patients Precise comparability; This represents the maximum and minimum value normalization function (and historical patients). At other opening angles, compared with the current patient (Normalization is performed between the numerical values). Indicates the current patient Corresponding to historical perspective The predicted stress change curve (expanding the historical angle) and the corresponding stretching time The input can simulate the current patient. Stress variation curves predicted by a practical digital twin model of bone creep behavior. Indicates historical patients From a historical perspective Historical stress variation curve over time; Indicates taking the absolute value; Indicates the time during which it is stretched. Integral within the range.

[0131] Reflected in the historical perspective Current patient The re-stress change curve and historical patients The degree of overlap between the corresponding historical stress change curves is indicated by the smaller the value. This indicates that the two curves overlap more closely, demonstrating similar bone creep patterns in the two patients from a mechanical perspective. It also indicates greater comparability between the historical patient and the current patient in terms of bone creep behavior.

[0132] It should be noted that since the practical model of digital twins changes dynamically with dynamic adjustments, this precise comparability also changes dynamically with dynamic adjustments.

[0133] The correction angle acquisition module 50 is used to obtain the optimal expansion angle range for the current patient based on accurate comparability and the corresponding historical expansion angle, and to determine the correction angle of the current patient's expansion stabilizer.

[0134] For osteotomy patients selected from historical patient databases who ultimately underwent successful treatment, if the historical stress change curve corresponding to the distraction angle used in these successful cases (historical patients) during the critical stages of treatment shows a high degree of overlap with the predicted stress change curve obtained by incorporating this distraction angle into the current patient's personalized digital twin model—meaning that the successful case and the current patient have extremely high accuracy at that distraction angle—then this distraction angle may fall within a "treatment window" that simultaneously promotes bone regeneration, maintains mechanical stability, and minimizes soft tissue damage. Therefore, this distraction angle can serve as a strong reference for the optimal distraction angle for the current patient.

[0135] Based on the above analysis, in some embodiments of the present invention, the correction angle acquisition module 50 obtains the optimal expansion angle range of the current patient after the last dynamic adjustment based on accurate comparability and the corresponding historical expansion angle, and determines the current correction angle of the expansion stabilizer for the current patient. Specifically, the correction angle acquisition module 50 is configured as follows:

[0136] First, based on precise comparability and corresponding historical expansion angles, the optimal expansion angle for the current patient is obtained, i.e., the current patient's expansion angle is constructed. The formula for calculating the optimal spreading angle is:

[0137] ;

[0138] In the formula, Indicates the current patient The optimal opening angle; This indicates the perspective of history. Patients with a history of time With current patients Precise comparability; Indicates historical patients The corresponding historical angle of expansion; This indicates that the historical patient database has historical expansion angles. The number of historical patients; Indicates historical patients The corresponding number of osteotomy surgeries (number of times the bone was opened).

[0139] Furthermore, based on the predictions of the current patient-specific digital twin practical model and the comparison of large datasets of historical successful osteotomy cases, the optimal opening angle is expanded to obtain a floating range of the optimal opening angle. For example, the calculated opening angle for the current patient... The optimal opening angle is 45°. The optimal opening angles corresponding to successful osteotomy cases in the past are 44°, 44.5°, 45°, 46°, etc. In this case, the optimal opening angle can be extended forward and backward by 1° respectively, that is, the floating range of the optimal opening angle is [44°, 46°].

[0140] Then, the real-time opening angle of the current patient is monitored; and it is determined whether the real-time opening angle is within the floating range; if not, that is, the current patient's opening angle is detected to deviate from this floating range, the control algorithm PID will respond immediately and calculate the minimum angle difference between the real-time opening angle and the floating range, and determine the correction angle of the current patient's opening stabilizer. This correction angle is obtained based on the data after the last adjustment, so it can effectively correct the deviation, avoid excessive action, maintain system stability, and avoid the problem of large and inaccurate adjustment angle amplitude caused by adjusting only once based on static feature data in the existing technology.

[0141] The system outputs this correction angle as a command to drive the execution machine to fine-tune the opening angle of the osteotomy stabilizer, thus forming a complete adaptive closed-loop control system that can self-optimize based on real-time feedback from the patient. This ensures that the opening process of the osteotomy stabilizer can automatically move towards the optimal opening angle range and remain stable within the optimal opening angle range throughout the entire osteotomy surgery.

[0142] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0143] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A system for calibrating the opening angle of an osteotomy spreader stabilizer based on medical big data, characterized in that, The system includes: The data acquisition module is used to collect multidimensional static feature data of the current patient, while building a historical patient database based on medical big data, as well as dynamic adjustment data of the current patient during osteotomy treatment; The general model generation module is used to analyze the bone creep pattern of the current patient based on the current patient's multidimensional static feature data and historical patient data, and generate a general digital twin model. A practical model generation module is used to calibrate a general digital twin model based on the dynamically adjusted data and perform data assimilation, thereby generating a practical digital twin model. The patient comparison module is used to simulate the stress change curve of the current patient based on the digital twin practical model, and combine the historical patient data to evaluate the accurate comparability between the historical patients and the current patients at different historical expansion angles. The correction angle acquisition module is used to obtain the optimal expansion angle range for the current patient based on the accurate comparability and the corresponding historical expansion angle, and to determine the correction angle of the current patient's expansion stabilizer. The general model generation module includes: The bone creep similarity analysis unit is used to analyze the similarity between the bone creep of the current patient and the corresponding multidimensional static feature data in the historical patient data. The expansion stage availability analysis unit is used to analyze the degree of change in bone creep rate at each expansion stage based on stress values ​​in historical patient data, and to obtain the availability of each expansion stage for each historical patient. A general model generation unit is used to generate a general digital twin model based on the degree of bone creep similarity and the availability, combined with biodynamic parameters translated from the mechanical behavior of historical patients at each stage of expansion. The availability analysis unit for the expansion phase is configured as follows: Based on the stress values ​​in historical patient data, the ratio between the cumulative stress change value in the forward period and the cumulative stress change value in the backward period is calculated at each moment in each expansion stage to obtain the degree of change in bone creep rate at each moment in each expansion stage. Obtain the maximum change in bone creep rate at all times during each expansion phase; Based on the maximum degree of change, the availability of each historical patient at each expansion stage is obtained, and the first... The first historical patient The availability calculation formula for each expansion phase is as follows: ； In the formula, Indicates the first The first historical patient Availability of each expansion phase; Indicates the first The first historical patient The first stage of expansion The stress value at each moment; Indicates the first The first historical patient The first stage of expansion The stress value at each moment; Indicates the first The first historical patient The first stage of expansion The stress value at each moment; Indicates the first The first historical patient The first stage of expansion The stress value at each moment; Represents the maximum value function; This represents the maximum and minimum value normalization function; Represents the minimum value greater than 0, and its dimensions are the same as... same; The patient comparison module is configured as follows: Based on the historical patient data, extract the historical expansion angle and the corresponding historical stress change curve of the historical patients; Based on the aforementioned digital twin practical model, the stress change curve of the current patient is simulated under the historical expansion angle to obtain the predicted stress change curve of the current patient again. By comparing the degree of overlap between the current patient's re-stress change curve and the historical stress change curve of the corresponding historical patient at the historical expansion angle, the accurate comparability of the historical patient and the current patient at different historical expansion angles is assessed, and the historical expansion angle is constructed. Patients with a history of time With current patients The formula for calculating precise comparability is: ； In the formula, This indicates the perspective of history. Patients with a history of time With current patients Precise comparability; This represents the maximum and minimum value normalization function; Indicates the current patient Corresponding to historical perspective The predicted stress change curve is then re-evaluated. Indicates historical patients From a historical perspective Historical stress variation curve over time; Indicates taking the absolute value; Indicates the time during which it is stretched. Integral within the range.

2. The osteotomy spreader stabilizer spreader angle calibration system based on medical big data according to claim 1, characterized in that, The bone creep similarity analysis unit is configured as follows: Based on the multidimensional static feature data of the current patient and the corresponding multidimensional static feature data of the historical patient data, the general similarity between the static feature data of the current patient and the historical patient in each dimension is obtained. For each dimension of static feature data in historical patient data, patients with the same value are grouped together to form several static feature groups; The consistency of bone creep among all patients within the static feature group was analyzed to obtain the degree of influence of static feature data of each dimension on bone creep; By weighting the degree of similarity with the degree of influence, the degree of skeletal creep similarity between the current patient and historical patients is obtained.

3. The osteotomy spreader stabilizer spreader angle calibration system based on medical big data according to claim 1, characterized in that, The practical model generation module includes: The optimal stress change curve estimation unit is used to analyze the Kalman gain of each dynamic adjustment when the current patient is undergoing osteotomy treatment based on the dynamic adjustment data, and combine it with the initial predicted stress change curve obtained based on the digital twin general model to obtain the optimal estimated stress change curve for each dynamic adjustment when the current patient is undergoing osteotomy treatment. A strong prior probability calculation unit is used to analyze the combination availability of each biodynamic parameter in the biodynamic parameter group of the digital twin general model with other biodynamic parameters based on the degree of bone creep similarity, and to determine the strong prior probability of each biodynamic parameter group of the digital twin general model by combining the general prior probability of the biodynamic parameter group. A practical model generation unit is used to generate a practical digital twin model based on the optimal estimated stress change curve and the strong prior probability. Biodynamic parameter set for constructing a general digital twin model The formula for calculating strong prior probability is: ; in: ； In the formula: Biodynamic parameter set representing a general model of digital twins Strong prior probability; Represents the biodynamic parameter set The number of biodynamic parameters in biodynamics; Biodynamic parameter set representing a general model of digital twins The Middle The availability of combinations of individual biodynamic parameters with other biodynamic parameters; Biodynamic parameter set representing a general model of digital twins The general prior probability; Indicates containing the first The and the first Biodynamic parameters The historical number of patients; Indicates the first The one containing the first The and the first Biodynamic parameters Historical patients and current patients The degree of similarity in skeletal creep.

4. The osteotomy spreader stabilizer spreader angle calibration system based on medical big data according to claim 3, characterized in that, The optimal stress variation curve estimation unit is configured as follows: Based on the measured stress change curve in the dynamic adjustment data, the stress value change at adjacent moments during each dynamic adjustment when the current patient is undergoing osteotomy is analyzed, and the Kalman gain of each dynamic adjustment during the current patient is obtained. Based on the aforementioned digital twin general model, the stress change curve of the current patient is simulated to obtain the initial predicted stress change curve; Based on the Kalman gain, the weights of the initial predicted stress change curve and the measured stress change curve are adjusted to obtain the optimal estimated stress change curve for each dynamic adjustment during osteotomy treatment of the current patient.

5. The osteotomy spreader stabilizer spreader angle calibration system based on medical big data according to claim 4, characterized in that, The practical model generation unit is configured as follows: Analyze the difference between the measured stress change curve and the initial predicted stress change curve, and construct a likelihood function; Based on Bayes' theorem, and combining the strong prior probability and the likelihood function, the kernel of the posterior probability is obtained; Based on the kernel of the posterior probability, a numerical optimization algorithm is used to drive the general digital twin model to infinitely approximate the optimal estimated stress change curve, thereby determining the maximum posterior probability estimate. A practical digital twin model is generated based on the set of biodynamic parameters corresponding to the maximum a posteriori probability estimate.

6. The osteotomy spreader stabilizer spreader angle calibration system based on medical big data according to claim 1, characterized in that, The correction angle acquisition module is configured as follows: Based on the precise comparability and the corresponding historical expansion angle, the optimal expansion angle for the current patient is obtained, and the optimal expansion angle is expanded to obtain the floating range of the optimal expansion angle; Monitor the current patient's real-time spread angle; Determine whether the real-time opening angle is within the floating range; If not, calculate the minimum angle difference between the real-time opening angle and the floating range to determine the corrected angle of the current patient's opening stabilizer.

7. The osteotomy spreader stabilizer spreader angle calibration system based on medical big data according to claim 1, characterized in that, The data acquisition module is configured as follows: Collect multidimensional static feature data of the current patient, including age, gender, bone density, etiological diagnosis, osteotomy location and initial deformity angle; The system is connected to medical big data, and the data within the medical big data is cleaned, standardized, and processed in a unified format to build a historical patient database. During osteotomy treatment of the current patient, dynamic adjustment data of the current patient is collected, including stress change curves.

Images