A method and related equipment for early warning of joint infection after cruciate ligament reconstruction surgery

Abstract

This invention provides a method and related equipment for early warning of joint infection after cruciate ligament reconstruction, applied in the field of data processing technology. This application acquires a training dataset and target electronic medical record information; preprocesses the training dataset to generate a training dataset with labeled information; processes the training dataset with labeled information based on preset processing rules to generate a training set and a validation set; acquires an initial postoperative infection early warning model matching the labeled information; trains the initial postoperative infection early warning model based on the training set and validation set to generate a target postoperative infection early warning model; preprocesses the target electronic medical record information to generate attribute information of the target user, wherein the target user's attribute information includes the target user's knee joint material parameters; processes the target user's attribute information based on the target postoperative infection early warning model to generate the target user's knee joint infection probability.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method and related equipment for early warning of joint infection after cruciate ligament reconstruction surgery. Background Technology

[0002] Anterior cruciate ligament (ACL) tears are a common knee injury among athletes. Especially in sports involving sudden stops and starts, and changes of direction, such as basketball, soccer, and skiing, ACL tears can significantly impact knee stability, necessitating arthroscopic ACL reconstruction. However, all surgical procedures carry a risk of infection; the postoperative infection rate for arthroscopic ACL reconstruction is approximately 1.4-18‰. Once postoperative infection occurs, the ligament graft and articular cartilage will be damaged, adversely affecting knee function. Furthermore, unplanned secondary surgeries and extensive antibiotic use place a significant burden on both patients and the healthcare system.

[0003] Postoperative joint infection following anterior cruciate ligament (ACL) reconstruction is a rare but serious complication, with a reported incidence of 0.14%–1.8%. It has the potential to affect articular cartilage integrity, graft ligamentization, and joint function. Although surgeons are highly vigilant about postoperative infection, there is still a delay between the onset of symptoms and diagnosis and treatment. Therefore, how to effectively diagnose and treat postoperative joint infection after symptoms appear, and even predict and prevent its occurrence, has become a major concern.

[0004] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this application is to provide a method and related equipment for early warning of joint infection after cruciate ligament reconstruction surgery, which at least partially overcomes the problems of existing technologies. This is achieved by collecting matched electronic medical record data, extracting key physiological factors, segmenting training and validation sets, training and optimizing an initial model to generate an accurate target model. This model analyzes patient attribute information, predicts infection risk, assists clinical decision-making, and maintains its accuracy through continuous updates.

[0006] Other features and advantages of this application will become apparent from the following detailed description, or may be learned in part by practice of the invention.

[0007] According to one aspect of this application, a method for early warning of joint infection after cruciate ligament reconstruction is provided, comprising: acquiring a training dataset and target electronic medical record information, wherein the training dataset consists of other electronic medical record information matching the target electronic medical record information; preprocessing the training dataset to generate a training dataset with identification information, wherein the identification information is used to characterize physiological factors affecting postoperative joint infection; processing the training dataset with identification information based on preset processing rules to generate a training set and a validation set; acquiring an initial postoperative infection early warning model matching the identification information, wherein the initial postoperative infection early warning model is set based on target electronic medical record information received within a preset time period; training the initial postoperative infection early warning model based on the training set and the validation set to generate a target postoperative infection early warning model; preprocessing the target electronic medical record information to generate attribute information of a target user, wherein the attribute information of the target user includes knee joint material parameter information of the target user; and processing the attribute information of the target user based on the target postoperative infection early warning model to generate the knee joint infection probability of the target user.

[0008] In one embodiment of this application, the preprocessing of the training dataset to generate a training dataset with identification information includes: extracting features from the training dataset to determine an original feature library; dividing the original feature library into various feature datasets to generate a training dataset and a validation set; using a classifier to predict the various validation sets in the original feature library and determining the prediction results; using a preset algorithm to train on the various training datasets in the original feature library to obtain the validation set class prediction results; and generating training samples with identification information based on the prediction results and the validation set class prediction results.

[0009] In one embodiment of this application, the step of extracting features from the training dataset to determine the original feature library includes: processing standard material information based on preset processing rules to generate standardized features, wherein the standardized features are features of a motif with clearly defined and known material composition; processing the standardized features based on preset feature filtering and dimensionality reduction rules to generate original features; and generating an original feature library from several original features.

[0010] In one embodiment of this application, training the initial postoperative infection early warning model based on the training set and the validation set to generate a target postoperative infection early warning model includes: extracting multiple sets of data from the training set, wherein each set of data contains a preset number of data samples, and at least one data sample includes identification information; training the initial postoperative infection early warning model based on the data samples in the multiple sets of data to generate a trained postoperative infection early warning model; processing the trained postoperative infection early warning model based on the validation set to generate a validation result; if the data sample containing identification information in the validation result indicates that the physiological factors affecting postoperative joint infection are in an abnormal state, then the trained postoperative infection early warning model is used as the target postoperative infection early warning model.

[0011] In one embodiment of this application, the target electronic medical record information is preprocessed to generate target user attribute information, including: performing data cleaning processing on the target electronic medical record information to generate target user information; processing the target user information based on a preset image extraction model to generate knee joint image information of the target user; and processing the knee joint image information of the target user to generate knee joint material parameter information of the target user, wherein the knee joint material parameter information includes the concentration of chemical components of synovial fluid, the thickness of articular cartilage, and the density of surrounding tissues.

[0012] Generate the target user's attribute information based on the target user's knee joint material parameter information;

[0013] The method includes a calculation formula for obtaining the concentration of chemical components in synovial fluid, the calculation formula being:

[0014]

[0015] Where C is the concentration of the chemical component, and S... water It is the water signal intensity in the synovial fluid, S baseine It is the baseline signal strength, S ref It is the signal strength of the reference material;

[0016] The method includes a formula for calculating the thickness of articular cartilage, the formula being:

[0017]

[0018] Where T is the average thickness of the cartilage, and D... max It is the maximum depth of cartilage, D min It is the minimum depth of cartilage;

[0019] The method includes a formula for calculating the density of tissues around the joint, the formula being:

[0020]

[0021] Where D is the average density of the tissue, HU i is the Hounsfield unit value of the i-th voxel, and n is the total number of voxels.

[0022] In one embodiment of this application, the target user's attribute information is processed based on the target postoperative infection early warning model to generate the target user's knee joint infection probability. The method further includes: the target postoperative infection early warning model includes a calculation formula for obtaining a knee joint infection matching value, the calculation formula being: Where yi represents the i-th label value of the current sample, pi represents the probability value of the i-th physiological feature in P, and k is the infection probability of the k main nodes.

[0023] In one embodiment of this application, the target user's attribute information is processed based on the target postoperative infection early warning model to generate the target user's knee joint infection probability, including: processing the target user's attribute information based on the target postoperative infection early warning model to generate knee joint image coding features; processing the knee joint image coding features based on the target postoperative infection early warning model to generate a knee joint infection matching value; if the knee joint infection matching value is higher than a preset matching threshold, it indicates that the target user's knee joint is in an abnormal state.

[0024] Another aspect of this application provides an early warning device for joint infection after cruciate ligament reconstruction surgery, characterized by comprising: an acquisition module for acquiring a training dataset and target electronic medical record information, wherein the training dataset consists of other electronic medical record information matching the target electronic medical record information; acquiring an initial postoperative infection early warning model matching identification information, wherein the initial postoperative infection early warning model is set based on the target electronic medical record information received within a preset time period; a processing module for preprocessing the training dataset to generate a training dataset with identification information, wherein the identification information is used to characterize physiological factors affecting postoperative joint infection; processing the training dataset with identification information based on preset processing rules to generate a training set and a validation set; training the initial postoperative infection early warning model based on the training set and the validation set to generate a target postoperative infection early warning model; preprocessing the target electronic medical record information to generate attribute information of a target user, wherein the attribute information of the target user includes knee joint material parameter information of the target user; and processing the attribute information of the target user based on the target postoperative infection early warning model to generate the knee joint infection probability of the target user.

[0025] According to another aspect of this application, an electronic device is characterized by comprising: a first processor; and a memory for storing executable instructions of the first processor; wherein the first processor is configured to execute the above-described method for early warning of joint infection after cruciate ligament reconstruction by executing the executable instructions.

[0026] According to another aspect of this application, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a second processor, implements the above-described method for early warning of joint infection after cruciate ligament reconstruction.

[0027] According to another aspect of this application, a computer program product is provided, comprising a computer program, characterized in that, when the computer program is executed by a third processor, it implements the above-described method for early warning of joint infection after cruciate ligament reconstruction.

[0028] This application provides a method and related equipment for early warning of joint infection after cruciate ligament reconstruction. In developing an early warning model for postoperative joint infection, the method first obtains a training dataset matching the target electronic medical record information and preprocesses it to generate a training dataset with labeled information. This labeled information keyly identifies physiological factors that may affect postoperative infection. Next, according to preset rules, the processed dataset is divided into a training set and a validation set, laying the foundation for model training and evaluation. Based on this, an initial postoperative infection early warning model is selected, which is constructed based on electronic medical record information collected within a specific time period. The initial model is then meticulously trained and optimized using the training and validation sets, ultimately generating an accurate early warning model, i.e., the target postoperative infection early warning model.

[0029] Subsequently, the target electronic medical record information is preprocessed to extract attribute information such as knee joint material parameters of the target user. Using a target postoperative infection early warning model, this attribute information is analyzed in depth to calculate the probability of knee joint infection in the target user. The purpose of this entire process is to provide a reliable tool to help medical professionals identify infection risks, optimize clinical decisions, and improve the quality of postoperative recovery for patients. Continuous monitoring of model performance and updates based on the latest medical data ensure the long-term effectiveness and adaptability of the model.

[0030] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0031] Figure 1 A flowchart illustrating an early warning method for joint infection after cruciate ligament reconstruction according to an embodiment of this application is shown.

[0032] Figure 2 This invention provides a schematic diagram of the structure of an early warning device for joint infection after cruciate ligament reconstruction according to an embodiment of the present application.

[0033] Figure 3 This illustration shows a schematic diagram of the structure of an electronic device according to an embodiment of this application;

[0034] Figure 4 A schematic diagram of a storage medium provided in one embodiment of this application is shown. Detailed Implementation

[0035] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0036] The following combination Figure 1 This application describes an early warning method for joint infection after cruciate ligament reconstruction according to an exemplary embodiment. It should be noted that the following application scenarios are shown only to facilitate understanding of the spirit and principles of this application, and the embodiments of this application are not limited in any way. Rather, the embodiments of this application can be applied to any applicable scenario.

[0037] In one embodiment, this application also proposes a method and related equipment for early warning of joint infection after cruciate ligament reconstruction surgery. Figure 1 A schematic flowchart illustrating an early warning method for joint infection after cruciate ligament reconstruction according to an embodiment of this application is shown. Figure 1 As shown, this method is applied to a server and includes:

[0038] S101, Obtain the training dataset and target electronic medical record information.

[0039] In one embodiment, electronic medical record data of the target user is collected, including the patient's basic information, clinical diagnosis, surgical records, treatment process, recovery status, etc. This invention does not limit this, and the applicant can select and set it according to actual needs.

[0040] The training dataset consists of other electronic medical record (EMR) data that match the target EMR information, identifying the target user group, such as patients with anterior cruciate ligament (ACL) tears. These users will be the focus of model training and analysis. Key information, such as injury type, surgical details, postoperative recovery status, and complications, is extracted from the target users' EMRs. Based on the target EMR information, similar EMR data is matched from a larger patient group to form the training dataset.

[0041] S102, preprocess the training dataset to generate a training dataset with identification information.

[0042] In one implementation, feature engineering is performed on the extracted electronic medical record information, including feature selection, feature transformation, and feature scaling. The training dataset is labeled, particularly for key clinical outcomes such as postoperative infection, to facilitate model learning. Methods for feature scaling mainly include normalization and standardization. The method used in this invention maps the original feature distribution x to a normal distribution space with a mean of 0 and a variance of 1 by calculating the standard deviation and mean of each feature item (such as height, weight, BMI, etc.), and uses this as the input to the neural network.

[0043] The identification information is used to characterize the physiological factors that affect postoperative joint infection. These physiological factors include, but are not limited to, basic patient information (age, sex, socioeconomic status, height, weight, BMI); past medical history (smoking history, alcohol consumption history, previous knee surgery history, comorbidities, history of immunosuppressant use); preoperative examination results (preoperative blood routine white blood cell count, preoperative blood glucose concentration); and surgery-related factors (time from injury to surgery, year of surgery, month of surgery, date of surgery, side of surgery, single or double bundle reconstruction, graft type, graft material, tibial fixation method, femoral fixation method, intraoperative procedures, start time of surgery, tourniquet duration, number of drainage tubes, number of incisions, and preoperative prophylactic antibiotic use).

[0044] Principal component analysis (PCA) of physiological factors influencing postoperative joint infection yields the following conclusions:

[0045] Generally, if the cumulative contribution rate of the top m influencing factors exceeds 80%, it can be considered that selecting the top m principal components can effectively preserve the information of the sample. In this experimental result, the value of m is 7, and the top 7 principal components are arranged according to their contribution rates as follows:

[0046]

[0047] In addition, imbalances in the data will be addressed, such as the possibility that there may be fewer infected cases than non-infected cases. This imbalance will be balanced using methods such as oversampling or undersampling, and the training dataset will be split into training and validation sets to evaluate the model's performance and generalization ability.

[0048] In another implementation, features are extracted from the training dataset to determine an original feature library. Key features, such as basic patient information, surgical details, and postoperative care measures, are extracted from the training dataset to construct the original feature library. The original feature library is then used to divide the dataset into different feature sets, generating training and validation sets. A classifier is used to predict on each validation set within the original feature library to determine the prediction results. A pre-defined algorithm is then used to train the model on each training dataset within the original feature library to obtain validation set prediction results. The trained model is then applied to the validation sets to generate class prediction results, evaluating the model's accuracy and generalization ability. Based on the prediction results and validation set class prediction results, training samples with labeled information are generated. The model parameters and algorithm are adjusted based on the prediction results and feedback to optimize the model and improve prediction accuracy.

[0049] S103, The training dataset with identification information is processed based on preset processing rules to generate a training set and a validation set.

[0050] In one implementation, features are extracted from a training dataset containing identification information to determine an original feature library. The original feature library is then processed using a neural network model to generate forged feature data, which differs from the feature data in the original feature library. The original feature library is then divided to generate a training dataset and a validation set, wherein at least one training dataset and one validation set are used, and the validation set includes the forged feature data. By generating forged feature data, preparation is made for subsequent validation of the initial postoperative infection warning model. That is, the forged feature data is actually abnormal data, but if the initial postoperative infection warning model fails to identify the current data as abnormal, it indicates that the current model's detection result is inaccurate.

[0051] S104, Obtain an initial postoperative infection early warning model that matches the identification information.

[0052] In one implementation, the initial postoperative infection early warning model is set based on the target electronic medical record information received within a preset time period.

[0053] A fully connected neural network (FCNN) is a fundamental artificial neural network architecture, also known as a multilayer perceptron (MLP). In a fully connected neural network, each neuron is connected to all neurons in the previous and next layers, forming a dense connection structure. By introducing nonlinear factors through activation functions, the neural network can fit complex nonlinear relationships.

[0054] The FCNN structure can be divided into an input layer, hidden layers, and an output layer. The modeling strategy adopted in this invention is as follows: the input layer dimension is set to 29, corresponding to each case data after feature scaling; there are two hidden layers with dimensions of 300 and 100 respectively, which effectively reduce the bottleneck effect through buffering; the output layer dimension is set to 1, corresponding to the binary classification result of infection / non-infection. The activation function is ReLU, the optimization function is Adam (during the experiment, it was found that using SGD gradient descent would lead to getting stuck in local minima, so the loss function was changed to Adam), and the loss function is cross-entropy.

[0055] S105, the initial postoperative infection early warning model is trained based on the training set and the validation set to generate the target postoperative infection early warning model.

[0056] In one implementation, multiple datasets are extracted from a training set, each dataset containing a predetermined number of data samples, at least one of which includes identification information. An initial postoperative infection warning model is trained based on these datasets to generate a trained postoperative infection warning model. In developing an accurate postoperative infection warning model, datasets containing a predetermined number of samples are first extracted from the training set, ensuring that each dataset contains at least one sample with key identification information. This identification information may indicate a patient's specific physiological state or infection risk factors. Next, these datasets are used to systematically train the initial postoperative infection warning model, forming the trained model.

[0057] The trained postoperative infection early warning model was processed using a validation set to generate validation results. Subsequently, the trained model was validated using an independent validation set to generate validation results. When analyzing the validation results, particular attention was paid to samples with labeled information to determine whether the model could accurately identify physiological factors affecting postoperative infection. If the model could effectively predict abnormal states, i.e., samples with abnormal physiological factors, then this trained model was considered the target postoperative infection early warning model. If the data samples containing labeled information in the validation results indicate that the physiological factors affecting postoperative joint infection are in an abnormal state, then the trained postoperative infection early warning model was used as the target postoperative infection early warning model.

[0058] Machine learning models, such as classification, regression, or deep learning models, are trained using training data. The model's performance is then evaluated on a validation set. The model parameters and structure are adjusted and optimized. The trained model is then used to predict the risk of postoperative complications such as infection in target users. Based on the model's predictions, personalized medical advice, such as preventative measures and treatment plans, is provided to doctors.

[0059] Furthermore, model optimization and iteration are ongoing processes that require continuous adjustments based on clinical feedback and validation results to improve the model's accuracy and usability. After the model is applied in clinical practice, its performance is continuously monitored and regularly evaluated to ensure it adapts to changes in the healthcare environment. Throughout this process, maintaining data privacy and ethical standards is crucial to ensuring the safe and compliant use of patient data. Through this process, healthcare professionals can build an effective early warning system to help identify infection risks in advance, optimize preventative measures, improve the quality of postoperative recovery for patients, and simultaneously reduce the burden on the healthcare system.

[0060] S106, preprocess the target electronic medical record information to generate the target user's attribute information.

[0061] In one implementation, the raw data in the electronic medical record is cleaned to ensure its accuracy and consistency, including removing duplicate records, correcting erroneous input, and formatting dates and times. Target user information is then extracted from the cleaned data, including basic information (such as age and gender), medical history (such as past illnesses and drug allergies), and surgical records (such as surgery type, surgery time, and anesthesia type).

[0062] Based on medical research and clinical guidelines, risk factors that may affect postoperative joint infection are identified. These factors include, but are not limited to: age: older patients may have lower immunity and weakened healing ability; gender: some studies suggest that gender may be associated with infection risk; smoking history: smoking may affect blood circulation and wound healing; surgical duration: longer surgical duration may increase the risk of infection; graft type: different types of grafts may have different infection risks. These factors are used as patient identification information, and relevant information is obtained from the surgical unit, nursing unit, etc., of the electronic medical record system. After the target user's surgery is completed, the above-mentioned physiological factors affecting postoperative joint infection are classified and converted. For example, the above risk factors are classified (e.g., age is divided into young, middle-aged, and elderly), and converted as needed (e.g., smoking history is converted to smoker and non-smoker).

[0063] Furthermore, based on a preset image extraction model, target user information is processed to generate knee joint image information for the target user. This preset image extraction model is used to process the target user information, extracting knee joint image information from medical imaging data. The extracted knee joint images are then further analyzed, utilizing image processing techniques to identify and quantify material parameters of the knee joint, such as articular cartilage thickness and ligament integrity. Specifically, this includes image denoising, contrast enhancement, and edge detection to improve image quality and prepare for subsequent analysis. Segmentation algorithms are used to distinguish different tissues of the knee joint, such as cartilage, ligaments, and bone—a crucial step in quantitative analysis. Features of the knee joint, such as cartilage thickness, joint space width, and ligament size, are extracted from the segmented images. Morphological operations, such as dilation, erosion, opening, and closing operations, are applied to improve the representation of image features. If the dataset contains multiple two-dimensional images, a three-dimensional model is synthesized using three-dimensional reconstruction techniques to more comprehensively evaluate the knee joint structure. Software tools are used to measure extracted features, such as cartilage volume, ligament length and width, and synovial fluid volume, analyzing the characteristics of knee joint tissues, such as the signal intensity distribution of cartilage, which helps in identifying lesions and injuries.

[0064] The system processes knee joint images of the target user to generate knee joint material parameters. Based on these parameters, detailed attribute information is generated, which is crucial for assessing knee condition and developing a treatment plan. The knee joint material parameters include the concentration of synovial fluid chemical composition, articular cartilage thickness, and periarticular tissue density.

[0065] The method includes formulas for calculating the concentration of chemical components in synovial fluid. The specific formulas are as follows:

[0066]

[0067] Where C is the concentration of the chemical component, and S... water It is the water signal intensity in the synovial fluid, S baseine It is the baseline signal strength, S ref It is the signal strength of the reference material;

[0068] The method includes obtaining a formula for calculating the thickness of articular cartilage. The specific formula is as follows:

[0069]

[0070] Where T is the average thickness of the cartilage, and D... max It is the maximum depth of cartilage, D min It is the minimum depth of cartilage;

[0071] The method includes obtaining a formula for calculating the density of tissues around the joint. The specific formula is as follows:

[0072]

[0073] Where D is the average density of the tissue, HU i is the Hounsfield unit value of the i-th voxel, and n is the total number of voxels.

[0074] The above provides a method for calculating material parameters of the knee joint, including the concentration of synovial fluid chemical composition, articular cartilage thickness, and periarticular tissue density. Calculating these parameters is crucial for assessing knee health, especially in the diagnosis and treatment of knee joint disorders. A detailed explanation of each parameter is provided below:

[0075] The concentration of chemical components in synovial fluid is a parameter measured using magnetic resonance spectroscopy (MRS) technology, which helps to understand the concentration of specific chemical components in synovial fluid.

[0076] In the formula, C represents the concentration of the chemical component, and S... water It is the signal intensity of water molecules in synovial fluid, S baseine It is the baseline signal strength, used for correction, while S ref This is the signal intensity of a reference substance with a known concentration. This formula allows us to determine the concentration of specific chemical components in synovial fluid, which can aid in the diagnosis of inflammation or other pathological conditions.

[0077] Articular cartilage thickness is measured using MRI. Articular cartilage is a smooth tissue that covers the ends of bones and helps reduce friction during joint movement. This thickness is very useful for assessing cartilage degeneration or damage.

[0078] In the formula, T represents the average thickness of the cartilage, and D... max and D min These represent the maximum and minimum thickness of cartilage measured on MRI images, respectively. This parameter helps detect cartilage wear and tear, commonly seen in conditions such as osteoarthritis.

[0079] Peri-joint tissue density, which can be measured by CT scan, is quantified by the Hounsfield Unit (HU) value.

[0080] In the formula, D represents the average density of the tissue, HU i The Hounsfield Unit value is obtained from a CT scan. This parameter helps assess fractures, osteoporosis, or other changes in bone density. Combining these parameters, doctors can obtain a comprehensive health assessment of the knee joint, which is crucial for developing treatment plans and monitoring disease progression.

[0081] In addition, by analyzing the thickness, surface smoothness, and signal intensity of cartilage, we can determine whether there is wear, degeneration, or damage; assess the structural integrity and functional status of the cruciate ligaments, collateral ligaments, and other periarticular tendons; measure the joint space width; assess joint stability and the presence of inflammation or degenerative changes; evaluate the morphology, density, and presence of fractures, bone spurs, or other bony changes; and observe whether the synovium is thickened and whether there is fluid accumulation in the joint capsule, which may indicate the presence of inflammation.

[0082] The procedure involves analyzing the soft tissues around the knee joint, such as muscles and fat pads, to check for swelling or injury. Dynamic imaging or patient activity data is used to assess the joint's range of motion and functional limitations. Biomechanical analysis of the image data is employed to understand the joint's mechanical properties and load distribution during movement. Any pathological features, such as cysts, osteophytes, or other abnormal structures, are identified and quantified. The patient's subjective pain score and functional status are correlated with image parameter values. Knee joint material parameters are used to predict the patient's potential response to different treatment options. Based on detailed knee joint properties, personalized treatment plans, including surgery, physical therapy, or pharmacological treatment, are developed.

[0083] S107, Based on the target postoperative infection early warning model, process the attribute information of the target user to generate the knee joint infection probability of the target user.

[0084] In one implementation, a postoperative infection early warning model based on the target infection risk generates an infection risk score for each patient according to the aforementioned risk factors. This may involve algorithms such as logistic regression, decision trees, and random forests to generate a preliminary infection status prediction for each target user. The output can be a numerical value of infection risk or a classification of low, medium, or high risk. The infection risk value can be obtained using the following calculation formula: Where K is the infection risk value, B is the number of decision trees, and t b (x) is the prediction result of the b-th tree for feature x. The calculated infection risk value is compared with a preset threshold to complete the classification of the corresponding risk. This application does not limit the specific preset threshold or the corresponding classification method; the applicant can set it according to actual needs.

[0085] If the infection risk value is higher than the preset threshold, the target user's attribute information is processed based on the target postoperative infection early warning model to generate knee joint image coding features. The target user's attribute information, including knee joint material parameters and other relevant clinical data, is analyzed using the target postoperative infection early warning model to convert knee joint image information into coding features. These features can represent key information in the image, such as cartilage condition and ligament integrity.

[0086] Based on the target postoperative infection early warning model, the encoded features of the knee joint image are processed to generate a knee joint infection matching value. The encoded features are further processed to extract potential patterns and trends related to postoperative infection. Using the processed encoded features and combined with the early warning model, the knee joint infection matching value is calculated. This value quantifies the similarity between the knee joint image and known infection features.

[0087] If the knee infection matching value is higher than the preset matching threshold, it indicates that the target user's knee joint is in an abnormal state. The knee infection matching value is compared with the preset matching threshold, which is determined based on clinical experience and statistical analysis. If the knee infection matching value is higher than the preset threshold, it indicates that the target user's knee joint may be in an abnormal state and there is a high risk of infection.

[0088] In addition, if the knee joint is in an abnormal state, an early warning signal is generated, prompting medical professionals to conduct further assessment and monitoring of the target user. Based on the warning results, corresponding preventive or therapeutic interventions are developed, such as antibiotic treatment, surgical intervention, or other medical measures. Users identified as high-risk are continuously monitored to ensure timely detection and treatment of any signs of infection. The early warning model is continuously optimized based on clinical feedback and new data to improve its predictive accuracy and reliability.

[0089] This application utilizes a targeted postoperative infection early warning model to predict infection based on the patient's basic information and surgery-related data. The data is highly real-time, outputting the infection probability immediately after surgery. If a patient is found to have a high probability of postoperative joint infection, an early warning message will be generated to alert clinicians and recommend a series of preventative measures, such as regular knee joint imaging examinations. After acquiring the images, they will be processed, analyzed, and parameters extracted, and compared with imaging images of infected knee joints to automatically diagnose the presence of infection.

[0090] In another implementation, the target postoperative infection early warning model includes a calculation formula for obtaining knee joint infection matching values, the calculation formula being: Where yi represents the i-th label value of the current sample, usually a binary value (0 or 1), where 1 indicates infection and 0 indicates no infection; pi represents the probability value of the i-th physiological feature in P; k is the infection probability of the k principal nodes, where k is the number of principal nodes, representing the main physiological features or risk factors considered by the model; and P is the knee infection matching value, used to measure the likelihood of knee infection. This formula calculates a weighted entropy value based on the sample label and physiological feature probabilities, where the weights are determined by the label value y. i Confirmed. If y i A value of 1 indicates that the feature is associated with infection, and its corresponding entropy value will be included in the sum; if y iIf the value is 0, the entropy of that feature is not included, meaning it is unrelated to infection. In this way, the model can assess the likelihood of a sample having knee joint infection.

[0091] In this application, a server acquires a training dataset and target electronic medical record (EMR) information. The training dataset consists of other EMR information that matches the target EMR information. The training dataset is preprocessed to generate a training dataset with labeled information, which characterizes physiological factors influencing postoperative joint infection. The labeled training dataset is then processed according to preset processing rules to generate a training set and a validation set. An initial postoperative infection warning model matching the labeled information is obtained, set based on the target EMR information received within a preset time period. The initial postoperative infection warning model is trained using the training and validation sets to generate a target postoperative infection warning model. The target EMR information is preprocessed to generate attribute information for the target user, including knee joint material parameters. The target user's attribute information is then processed based on the target postoperative infection warning model to generate the target user's knee joint infection probability. Matching EMR data is collected, key physiological factors are extracted, the training and validation sets are segmented, and the initial model is trained and optimized to generate an accurate target model. This model analyzes patient attribute information, predicts infection risk, assists clinical decision-making, and maintains its accuracy through continuous updates.

[0092] Optionally, in another embodiment based on the method described above in this application, the step of extracting features from the training dataset to determine the original feature library includes:

[0093] The standard material information is processed based on preset processing rules to generate standardized features, wherein the standardized features are features of a motif whose material composition is clearly defined and known.

[0094] The standardized features are processed based on preset feature filtering and dimensionality reduction rules to generate original features;

[0095] Generate an original feature library from several original features.

[0096] In one implementation, data processing and feature library construction are key steps in improving diagnostic and treatment outcomes in knee joint research and clinical applications. First, medical imaging and clinical data of the knee joint, including MRI, X-rays, and patient medical records, are collected to provide a foundation for analysis. Next, morphological features of the knee joint, such as cartilage thickness and ligament integrity, are extracted from the images according to pre-defined processing rules, and these features are standardized.

[0097] Furthermore, through feature selection and dimensionality reduction techniques, the most representative features are chosen from the standardized features to construct an original feature library. This feature library comprehensively reflects the state of the knee joint, providing support for the training of machine learning models. Appropriate evaluation metrics, such as accuracy, sensitivity, and specificity, are used to assess the model's diagnostic and predictive capabilities. Through training, these models can predict the type of knee injury or postoperative recovery time, assisting doctors in making more accurate diagnoses and treatment plans.

[0098] Model evaluation is a crucial step in ensuring predictive accuracy, using clinical data to validate the model's diagnostic and predictive capabilities. As new data accumulates and feedback from model evaluations is received, the original feature library will be continuously updated to improve its timeliness and accuracy. Ultimately, these validated models and feature libraries can be integrated into clinical decision support systems, providing physicians with scientific evidence to optimize patient treatment.

[0099] This process allows medical professionals in the field of knee joint surgery to gain a deeper understanding of knee joint pathology, provide personalized treatment plans for patients, and improve the success rate of surgery and treatment. This not only helps improve patient treatment outcomes and satisfaction but also provides valuable data support for knee joint disease research.

[0100] By applying the above technical solutions, the server acquires a training dataset and target electronic medical record information. The training dataset consists of other electronic medical record information that matches the target electronic medical record information. Standardized features are generated by processing standard material information according to preset processing rules. These standardized features are features of motifs with clearly defined and known material compositions. The standardized features are then processed using preset feature selection and dimensionality reduction rules to generate original features. Several original features are used to generate an original feature library. The original feature library is then divided into various feature datasets to generate training and validation datasets. A classifier is used to predict the results of each validation dataset in the original feature library. A preset algorithm is used to train the training datasets in the original feature library to obtain validation class prediction results. Based on the prediction results and validation class prediction results, training samples with labeled information are generated. This labeled information is used to characterize physiological factors affecting postoperative joint infection. The training dataset with labeled information is then processed according to preset processing rules to generate training and validation sets.

[0101] An initial postoperative infection warning model matching the identification information is obtained, wherein the initial postoperative infection warning model is set based on the target electronic medical record information received within a preset time period; multiple sets of data are extracted from the training set, wherein each set of data contains a preset number of data samples, wherein at least one data sample includes identification information; the initial postoperative infection warning model is trained based on the data samples in the multiple sets of data to generate a trained postoperative infection warning model; the trained postoperative infection warning model is processed based on the validation set to generate validation results; if the data sample containing the identification information in the validation results indicates that the physiological factors affecting postoperative joint infection are in an abnormal state, then the trained postoperative infection warning model is used as the target postoperative infection warning model; the target electronic medical record information is cleaned to generate target user information; the target user information is processed based on a preset image extraction model to generate knee joint image information of the target user; the knee joint image information of the target user is processed to generate knee joint material parameter information of the target user.

[0102] Attribute information of the target user is generated based on the target user's knee joint material parameter information, including the target user's knee joint material parameter information. The target user's attribute information is then processed based on the target postoperative infection early warning model to generate knee joint image coding features. These knee joint image coding features are then processed again based on the target postoperative infection early warning model to generate a knee joint infection matching value. The target postoperative infection early warning model includes a calculation formula for obtaining the knee joint infection matching value, which is as follows: Where yi represents the i-th label value of the current sample, pi represents the probability value of the i-th physiological feature in P, and k is the infection probability of the k main nodes; if the knee joint infection matching value is higher than the preset matching threshold, it indicates that the target user's knee joint is in an abnormal state. By collecting matched electronic medical record data, extracting key physiological factors, splitting the training and validation sets, training and optimizing the initial model to generate an accurate target model. This model analyzes patient attribute information, predicts infection risk, assists clinical decision-making, and maintains its accuracy through continuous updates.

[0103] In one implementation, such as Figure 2 As shown, this application also provides an early warning device for joint infection after cruciate ligament reconstruction, comprising:

[0104] The acquisition module 201 is used to acquire a training dataset and target electronic medical record information, wherein the training dataset consists of other electronic medical record information that matches the target electronic medical record information; and to acquire an initial postoperative infection warning model that matches the identification information, wherein the initial postoperative infection warning model is set based on the target electronic medical record information received within a preset time period.

[0105] Processing module 202 is used to preprocess the training dataset to generate a training dataset with labeling information, wherein the labeling information is used to characterize physiological factors affecting postoperative joint infection; process the training dataset with labeling information based on preset processing rules to generate a training set and a validation set; train the initial postoperative infection early warning model based on the training set and the validation set to generate a target postoperative infection early warning model; preprocess the target electronic medical record information to generate target user attribute information, wherein the target user attribute information includes the target user's knee joint material parameter information; process the target user attribute information based on the target postoperative infection early warning model to generate the target user's knee joint infection probability.

[0106] In another embodiment of this application, the processing module 202 is configured to preprocess the training dataset to generate a training dataset with identification information, including:

[0107] Feature extraction is performed on the training dataset to determine the original feature library;

[0108] Based on the original feature library, divide each feature dataset to generate a training dataset and a validation set;

[0109] The original feature library is divided into various validation sets using a classifier for prediction, and the prediction results are determined.

[0110] The original feature library is divided into various training datasets for training using a preset algorithm, and the validation set class prediction results are obtained.

[0111] Based on the prediction results and the validation set prediction results, training samples with labeling information are generated.

[0112] In another embodiment of this application, the processing module 202 is configured to perform feature extraction on the training dataset to determine the original feature library, including:

[0113] The standard material information is processed based on preset processing rules to generate standardized features, wherein the standardized features are features of a motif whose material composition is clearly defined and known.

[0114] The standardized features are processed based on preset feature filtering and dimensionality reduction rules to generate original features;

[0115] Generate an original feature library from several original features.

[0116] In another embodiment of this application, the processing module 202 is configured to train the initial postoperative infection early warning model based on the training set and the validation set to generate a target postoperative infection early warning model, including:

[0117] Multiple sets of data are extracted from the training set, wherein each set of data contains a preset number of data samples, and at least one data sample includes identification information.

[0118] The initial postoperative infection early warning model is trained based on data samples from multiple sets of data groups to generate a trained postoperative infection early warning model.

[0119] The trained postoperative infection early warning model is processed based on the validation set to generate validation results;

[0120] If the data sample containing the identification information in the verification result is in an abnormal state of physiological factors affecting postoperative joint infection, then the trained postoperative infection early warning model will be used as the target postoperative infection early warning model.

[0121] In another embodiment of this application, the processing module 202 is configured to preprocess the target electronic medical record information to generate attribute information of the target user, including:

[0122] The target electronic medical record information is cleaned to generate target user information;

[0123] The target user information is processed based on a preset image extraction model to generate knee joint image information of the target user;

[0124] The knee joint image information of the target user is processed to generate the knee joint material parameter information of the target user, wherein the knee joint material parameter information includes the chemical composition concentration of synovial fluid, the thickness of articular cartilage and the density of periarticular tissue;

[0125] Generate the target user's attribute information based on the target user's knee joint material parameter information;

[0126] The method includes a calculation formula for obtaining the concentration of chemical components in synovial fluid, the calculation formula being:

[0127]

[0128] Where C is the concentration of the chemical component, and S... water It is the water signal intensity in the synovial fluid, S baseine It is the baseline signal strength, S ref It is the signal strength of the reference material;

[0129] The method includes a formula for calculating the thickness of articular cartilage, the formula being:

[0130]

[0131] Where T is the average thickness of the cartilage, and D... max It is the maximum depth of cartilage, D min It is the minimum depth of cartilage;

[0132] The method includes a formula for calculating the density of tissues around the joint, the formula being:

[0133]

[0134] Where D is the average density of the tissue, HU i is the Hounsfield unit value of the i-th voxel, and n is the total number of voxels.

[0135] In another embodiment of this application, the processing module 202 is configured to process the attribute information of the target user based on the target postoperative infection early warning model to generate the knee joint infection probability of the target user, including:

[0136] Based on the target postoperative infection early warning model, the attribute information of the target user is processed to generate knee joint image coding features;

[0137] Based on the target postoperative infection early warning model, the encoded features of the knee joint image are processed to generate a knee joint infection matching value;

[0138] If the knee joint infection matching value is higher than the preset matching threshold, it indicates that the target user's knee joint is in an abnormal state.

[0139] In another embodiment of this application, the processing module 202 is configured to process the attribute information of the target user based on the target postoperative infection early warning model to generate the knee joint infection probability of the target user, and further includes:

[0140] The target postoperative infection early warning model includes a calculation formula for obtaining knee joint infection matching values, the calculation formula being:

[0141]

[0142] Where yi represents the i-th label value of the current sample, pi represents the probability value of the i-th physiological feature in P, and k is the infection probability of the k main nodes.

[0143] This application provides an electronic device, such as... Figure 3As shown, the electronic device 3 includes a first processor 300, a memory 301, a bus 302, and a communication interface 303. The first processor 300, the communication interface 303, and the memory 301 are connected via the bus 302. The memory 301 stores a computer program that can run on the first processor 300. When the first processor 300 runs the computer program, it executes the early warning method for joint infection after cruciate ligament reconstruction provided in any of the foregoing embodiments of this application.

[0144] The memory 301 may include high-speed random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 303 (which can be wired or wireless), such as the Internet, wide area network, local area network, or metropolitan area network.

[0145] Bus 302 can be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. Memory 301 is used to store programs. After receiving an execution instruction, the first processor 300 executes the program. The early warning method for joint infection after cruciate ligament reconstruction disclosed in any of the foregoing embodiments of this application can be applied to the first processor 300, or implemented by the first processor 300.

[0146] The first processor 300 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of the first processor 300 or by instructions in software form. The first processor 300 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be implemented by a hardware decoding processor, or by a combination of hardware and software modules in the decoding processor. The software modules may reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 301. The first processor 300 reads the information in memory 301 and, in conjunction with its hardware, completes the steps of the above method.

[0147] The electronic device provided in the above embodiments of this application and the early warning method for joint infection after cruciate ligament reconstruction provided in the embodiments of this application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the applications stored therein.

[0148] This application provides a computer-readable storage medium, such as... Figure 4 As shown, the computer-readable storage medium 401 stores a computer program, which is read and executed by the second processor 402 to implement the aforementioned method for early warning of joint infection after cruciate ligament reconstruction.

[0149] The technical solutions of this application embodiment, in essence, or the parts that contribute to the prior art, or all or part of the technical solutions, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause an electronic device (which may be an air conditioner, refrigeration unit, personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the method described in the embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0150] The computer-readable storage medium provided in the above embodiments of this application and the early warning method for joint infection after cruciate ligament reconstruction provided in the embodiments of this application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the applications stored therein.

[0151] This application provides a computer program product, including a computer program, which is executed by a third processor to implement the method described above.

[0152] The computer program product provided in the above embodiments of this application and the early warning method for joint infection after cruciate ligament reconstruction provided in the embodiments of this application are based on the same inventive concept and have the same beneficial effects as the methods used, run or implemented by their stored applications.

[0153] It should be noted that in this application, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0154] The various embodiments in this application are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of the method, electronic device, electronic device, and readable storage medium for early warning of joint infection after cruciate ligament reconstruction are basically similar to the embodiments of the method for early warning of joint infection after cruciate ligament reconstruction described above, and are therefore described simply. Relevant parts can be referred to in the description of the embodiments of the method for early warning of joint infection after cruciate ligament reconstruction described above.

[0155] While this application discloses the above information, it is not limited thereto. Any person skilled in the art can make various modifications and alterations without departing from the spirit and scope of this application; therefore, the scope of protection of this application shall be determined by the scope defined in the claims.

Claims

1. A method for early warning of joint infection after cruciate ligament reconstruction, characterized in that, include: Obtain a training dataset and target electronic medical record information, wherein the training dataset consists of other electronic medical record information that matches the target electronic medical record information; The training dataset is preprocessed to generate a training dataset with labeling information, wherein the labeling information is used to characterize the physiological factors affecting postoperative joint infection; The training dataset with identification information is processed based on preset processing rules to generate a training set and a validation set; Obtain an initial postoperative infection early warning model that matches the identification information, wherein the initial postoperative infection early warning model is set based on the target electronic medical record information received within a preset time period; The initial postoperative infection early warning model is trained based on the training set and the validation set to generate the target postoperative infection early warning model; The target electronic medical record information is preprocessed to generate target user attribute information, including: data cleaning of the target electronic medical record information to generate target user information; processing the target user information based on a preset image extraction model to generate knee joint image information of the target user; processing the knee joint image information of the target user to generate knee joint material parameter information of the target user, wherein the knee joint material parameter information includes the chemical composition concentration of synovial fluid, the thickness of articular cartilage, and the density of periarticular tissues; and generating target user attribute information based on the knee joint material parameter information of the target user; the formula for calculating the chemical composition concentration of synovial fluid is: Where C is the concentration of the chemical component; It is the water signal intensity in the synovial fluid. It is the baseline signal strength. It refers to the signal intensity of the reference material; the formula for calculating the thickness of articular cartilage is: Where T is the average thickness of the cartilage. It is the maximum depth of cartilage. It is the minimum depth of cartilage; the formula for calculating the density of tissues around the joint is: Where D is the average density of the tissue; is the Hounsfield unit value of the i-th voxel, and n is the total number of voxels. The attribute information of the target user includes the knee joint material parameter information of the target user. Based on the target postoperative infection early warning model, the attribute information of the target user is processed to generate the knee joint infection probability of the target user. This includes: processing the attribute information of the target user based on the target postoperative infection early warning model to generate knee joint image coding features; processing the knee joint image coding features based on the target postoperative infection early warning model to generate a knee joint infection matching value; if the knee joint infection matching value is higher than a preset matching threshold, it indicates that the target user's knee joint is in an abnormal state; the calculation formula for the knee joint infection matching value is as follows: ;in, This represents the i-th label value of the current sample. Let represent the probability value of the i-th physiological characteristic in P, and k be the infection probability of the k main nodes.

2. The method as described in claim 1, characterized in that, The preprocessing of the training dataset to generate a training dataset with labeling information includes: Feature extraction is performed on the training dataset to determine the original feature library; Based on the original feature library, divide each feature dataset to generate a training dataset and a validation set; The original feature library is divided into various validation sets using a classifier for prediction, and the prediction results are determined. The original feature library is divided into various training datasets for training using a preset algorithm, and the validation set class prediction results are obtained. Based on the prediction results and the validation set prediction results, training samples with labeling information are generated.

3. The method as described in claim 2, characterized in that, The step of extracting features from the training dataset to determine the original feature library includes: The standard material information is processed based on preset processing rules to generate standardized features, wherein the standardized features are features of a motif whose material composition is clearly defined and known. The standardized features are processed based on preset feature filtering and dimensionality reduction rules to generate original features; Generate an original feature library from several original features.

4. The method as described in claim 2, characterized in that, The step of training the initial postoperative infection early warning model based on the training set and the validation set to generate a target postoperative infection early warning model includes: Multiple sets of data are extracted from the training set, wherein each set of data contains a preset number of data samples, and at least one data sample includes identification information. The initial postoperative infection early warning model is trained based on data samples from multiple sets of data groups to generate a trained postoperative infection early warning model. The trained postoperative infection early warning model is processed based on the validation set to generate validation results; If the data sample containing the identification information in the verification result is in an abnormal state of physiological factors affecting postoperative joint infection, then the trained postoperative infection early warning model will be used as the target postoperative infection early warning model.

5. An early warning device for joint infection after cruciate ligament reconstruction surgery, characterized in that, For implementing the method of claim 1, the apparatus comprises: The acquisition module is used to acquire a training dataset and target electronic medical record information, wherein the training dataset consists of other electronic medical record information that matches the target electronic medical record information; and to acquire an initial postoperative infection warning model that matches the identification information, wherein the initial postoperative infection warning model is set based on the target electronic medical record information received within a preset time period. The processing module is used to preprocess the training dataset to generate a training dataset with labeled information, wherein the labeled information is used to characterize physiological factors affecting postoperative joint infection; process the training dataset with labeled information based on preset processing rules to generate a training set and a validation set; train the initial postoperative infection early warning model based on the training set and the validation set to generate a target postoperative infection early warning model; preprocess the target electronic medical record information to generate target user attribute information, wherein the target user attribute information includes the target user's knee joint material parameter information; process the target user's attribute information based on the target postoperative infection early warning model to generate the target user's knee joint infection probability.

6. An electronic device, characterized in that, include: First processor; and memory for storing executable instructions of the first processor; The first processor is configured to execute the method for early warning of joint infection after cruciate ligament reconstruction as described in any one of claims 1 to 4 by executing the executable instructions.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the second processor, it implements the early warning method for joint infection after cruciate ligament reconstruction as described in any one of claims 1 to 4.

Images