Personalized orthopedic surgery planning method based on machine learning

Abstract

The invention discloses a personalized orthopedic surgery planning method based on machine learning, and relates to the technical field of personalized medical treatment, and the method comprises the steps: collecting medical data of a patient, generating a personalized basic data set of the patient, building a risk prediction model, and obtaining a predicted surgical risk score; preprocessing the basic data set, and extracting features of the skeleton structure of the patient based on a deep learning algorithm; training the personalized orthopedic surgery planning model, optimizing the historical patient data based on the personalized orthopedic surgery planning model, and predicting an optimal surgery plan; and outputting a personalized operation scheme according to the trained machine learning model. According to the method, machine learning and personalized orthopedic surgery planning are combined, a personalized surgery scheme is customized for each patient, the most suitable surgery scheme is provided for the patient in combination with historical data and the risk prediction model, the surgical treatment effect and the life quality of the patient are remarkably improved, and the method has wide clinical application prospects.

Description

Technical Field

[0006] The present invention relates to the field of personalized medicine technology, and specifically to a personalized orthopedic surgery planning method based on machine learning. Background Art

[0007] With the continuous development of medical imaging technology and machine learning technology, personalized medicine has become a research hotspot in the medical field. Especially in orthopedic surgery, due to the individual differences of patients, traditional surgical plans often cannot meet the special needs of different patients, resulting in significant differences in surgical effects and recovery times. Therefore, it is of great clinical significance to use machine learning and deep learning algorithms to customize personalized orthopedic surgery plans based on information such as the individual characteristics, imaging data, historical surgical data, and physiological parameters of patients.

[0008] At present, the application of machine learning technology in medical image processing, disease diagnosis, and treatment planning has achieved remarkable results. By extracting features from medical images and using supervised learning algorithms to train personalized orthopedic surgery planning models, the accuracy and personalization of surgical plans can be effectively improved, postoperative complications can be reduced, and the quality of life and recovery speed of patients can be improved. However, there is currently a lack of a unified and efficient personalized orthopedic surgery planning method based on machine learning that can comprehensively consider the patient's overall condition, bone structure, medical imaging data, and historical surgical data to provide a comprehensive and personalized surgical plan. Summary of the Invention

[0009] To solve the above technical problems and provide a personalized orthopedic surgery planning method based on machine learning, the present technical solution solves the above problems.

[0010] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0011] A personalized orthopedic surgery planning method based on machine learning, comprising:

[0012] Collect medical imaging data, medical history data, bone structure data, and surgery-related data of a patient to generate a personalized basic data set for the patient, establish a risk prediction model, obtain a predicted surgical risk score, and determine the progress of the surgery;

[0013] Perform denoising, normalization, standardization, and data augmentation processing on the basic data set, extract features of the patient's bone structure based on deep learning algorithms, and map them into feature vectors suitable for surgical planning;

[0014] Based on the feature data of the patient, use supervised learning algorithms to train a personalized orthopedic surgery planning model, and optimize the historical patient data based on the personalized orthopedic surgery planning model to predict the best surgical plan;

[0015] Diagnose the orthopedic diseases of the patient according to the trained machine learning model, and output a personalized surgical plan, including the surgical site, surgical method, selection of surgical instruments, and surgical time.

[0016] Preferably, the collection of the patient's medical imaging data, medical history data, bone structure data, and surgery-related data to generate a personalized basic dataset for the patient specifically includes:

[0017] Collect medical imaging data from the patient's imaging examinations, including but not limited to X-ray, CT, and MRI images, convert the collected imaging data into a standard format, and unify images from different sources and types into a standardized format;

[0018] Combined with clinical expert annotations, annotate the images of each patient to clarify the lesion area and bone morphological characteristics;

[0019] Collect the patient's basic information, including gender, age, weight, and height, collect the patient's medical history and family history, and understand whether there are genetic diseases and factors that are likely to cause orthopedic problems;

[0020] Record the patient's previous surgical history and treatment plan, collect the patient's drug allergy history and medications being taken to ensure the safety of drug use during the surgery;

[0021] Obtain the density information of the patient's bones through bone density detection, evaluate the strength and vulnerability of the bone mass, which affects the fracture risk and the selection of the surgical plan. Based on the imaging data, obtain a three-dimensional data model of the patient's bones through three-dimensional reconstruction technology, and analyze the bone morphology, joint space, and fracture position characteristics;

[0022] Extract the geometric parameters of the patient's bones, including the length, angle, and curvature of the bones, and evaluate the range of joint movement through imaging;

[0023] Collect the historical surgical data of similar patients, including the type of surgery, surgical process, instruments used, and intraoperative complications;

[0024] Evaluate the risk of the surgery by assessing the patient's general condition, combining historical surgical data and a risk prediction model;

[0025] Integrate the patient's medical imaging data, medical history data, bone structure data, and surgery-related data to construct a personalized basic dataset for the patient. Based on data fusion technology, form these data into a unified structured dataset.

[0026] Preferably, the evaluation of the risk of the surgery by assessing the patient's general condition, combining historical surgical data and a risk prediction model specifically includes:

[0027] Among them, the risk prediction model formula is:

[0028]

[0029] In the formula, R is the predicted risk score, α1, α2, α n are the weighted coefficients of the features, F1, F2, F n are the physiological and clinical features, L is the total surgical risk score, R asa is the risk score obtained according to the ASA score, w asa is the weight coefficient of R asa is the risk score of the historical surgical data, w m is the weight coefficient of R m is the risk score obtained through the prediction model, and w is the weight coefficient of R; m Based on the risk score of the surgery, it is determined whether the surgery can be performed.

[0030] Preferably, the denoising, normalization, standardization, and data augmentation processing of the basic data set, extracting the features of the patient's bone structure based on the deep learning algorithm, and mapping them into feature vectors suitable for surgical planning specifically include:

[0031] Performing Gaussian blur on the image based on the Gaussian filter formula to remove high-frequency noise and performing normalization processing on the data;

[0032] Performing data augmentation processing through rotation, translation, mirror flipping, scaling, and random noise, and generating new samples by transforming the original data;

[0033] Automatically extracting high-level features of the bone structure from the medical images based on the convolutional neural network;

[0034] Mapping the features extracted from the deep learning model into feature vectors suitable for surgical planning through principal component analysis.

[0035] Preferably, the performing Gaussian blur on the image based on the Gaussian filter formula to remove high-frequency noise and performing normalization processing on the data specifically includes:

[0036] Among them, the Gaussian filter formula is:

[0037]

[0038]

[0039] In the formula, I d (x, y) is the original image data after Gaussian filter processing, G(i, j, σ) is the Gaussian kernel, σ is the standard deviation, k is the kernel size, I(x + i, y + j) is the pixel value at the coordinate (x + i, y + j) in the original image I, and i and j are the offsets relative to the center position in the convolution kernel.​

[0040] Preferably, based on the patient's characteristic data, a personalized orthopedic surgery planning model is trained using a supervised learning algorithm, and the historical patient data is optimized based on the personalized orthopedic surgery planning model to predict the best surgical plan, which specifically includes:

[0041] Based on the decision tree algorithm, a personalized orthopedic surgery planning model is established through the extracted characteristic data;

[0042] The personalized orthopedic surgery planning model is trained using the training set, and the model parameters are adjusted through Bayesian optimization;

[0043] Based on the results predicted by the model, the historical patient data is analyzed, the surgical plan of the patient is optimized, different surgical plans are simulated on the historical patient data, and the effects of different surgical plans are compared;

[0044] According to the specific characteristics and surgical needs of the patient, an optimal surgical plan is customized for each patient;

[0045] For new patients, based on the trained model, combined with the personal characteristic data of the patient, the best surgical plan is predicted. The characteristic data of the new patient is input, the prediction result is obtained using the trained model, the recommended surgical plan is output, the prediction result is combined with the recommendation system, and a personalized surgical plan report is generated through natural language processing.

[0046] Preferably, the establishment of the personalized orthopedic surgery planning model based on the decision tree algorithm through the extracted characteristic data specifically includes:

[0047] The data is divided by selecting the best feature, and the feature division is performed based on the information gain formula;

[0048] Based on the selected feature, a personalized orthopedic surgery planning model is established, where the formula of the personalized orthopedic surgery planning model is:

[0049] Y = DecisionTree(X) where X = (x1, x2, …, x n )

[0050] In the formula, Y is the prediction result, DecisionTree is the decision tree model, X is the input feature, n is the number of features, and x1, x2, …, x n are sample features.

[0051] Preferably, the division of the data by selecting the best feature and the feature division based on the information gain formula specifically include:

[0052] Among them, the information gain formula is:

[0053]

[0054] Wherein, H(S) is the entropy of the data set S, and S v is the subset obtained after partitioning according to feature A, values(A) are all possible values that feature A can take, and IG(S,A) is the information gain after partitioning the data set S by feature A.

[0055] Preferably, based on the results of model prediction, analyze the historical patient data, optimize the surgical plan of the patient, simulate different surgical plans on the historical patient data, and compare the effects of different surgical plans, which specifically include:

[0056] Input the feature data of the patient, simulate the expected effects of different surgical plans, set different surgical plans on the historical data, and make predictions;

[0057] Select the key indicators for evaluating the surgical effect, including the postoperative complication rate, postoperative recovery time, hospitalization time, mortality rate, and patient satisfaction;

[0058] After simulating different surgical plans through the model, compare the effects of each plan, and find the optimal surgical plan;

[0059] Based on the prediction results output by the model, combined with the personalized characteristics of the patient, propose a plan with the best effect, and verify the optimized surgical plan through clinical feedback and actual surgical results;

[0060] Visualize the effects and comparison results of different surgical plans, and use charts to show the effect differences of different plans.

[0061] Preferably, after simulating different surgical plans through the model, comparing the effects of each plan, and finding the optimal surgical plan specifically includes:

[0062] Use the multi-index evaluation method to comprehensively consider multiple indicators to compare surgical plans. Among them, the formula of the multi-index evaluation method is:

[0063]

[0064] Wherein, S i is the comprehensive score of the i-th surgical plan, w k is the weight of the k-th indicator, x ik is the score of the i-th surgical plan on the k-th indicator, and n is the number of evaluation indicators.

[0065] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0066] The present invention proposes to combine machine learning with personalized orthopedic surgical planning. By comprehensively considering the patient's medical imaging data, medical history, bone structure data, and surgical-related data, a personalized surgical plan is customized for each patient to optimize the surgical effect and surgical risk assessment. Through deep learning and data augmentation techniques, bone features can be extracted more precisely. Combining historical data and risk prediction models, the most suitable surgical plan is provided for the patient, significantly improving the surgical treatment effect and the patient's quality of life, and having broad clinical application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 It is a flow chart framework diagram of the steps of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0068] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments in the following description are only examples, and other obvious variations can be thought of by those skilled in the art.

[0069] Refer to Figure 1 As shown, the personalized orthopedic surgical planning method based on machine learning includes:

[0070] Step 1:

[0071] Collect medical imaging data from the patient's imaging examinations, including but not limited to X-ray, CT, and MRI images. Convert the collected image data into a standardized format to unify images from different sources and types.

[0072] Combined with clinical expert annotations, annotate the images of each patient to clarify the lesion area and bone morphological features.

[0073] Collect the patient's basic information, including gender, age, weight, and height. Collect the patient's medical history and family history to understand whether there are genetic diseases and factors prone to orthopedic problems.

[0074] Record the patient's previous surgical history and treatment plan, collect the patient's drug allergy history and currently taking medications to ensure the safety of drug use during the operation.

[0075] Obtain the density information of the patient's bones through bone density detection to evaluate the strength and vulnerability of the bone mass, which affects the fracture risk and the choice of surgical plan. Based on the imaging data, obtain a three-dimensional data model of the patient's bones through three-dimensional reconstruction technology, and analyze the bone morphology, joint space, and fracture position features.

[0076] Extract the geometric parameters of the patient's bones, including the length, angle, and curvature of the bones, and evaluate the joint range of motion through imaging.

[0077] Collect the historical surgical data of similar patients, including the type of surgery, the surgical process, the instruments used, and intraoperative complications;

[0078] Evaluate the risk of surgery by assessing the patient's general condition, combining historical surgical data and a risk prediction model;

[0079] Among them, the risk prediction model formula is:

[0080]

[0081] In the formula, R is the predicted risk score, α1α2α n are the weighted coefficients of the features, F1F1F n are the physiological and clinical features, L is the total surgical risk score, R asa is the risk score obtained according to the ASA score, w asa is for R asa the weight coefficient of, R m is the risk score of the historical surgical data, w m is for R m the weight coefficient of, R is the risk score obtained through the prediction model, and w is the weight coefficient of R;

[0082] Based on the risk score of the surgery, determine whether the surgery can be performed;

[0083] Integrate the patient's medical image data, medical history data, bone structure data, and surgery-related data to construct a personalized basic dataset for the patient, and based on data fusion technology, form these data into a unified structured dataset.

[0084] Step two:

[0085] Perform Gaussian blur on the image based on the Gaussian filtering formula to remove high-frequency noise and normalize the data;

[0086] Among them, the Gaussian filtering formula is:

[0087]

[0088] In the formula, I d (x,y) is the original image data after Gaussian filtering, G(i,j,σ) is the Gaussian kernel, σ is the standard deviation, k is the kernel size, I(x+i,t+j) is the pixel value at the coordinate (x+i,y+j) in the original image I, and i and j are the offsets relative to the center position in the convolution kernel;

[0089] Perform data augmentation processing on the data through rotation, translation, mirror flipping, scaling, and random noise, and generate new samples by transforming the original data;

[0090] Automatically extract high-level features of bone structures from medical images based on convolutional neural networks;

[0091] Map the features extracted from the deep learning model into feature vectors suitable for surgical planning through principal component analysis.

[0092] Step Three:

[0093] Based on the decision tree algorithm, establish a personalized orthopedic surgical planning model through the extracted feature data;

[0094] Divide the data by selecting the best features, and perform feature division based on the information gain formula;

[0095] Among them, the information gain formula is:

[0096]

[0097] In the formula, H(S) is the entropy of the dataset S, S v is the subset obtained after division according to the feature A, values(A) are all possible values that the feature A can take, and IG(S,A) is the information gain after dividing the dataset S by the feature A;

[0098] Based on the selected features, establish a personalized orthopedic surgical planning model. Among them, the formula for the personalized orthopedic surgical planning model is:

[0099] Y = DecisionTree(X) where X = (x1,x2,…,x n )

[0100] In the formula, Y is the prediction result, DecisionTree is the decision tree model, X is the input feature, n is the number of features, and x1,x2,…,x n are the sample features;

[0101] Use the training set to train the personalized orthopedic surgical planning model, and adjust the model parameters through Bayesian optimization;

[0102] Based on the results predicted by the model, analyze the historical patient data, optimize the surgical planning of the patients, simulate different surgical plans on the historical patient data, and compare the effects of different surgical plans;

[0103] Input the feature data of the patient, simulate the expected effects of different surgical plans, set different surgical plans on the historical data, and make predictions;

[0104] Select the key indicators for evaluating the surgical effect, including the postoperative complication rate, postoperative recovery time, hospitalization time, mortality rate, and patient satisfaction;

[0105] After simulating different surgical plans through the model, compare the effects of each plan to find the optimal surgical plan;

[0106] Use a multi-index evaluation method to comprehensively consider multiple indicators to compare surgical plans. Among them, the formula for the multi-index evaluation method is:

[0107]

[0108] In the formula, S i is the comprehensive score of the i-th surgical plan, w k is the weight of the k-th indicator, x ik is the score of the i-th surgical plan on the k-th indicator, and n is the number of evaluation indicators;

[0109] Based on the prediction results output by the model and combined with the personalized characteristics of the patient, propose a plan with the best effect, and verify the optimized surgical plan through clinical feedback and actual surgical results;

[0110] Visualize the effects and comparison results of different surgical plans, and use charts to show the effect differences of different plans;

[0111] According to the specific characteristics and surgical needs of the patient, customize the optimal surgical plan for each patient;

[0112] For new patients, based on the trained model and combined with the personal characteristic data of the patient, predict the best surgical plan. Input the characteristic data of the new patient, use the trained model to obtain the prediction result, output the recommended surgical plan, combine the prediction result with the recommendation system, and generate a personalized surgical plan report through natural language processing.

[0113] Step Four:

[0114] According to the trained machine learning model, diagnose the orthopedic diseases of the patient and output a personalized surgical plan, including the surgical site, surgical method, selection of surgical instruments, and surgical time.

[0115] In summary, the advantages of the present invention are as follows:

[0116] By collecting information such as the medical image data, medical history data, and bone structure data of the patient, based on machine learning and deep learning technologies, it is possible to customize the most suitable surgical plan for each patient. This method not only considers the specific condition of the patient but also can optimize by integrating historical surgical data, improving the treatment effect;

[0117] Accurate surgical risk assessment: By establishing a risk prediction model and comprehensively considering physiological and clinical characteristics, ASA score, historical surgical data, etc., the surgical risk of patients can be accurately evaluated. This comprehensive assessment method can effectively help doctors determine whether a surgery is suitable and avoid unnecessary surgical risks;

[0118] Based on the trained machine learning model, optimize the historical patient data, predict the best surgical plan, and provide personalized surgical planning according to the individual characteristics of the patient. By simulating the effects of different surgical plans, the surgical decision of the patient can be optimized, and a surgical plan with a short postoperative recovery time, few complications, and high satisfaction can be selected;

[0119] By using deep learning algorithms to extract high-level features of the patient's bone structure and combining data augmentation techniques, more potential and useful features can be extracted from medical images, providing a more accurate basis for surgical planning. In addition, techniques such as Gaussian filtering and principal component analysis can further improve the effects of image processing and data dimensionality reduction, thereby improving the accuracy of surgical planning;

[0120] By using a multi-index evaluation method to comprehensively compare the effects of different surgical plans, the plans can be evaluated from multiple dimensions, and the effect differences of different surgical plans can be displayed through visual charts, facilitating doctors and patients to make more informed decisions;

[0121] By using decision tree algorithms and combining information gain for feature selection, an accurate personalized orthopedic surgical planning model can be established. Based on this model, the system can simulate historical patient data and predict the effects of different surgical plans, so as to customize the best treatment plan for each patient;

[0122] This method uses machine learning and artificial intelligence technologies to assist doctors in accurate diagnosis and treatment plan planning, which can effectively improve clinical efficiency, reduce human errors, optimize the surgical process, reduce surgical complications, and improve the overall treatment effect and recovery of patients.

[0123] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art of this industry should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification is only the principle of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A personalized orthopedic surgery planning method based on machine learning, characterized in that Including: Collecting the patient's medical imaging data, medical history data, bone structure data, and surgery-related data, generating a personalized basic dataset for the patient, establishing a risk prediction model, obtaining a predicted surgical risk score, and determining whether the surgery can be performed; Performing denoising, normalization, standardization, and data augmentation on the basic dataset, extracting the features of the patient's bone structure based on deep learning algorithms, and mapping them into feature vectors suitable for surgical planning; Training a personalized orthopedic surgery planning model using supervised learning algorithms based on the patient's feature data, optimizing the historical patient data based on the personalized orthopedic surgery planning model, and predicting the best surgical plan; Diagnosing the patient's orthopedic disease according to the trained machine learning model and outputting a personalized surgical plan, including the surgical site, surgical method, selection of surgical instruments, and surgical time.

2. The personalized orthopedic surgery planning method based on machine learning according to claim 1, wherein The collecting of the patient's medical imaging data, medical history data, bone structure data, and surgery-related data to generate a personalized basic dataset for the patient specifically includes: Collecting medical imaging data from the patient's imaging examinations, including but not limited to X-ray, CT, and MRI images, performing format conversion on the collected imaging data, and unifying images from different sources and types into a standardized format; Combined with clinical expert annotations, annotating each patient's image to clarify the lesion area and bone morphological features; Collecting the patient's basic information, including gender, age, weight, and height, collecting the patient's disease history and family history, and understanding whether there are genetic diseases and factors prone to orthopedic problems; Recording the patient's previous surgical history and treatment plan, collecting the patient's drug allergy history and medications being taken to ensure the safety of drug use during the surgery; Obtaining the density information of the patient's bones through bone density detection, evaluating the strength and vulnerability of the bone mass, which affects the fracture risk and the selection of the surgical plan. Based on the imaging data, obtaining a three-dimensional data model of the patient's bones through three-dimensional reconstruction technology, and analyzing the bone morphology, joint space, and fracture position features; Extracting the geometric parameters of the patient's bones, including the length, angle, and curvature of the bones, and evaluating the range of joint movement through imaging; Collecting historical surgical data of similar patients, including the type of surgery, surgical process, instruments used, and intraoperative complications; Evaluating the risk of surgery by assessing the patient's overall condition, combining historical surgical data and the risk prediction model; Integrating the patient's medical imaging data, medical history data, bone structure data, and surgery-related data to construct a personalized basic dataset for the patient, and forming a unified structured dataset for these data based on data fusion technology.

3. The personalized orthopedic surgery planning method based on machine learning according to claim 2, wherein, The evaluating the risk of surgery by assessing the patient's overall condition, combining historical surgical data and the risk prediction model specifically includes: Among them, the risk prediction model formula is: where R is the predicted risk score, α1, α2, α n are the weighting coefficients of the features, F1, F2, F n are the physiological and clinical features, L is the total surgical risk score, R asa is the risk score obtained from the ASA score, w asa is the weighting coefficient of R ase ; R m is the risk score of the historical surgical data, w m is the weighting coefficient of R m ; R is the risk score obtained from the prediction model, and w is the weighting coefficient of R; Based on the risk score of the surgery, judging whether the surgery can be performed.

4. The personalized orthopedic surgery planning method based on machine learning according to claim 3, wherein, The performing of denoising, normalization, standardization, and data augmentation on the basic dataset, extracting the features of the patient's bone structure based on deep learning algorithms, and mapping them into feature vectors suitable for surgical planning specifically includes: Perform Gaussian blur on the image based on the Gaussian filter formula to remove high-frequency noise and normalize the data; Perform data augmentation processing on the data through rotation, translation, mirror flipping, scaling, and random noise, and generate new samples by transforming the original data; Automatically extract high-level features of bone structures from medical images based on a convolutional neural network; Map the features extracted from the deep learning model to feature vectors suitable for surgical planning through principal component analysis.

5. The personalized orthopedic surgery planning method based on machine learning according to claim 4, wherein, The specific process of performing Gaussian blur on the image based on the Gaussian filter formula to remove high-frequency noise and normalize the data includes: where the Gaussian filter formula is: where I d (x, y) is the original image data after Gaussian filtering, G(i, j, σ) is the Gaussian kernel, σ is the standard deviation, k is the kernel size, I(x + i, y + j) is the pixel value at the coordinate (x + i, y + j) in the original image I, and i and j are the offsets relative to the center position in the convolution kernel.

6. The personalized orthopedic surgery planning method based on machine learning according to claim 5, wherein The specific process of training a personalized orthopedic surgical planning model based on the patient's feature data using a supervised learning algorithm, optimizing the historical patient data based on the personalized orthopedic surgical planning model, and predicting the best surgical plan includes: Based on the decision tree algorithm, establish a personalized orthopedic surgical planning model through the extracted feature data; Use the training set to train the personalized orthopedic surgical planning model and adjust the model parameters through Bayesian optimization; Based on the results predicted by the model, analyze the historical patient data, optimize the patient's surgical plan, simulate different surgical plans on the historical patient data, and compare the effects of different surgical plans; Tailor the optimal surgical plan for each patient according to the patient's specific characteristics and surgical needs; For new patients, based on the trained model, combined with the patient's personal feature data, predict the best surgical plan, input the feature data of the new patient, use the trained model to obtain the prediction result, output the recommended surgical plan, combine the prediction result with the recommendation system, and generate a personalized surgical plan report through natural language processing.

7. The personalized orthopedic surgery planning method based on machine learning according to claim 6, wherein, The specific process of establishing a personalized orthopedic surgical planning model based on the decision tree algorithm through the extracted feature data includes: Divide the data by selecting the best features and perform feature division based on the information gain formula; Based on the selected features, establish a personalized orthopedic surgical planning model, where the personalized orthopedic surgical planning model formula is: Y = DecisionTree(X) where X = (x1,x2,…,x n ) Where Y is the prediction result, DecisionTree is the decision tree model, X is the input feature, n is the number of features, and x1, x2, …, x n are sample features.

8. The personalized orthopedic surgery planning method based on machine learning according to claim 7, wherein The specific process of dividing the data by selecting the best features and performing feature division based on the information gain formula includes: where the information gain formula is: where \(H(S)\) is the entropy of the data set \(S\), and \(S\) v is the subset obtained by partitioning according to feature \(A\), \(values(A)\) are all possible values that feature \(A\) can take, and \(IG(S, A)\) is the information gain obtained by partitioning the data set \(S\) by feature \(A\).

9. The personalized orthopedic surgery planning method based on machine learning according to claim 8, characterized in that The specific process of analyzing the historical patient data based on the results predicted by the model, optimizing the patient's surgical plan, simulating different surgical plans on the historical patient data, and comparing the effects of different surgical plans includes: Input the patient's feature data, simulate the expected effects of different surgical plans, set different surgical plans on the historical data, and make predictions; Select the key indicators for evaluating the surgical effect, including the postoperative complication rate, postoperative recovery time, hospital stay, mortality rate, and patient satisfaction; After simulating different surgical plans through the model, compare the effects of each plan to find the optimal surgical plan; Based on the prediction results output by the model, combined with the patient's personalized characteristics, propose the plan with the best effect, and verify the optimized surgical plan through clinical feedback and actual surgical results; Visualize the effects and comparison results of different surgical plans and use charts to show the effect differences of different plans.

10. The personalized orthopedic surgery planning method based on machine learning according to claim 9, wherein After simulating different surgical plans through the model, comparing the effects of each plan, and finding the optimal surgical plan specifically includes: Using a multi-index evaluation method to comprehensively consider multiple indicators to compare surgical plans. Among them, the formula for the multi-index evaluation method is: Where, S i is the comprehensive score of the i-th surgical plan, w k is the weight of the k-th index, x ik is the score of the i-th surgical plan on the k-th index, and n is the number of evaluation indicators.

Images