Myocardial disease auxiliary diagnosis system based on artificial intelligence
By integrating history data and disease sensitive factors in the myocardial disease assisted diagnosis system, dynamically adjusting model parameters, identifying and extracting myocardial disease-related features, the problem of staticity and inaccurate feature extraction of diagnostic models in the prior art is solved, and the problem of inaccurate feature extraction of diagnostic models in the prior art is achieved. Higher diagnostic accuracy and personalized treatment.
Patent Information
- Application Number
- CN202411155149.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-22
- Publication Date
- 2024-11-29
AI Technical Summary
The existing myocardial disease assisted diagnosis system fails to fully integrate the patient's comprehensive medical history data, real-time disease characteristics and disease sensitivity factors, resulting in a lack of dynamicity and flexibility in the diagnostic model, reducing the accuracy and timeliness of diagnosis, and insufficient feature extraction , easily disturbed by noise data, increasing the computing burden.
Design an artificial intelligence-based myocardial disease assisted diagnosis system, including data acquisition and preprocessing module, feature extraction module, auxiliary diagnosis module, condition evaluation module and personalized diagnosis and treatment module, by integrating patient's historical history data, condition characteristic data and disease Sensitive factors dynamically adjust model parameters, identify and extract the characteristics of the disorders that are highly related to myocardial disease, build the symptoms sensitive factors and diagnostic evolution model, adjust the model weights in real time, and conduct highly diagnosed and personalized treatment suggestions for myocardial disease.
It improves the accuracy and real-time diagnosis, ensures that the model focuses on the most relevant data points, reduces the data burden of subsequent processing, provides a personalized treatment plan, and enhances the flexibility and treatment effect of diagnosis.
Smart Images

Figure CN119049690A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of artificial intelligence technology, and specifically refers to an artificial intelligence-based myocardial disease auxiliary diagnosis system. Background Art
[0002] With the aging of the population and changes in lifestyle, myocardial disease has become a major health issue worldwide. Traditional methods for diagnosing myocardial disease rely on the doctor's experience and the patient's clinical manifestations, and have problems such as strong subjectivity and high misdiagnosis rate.
[0003] However, the existing auxiliary diagnosis system for myocardial diseases still has certain defects. The existing auxiliary diagnosis system for myocardial diseases fails to fully integrate the patient's comprehensive medical history data, real-time symptom characteristics and symptom sensitivity factors, resulting in the lack of sufficient dynamics and flexibility in the diagnosis model. The static model parameters cannot effectively respond to the fluctuations in the patient's condition, reducing the accuracy of the diagnosis. The medical history data is not updated in a timely manner or is not taken into account in the model, so that the diagnosis process cannot fully reflect the patient's latest health status, affecting the timeliness and accuracy of the diagnosis. The feature extraction may not be accurate enough and cannot effectively distinguish between key symptom characteristics that are highly related to myocardial disease and other irrelevant information, resulting in low model training efficiency. The diagnostic results are easily disturbed by noise data. There is a lack of evaluation and sorting mechanism for the extracted features, resulting in the need to process a large amount of redundant data in the subsequent processing stage, which increases the computational burden and response time. Therefore, an auxiliary diagnosis system for myocardial diseases based on artificial intelligence is proposed. Summary of the invention
[0004] The purpose of the present invention is to provide an artificial intelligence-based myocardial disease auxiliary diagnosis system to solve the problems raised in the above-mentioned background technology.
[0005] To achieve the above-mentioned object, the present invention provides the following technical solutions: an artificial intelligence-based myocardial disease auxiliary diagnosis system, comprising a data acquisition preprocessing module, a feature extraction module, an auxiliary diagnosis module, a symptom assessment module, a symptom analysis module and a personalized diagnosis and treatment module;
[0006] The feature extraction module identifies and extracts the symptom features that are highly related to myocardial diseases through data acquisition preprocessed clinical data, and evaluates and ranks the extracted symptom features;
[0007] The auxiliary diagnosis module integrates the patient's historical medical history data, disease characteristic data and disease sensitivity factors, dynamically adjusts the model parameters, and performs a high-level diagnosis of myocardial diseases;
[0008] The specific operation method of the auxiliary diagnosis module includes the following steps:
[0009] A1: Constructing disease sensitivity factors: According to the disease feature results evaluated by the feature extraction module, assign corresponding weights to each disease feature, evaluate the impact of disease features on the diagnosis results through sensitivity analysis, and construct disease sensitivity factors by combining disease feature weights with sensitivity analysis results;
[0010] A2: Historical medical history analysis: Based on historical medical history data, the fluctuation of symptom characteristics is evaluated by calculating the standard deviation of symptom characteristics;
[0011] A3: Dynamically adjust the disease sensitivity factor: Update the disease sensitivity factor according to the fluctuation of disease characteristics;
[0012] A4: Construction of disease evolution matrix: Construct a correlation matrix between features based on historical medical history data. Based on the correlation matrix, a diagnostic evolution model is constructed through a Markov chain to predict the changing trend of features.
[0013] A5: Dynamic adaptive adjustment: Dynamic adaptation is performed based on the disease characteristic vector, the adjusted sensitivity factor and the disease evolution matrix;
[0014] A6: Real-time weight adjustment: The weight is adjusted dynamically in real time based on the patient data information fed back.
[0015] Wherein, A2 evaluates the volatility of the symptom characteristics by calculating the standard deviation of the symptom characteristics based on the historical medical history data, and the implementation formula is:
[0016]
[0017] In formula (1), σ j Table 1 shows the standard deviation of the jth feature, N represents the number of samples, x jj represents the jth eigenvalue of the i-th sample, represents the average value of the j-th feature.
[0018] Among them, A3 updates the disease sensitivity factor according to the fluctuation of the disease characteristics. Let the disease sensitivity factor be F, and combine formula (1) to obtain the updated disease sensitivity factor, and the implementation formula is:
[0019] F′=F+α(σ j -σ j ′)(2),
[0020] In formula (2), F′ represents the updated disease sensitivity factor, α represents the adjustment coefficient, and σ j and σ j ′ represents the standard deviation before and after adjustment.
[0021] Among them, A4 constructs a correlation matrix between features based on historical medical history data, and the implementation formula is:
[0022]
[0023] In the formula, R jk represents the correlation coefficient between the jth feature and the kth feature, x ij and x ik denote the jth and kth eigenvalues of the ith sample, respectively. and Represent the average values of the j-th and k-th features respectively, and N represents the number of samples;
[0024] Based on the correlation matrix, the diagnostic evolution model is constructed through the Markov chain, and the implementation formula is:
[0025]
[0026] In the formula, P jk represents the transition probability from state j to state k, and They represent indicator functions respectively, when the condition x ij =j and x i ( k+1 )=k holds, then return 1; otherwise return 0.
[0027] Wherein, the A5 is dynamically adapted according to the disease characteristic vector, the adjusted sensitivity factor and the disease evolution matrix, and the implementation formula is:
[0028] y=f(x,P,F′),
[0029] In the formula, f represents the diagnostic model function, p represents the disease evolution matrix, F′ represents the adjusted disease sensitivity factor, and the probability of myocardial disease is predicted by the formula, and x represents the disease important feature matrix;
[0030] The eigenvector is dynamically adjusted according to the disease evolution matrix, and the implementation formula is:
[0031] x′=x+Δx,
[0032] In the formula, x′ represents the adjusted disease characteristic vector, and Δx represents the adjustment amount of the disease characteristic vector.
[0033] The auxiliary diagnosis module integrates the patient's medical history data, disease characteristic data and disease sensitivity factors, dynamically adjusts the model parameters, and makes a high-level diagnosis of myocardial diseases. The disease sensitivity factors are updated according to the fluctuations of the patient's medical history data, ensuring the flexibility and accuracy of the diagnosis process. By dynamically adjusting the model parameters, the actual situation of the patient can be more accurately matched, improving the real-time nature of the diagnosis results.
[0034] The feature extraction module collects preprocessed clinical data, identifies and extracts symptom features that are highly correlated with myocardial disease, and evaluates the extracted symptom features; the importance of the symptom features of the clinical data is evaluated to obtain features that are highly correlated with the symptom, and the implementation formula is:
[0035]
[0036] In the formula, I j represents the disease importance score of the jth feature, N represents the number of samples, y i and They represent the true value and predicted value of the i-th sample respectively. The feature with high disease importance score is recorded as X′. X′ represents the important feature matrix of the disease: X′=[x1x2...x J ], x J According to I j The important features of the disease are selected, J represents the number of important features of the disease, and the formula is realized by reducing the dimension of the important feature matrix of the disease:
[0037] x=W T X′,
[0038] In the formula, x represents the important feature matrix of the disease after dimensionality reduction, W T Represents the transpose of the weight matrix.
[0039] The symptom assessment module scores the severity of the symptom according to the symptom prediction results of the auxiliary diagnosis module, and sets the threshold between 0 and 10, with a score above 7 indicating severe symptoms and a score below 4 indicating mild symptoms.
[0040] The symptom assessment module scores the severity of the symptom based on the results of the auxiliary diagnosis module, and divides the severity of the symptom into different levels by setting thresholds, so that doctors can quickly understand the patient's condition
[0041] The symptom analysis module performs analysis based on the evaluation results of the symptom assessment module, and provides detailed diagnostic basis and diagnostic suggestions based on the analysis results.
[0042] Among them, the personalized diagnosis and treatment module provides personalized diagnosis and treatment based on the diagnosis suggestions and diagnosis results provided by the symptom analysis module.
[0043] Compared with the prior art, the present invention has the following beneficial effects:
[0044] 1. In order to solve the problem of traditional diagnostic accuracy and improve the accuracy of the diagnostic process, the present invention integrates the patient's medical history data, symptom characteristic data and symptom sensitivity factors through the auxiliary diagnosis module, dynamically adjusts the model parameters, and performs a high-level diagnosis of myocardial diseases. The symptom sensitivity factors are updated according to the fluctuations of the patient's medical history data, thereby ensuring the flexibility and accuracy of the diagnostic process. By dynamically adjusting the model parameters, the actual situation of the patient can be more accurately matched, and the real-time nature of the diagnostic results can be improved;
[0045] 2. The present invention uses the symptom assessment module to score the severity of the symptom based on the results of the auxiliary diagnosis module, and divides the severity of the symptom into different levels by setting a threshold, so that the doctor can quickly understand the patient's condition;
[0046] 3. The present invention accurately identifies and extracts disease features that are highly relevant to myocardial diseases through a feature extraction module, which can ensure that the model focuses on the most relevant data points, thereby improving the accuracy of diagnosis. The extracted disease features are evaluated and sorted, which helps to remove irrelevant information and reduce the data processing burden in the subsequent processing stage, thereby improving the overall processing efficiency;
[0047] 4. The present invention provides personalized treatment suggestions based on the results of symptom analysis through a personalized diagnosis and treatment module, which can treat each patient according to their specific conditions and improve the treatment effect. BRIEF DESCRIPTION OF THE DRAWINGS
[0048] Figure 1 This is a schematic diagram of the structure of an artificial intelligence-based myocardial disease auxiliary diagnosis system of the present invention;
[0049] Figure 2 This is an operation flow chart of an artificial intelligence-based myocardial disease auxiliary diagnosis system of the present invention;
[0050] Figure 3 This is an operation flow chart of an auxiliary diagnosis module of an artificial intelligence-based myocardial disease auxiliary diagnosis system of the present invention. DETAILED DESCRIPTION
[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0052] Example
[0053] See also Figure 1-Figure 3As shown, the present invention provides a technical solution: including a data acquisition preprocessing module, a feature extraction module, an auxiliary diagnosis module, a symptom assessment module, a symptom analysis module and a personalized diagnosis and treatment module;
[0054] The feature extraction module identifies and extracts the symptom features that are highly related to myocardial diseases through data acquisition preprocessed clinical data, and evaluates and ranks the extracted symptom features;
[0055] The auxiliary diagnosis module integrates the patient's historical medical history data, disease characteristic data and disease sensitivity factors, dynamically adjusts the model parameters, and performs a high-level diagnosis of myocardial diseases;
[0056] The specific operation method of the auxiliary diagnosis module includes the following steps:
[0057] A1: Constructing disease sensitivity factors: According to the disease feature results evaluated by the feature extraction module, assign corresponding weights to each disease feature, evaluate the impact of disease features on the diagnosis results through sensitivity analysis, and construct disease sensitivity factors by combining disease feature weights with sensitivity analysis results;
[0058] A2: Historical medical history analysis: Based on historical medical history data, the fluctuation of symptom characteristics is evaluated by calculating the standard deviation of symptom characteristics;
[0059] A3: Dynamically adjust the disease sensitivity factor: Update the disease sensitivity factor according to the fluctuation of disease characteristics;
[0060] A4: Construction of disease evolution matrix: Construct a correlation matrix between features based on historical medical history data. Based on the correlation matrix, a diagnostic evolution model is constructed through a Markov chain to predict the changing trend of features.
[0061] A5: Dynamic adaptive adjustment: Dynamic adaptation is performed based on the disease characteristic vector, the adjusted sensitivity factor and the disease evolution matrix;
[0062] A6: Real-time weight adjustment: The weight is adjusted dynamically in real time based on the patient data information fed back.
[0063] Wherein, A2 evaluates the volatility of the symptom characteristics by calculating the standard deviation of the symptom characteristics based on the historical medical history data, and the implementation formula is:
[0064]
[0065] In formula (1), σ j Table 1 shows the standard deviation of the jth feature, N represents the number of samples, x jj represents the jth eigenvalue of the i-th sample, represents the average value of the j-th feature.
[0066] Among them, A3 updates the disease sensitivity factor according to the fluctuation of the disease characteristics. Let the disease sensitivity factor be F, and combine formula (1) to obtain the updated disease sensitivity factor, and the implementation formula is:
[0067] F′=F+α(σ j -σ j ′)(2),
[0068] In formula (2), F′ represents the updated disease sensitivity factor, α represents the adjustment coefficient, and σ j and σ j ′ represents the standard deviation before and after adjustment.
[0069] Among them, A4 constructs a correlation matrix between features based on historical medical history data, and the implementation formula is:
[0070]
[0071] In the formula, R jk represents the correlation coefficient between the jth feature and the kth feature, x ij and x ik denote the jth and kth eigenvalues of the ith sample, respectively. and Represent the average values of the j-th and k-th features respectively, and N represents the number of samples;
[0072] Based on the correlation matrix, the diagnostic evolution model is constructed through the Markov chain, and the implementation formula is:
[0073]
[0074] In the formula, P jk represents the transition probability from state j to state k, and They represent indicator functions respectively, when the condition x ij =j and x i ( k+1 )=k holds, then return 1; otherwise return 0.
[0075] Wherein, the A5 is dynamically adapted according to the disease characteristic vector, the adjusted sensitivity factor and the disease evolution matrix, and the implementation formula is:
[0076] y=f(x,P,F′),
[0077] In the formula, f represents the diagnostic model function, p represents the disease evolution matrix, F′ represents the adjusted disease sensitivity factor, and the probability of myocardial disease is predicted by the formula, and x represents the disease important feature matrix;
[0078] The eigenvector is dynamically adjusted according to the disease evolution matrix, and the implementation formula is:
[0079] x′=x+Δx,
[0080] In the formula, x′ represents the adjusted disease characteristic vector, and Δx represents the adjustment amount of the disease characteristic vector.
[0081] The auxiliary diagnosis module integrates the patient's medical history data, disease characteristic data and disease sensitivity factors, dynamically adjusts the model parameters, and makes a high-level diagnosis of myocardial diseases. The disease sensitivity factors are updated according to the fluctuations of the patient's medical history data, ensuring the flexibility and accuracy of the diagnosis process. By dynamically adjusting the model parameters, the actual situation of the patient can be more accurately matched, improving the real-time nature of the diagnosis results.
[0082] The feature extraction module collects preprocessed clinical data, identifies and extracts symptom features that are highly correlated with myocardial disease, and evaluates the extracted symptom features; the importance of the symptom features of the clinical data is evaluated to obtain features that are highly correlated with the symptom, and the implementation formula is:
[0083]
[0084] In the formula, I j represents the disease importance score of the jth feature, N represents the number of samples, y i and They represent the true value and predicted value of the i-th sample respectively. The feature with high disease importance score is recorded as X′. X′ represents the important feature matrix of the disease: X′=[x1x2...x J ], x J According to I j The important features of the disease are selected, J represents the number of important features of the disease, and the formula is realized by reducing the dimension of the important feature matrix of the disease:
[0085] x=W T X′,
[0086] In the formula, x represents the important feature matrix of the disease after dimensionality reduction, W T Represents the transpose of the weight matrix.
[0087] The symptom assessment module scores the severity of the symptom according to the symptom prediction results of the auxiliary diagnosis module, and sets the threshold between 0 and 10, with a score above 7 indicating severe symptoms and a score below 4 indicating mild symptoms.
[0088] The symptom assessment module scores the severity of the symptom based on the results of the auxiliary diagnosis module, and divides the severity of the symptom into different levels by setting thresholds, so that doctors can quickly understand the patient's condition
[0089] The symptom analysis module performs analysis based on the evaluation results of the symptom assessment module, and provides detailed diagnostic basis and diagnostic suggestions based on the analysis results.
[0090] Among them, the personalized diagnosis and treatment module provides personalized diagnosis and treatment based on the diagnosis suggestions and diagnosis results provided by the symptom analysis module.
[0091] In this example, specifically: A2, based on historical medical data, evaluates the volatility of symptom characteristics by calculating the standard deviation of symptom characteristics, and the implementation formula is:
[0092]
[0093] In formula (1), σ j Table 1 shows the standard deviation of the jth feature, N represents the number of samples, x jj represents the jth eigenvalue of the i-th sample, represents the average value of the j-th feature.
[0094] In this example, specifically: A3 updates the disease sensitivity factor according to the fluctuation of the disease characteristics. Assuming the disease sensitivity factor is F, the updated disease sensitivity factor is obtained by combining formula (1), and the implementation formula is:
[0095] F′=F+α(σ j -σ j ′)(2),
[0096] In formula (2), F′ represents the updated disease sensitivity factor, α represents the adjustment coefficient, and σ j and σ j ′ represents the standard deviation before and after adjustment.
[0097] In this example, specifically: A4 constructs a correlation matrix between features based on historical medical history data, and the implementation formula is:
[0098]
[0099] In the formula, R jk represents the correlation coefficient between the jth feature and the kth feature, x ij and x ik denote the jth and kth eigenvalues of the ith sample, respectively. and Represent the average values of the j-th and k-th features respectively, and N represents the number of samples;
[0100] Based on the correlation matrix, the diagnostic evolution model is constructed through the Markov chain, and the implementation formula is:
[0101]
[0102] In the formula, P jk represents the transition probability from state j to state k, and They represent indicator functions respectively, when the condition x ij =j and x i(k+1) =k holds true, returns 1; otherwise returns 0.
[0103] In this example, specifically: A5 is dynamically adapted according to the disease characteristic vector, the adjusted sensitivity factor and the disease evolution matrix, and the implementation formula is:
[0104] y=f(x,P,F′),
[0105] In the formula, f represents the diagnostic model function, p represents the disease evolution matrix, F′ represents the adjusted disease sensitivity factor, and the probability of myocardial disease is predicted by the formula, and x represents the disease important feature matrix;
[0106] The eigenvector is dynamically adjusted according to the disease evolution matrix, and the implementation formula is:
[0107] x′=x+Δx,
[0108] In the formula, x′ represents the adjusted disease characteristic vector, and Δx represents the adjustment amount of the disease characteristic vector.
[0109] Working principle: The data acquisition and preprocessing module collects and preprocesses the patient's clinical data, including electrocardiogram, echocardiogram, etc. The feature extraction module conducts in-depth analysis of the preprocessed clinical data, identifies and extracts disease features that are highly related to myocardial diseases, evaluates the extracted disease features, and ranks them according to their importance to the diagnosis results;
[0110] The patient's historical medical history data, disease characteristic data and disease sensitivity factors are integrated through the diagnosis auxiliary module. The model parameters are dynamically adjusted to perform a high-level diagnosis of myocardial diseases. According to the evaluation results of the disease characteristics, weights are assigned to each feature, and the disease sensitivity factor is constructed in combination with sensitivity analysis. The standard deviation of the disease characteristics is calculated, and its volatility is evaluated. The disease sensitivity factor is updated according to the fluctuation of the disease characteristics. The correlation matrix between the features is constructed based on the historical medical history data, and the diagnosis evolution model is constructed through the Markov chain. Dynamic adaptation is performed according to the disease characteristic vector, the adjusted sensitivity factor and the disease evolution matrix to optimize the diagnosis model: According to the feedback of the patient data information, the model weight is adjusted in real time to maintain the accuracy and adaptability of the diagnosis model. According to the disease prediction results of the auxiliary diagnosis module, the severity of the disease is scored. A score above 7 indicates severe and a score below 4 indicates mild symptoms. According to the results of the disease evaluation module, an in-depth analysis is conducted to provide detailed diagnostic basis and diagnostic suggestions. According to the diagnostic suggestions and diagnostic results provided by the disease analysis module, personalized diagnosis and treatment plans are provided for patients.
[0111] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
[0112] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.
Claims
1. An artificial intelligence-based myocardial disease auxiliary diagnosis system, characterized by: It includes data collection and preprocessing module, feature extraction module, auxiliary diagnosis module, symptom assessment module, symptom analysis module and personalized diagnosis and treatment module; The feature extraction module identifies and extracts the symptom features that are highly related to myocardial diseases through data acquisition preprocessed clinical data, and evaluates and ranks the extracted symptom features; The auxiliary diagnosis module integrates the patient's historical medical history data, disease characteristic data and disease sensitivity factors, dynamically adjusts the model parameters, and performs a high-level diagnosis of myocardial diseases; The specific operation method of the auxiliary diagnosis module includes the following steps: A1: Constructing disease sensitivity factors: According to the disease feature results evaluated by the feature extraction module, assign corresponding weights to each disease feature, evaluate the impact of disease features on the diagnosis results through sensitivity analysis, and construct disease sensitivity factors by combining disease feature weights with sensitivity analysis results; A2: Historical medical history analysis: Based on historical medical history data, the fluctuation of symptom characteristics is evaluated by calculating the standard deviation of symptom characteristics; A3: Dynamically adjust the disease sensitivity factor: Update the disease sensitivity factor according to the fluctuation of disease characteristics; A4: Construction of disease evolution matrix: Construct a correlation matrix between features based on historical medical history data. Based on the correlation matrix, a diagnostic evolution model is constructed through a Markov chain to predict the changing trend of features. A5: Dynamic adaptive adjustment: Dynamic adaptation is performed based on the disease characteristic vector, the adjusted sensitivity factor and the disease evolution matrix; A6: Real-time weight adjustment: The weight is adjusted dynamically in real time based on the patient data information fed back.
2. The artificial intelligence-based myocardial disease auxiliary diagnosis system according to claim 1, characterized in that: A2, based on historical medical data, evaluates the volatility of symptom characteristics by calculating the standard deviation of symptom characteristics, and the implementation formula is: In formula (1), σ j Table 1 shows the standard deviation of the jth feature, N represents the number of samples, x jj represents the jth eigenvalue of the i-th sample, represents the average value of the j-th feature.
3. The artificial intelligence-based myocardial disease auxiliary diagnosis system according to claim 1, characterized in that: A3 updates the disease sensitivity factor according to the fluctuation of the disease characteristics. Assuming the disease sensitivity factor is F, the updated disease sensitivity factor is obtained by combining formula (1). The implementation formula is: F′=F+α(σ j -s j ′)(2), In formula (2), F′ represents the updated disease sensitivity factor, α represents the adjustment coefficient, and σ j and σ j ′ represents the standard deviation before and after adjustment.
4. The artificial intelligence-based myocardial disease auxiliary diagnosis system according to claim 1, characterized in that: A4 constructs a correlation matrix between features based on historical medical history data, and the implementation formula is: In the formula, R jk represents the correlation coefficient between the jth feature and the kth feature, x ij and x ik denote the jth and kth eigenvalues of the ith sample, respectively. and Represent the average values of the j-th and k-th features respectively, and N represents the number of samples; Based on the correlation matrix, the diagnostic evolution model is constructed through the Markov chain, and the implementation formula is: In the formula, P jk represents the transition probability from state j to state k, and They represent indicator functions respectively, when the condition x ij =j and x i ( k+1 )=k holds, then return 1; otherwise return 0.
5. The artificial intelligence-based myocardial disease auxiliary diagnosis system according to claim 1, characterized in that: A5 is dynamically adapted according to the disease characteristic vector, the adjusted sensitivity factor and the disease evolution matrix, and the implementation formula is: y=f(x,P,F′), In the formula, f represents the diagnostic model function, p represents the disease evolution matrix, F′ represents the adjusted disease sensitivity factor, and the probability of myocardial disease is predicted by the formula, and x represents the disease important feature matrix; The eigenvector is dynamically adjusted according to the disease evolution matrix, and the implementation formula is: x′=x+Δx, In the formula, x′ represents the adjusted disease characteristic vector, and Δx represents the adjustment amount of the disease characteristic vector.
6. The artificial intelligence-based myocardial disease auxiliary diagnosis system according to claim 1, characterized in that: The feature extraction module collects preprocessed clinical data, identifies and extracts symptom features that are highly correlated with myocardial disease, and evaluates the extracted symptom features; evaluates the importance of the symptom features of the clinical data to obtain features that are highly correlated with the symptom, and the implementation formula is: In the formula, I j represents the disease importance score of the jth feature, N represents the number of samples, y i and They represent the true value and predicted value of the i-th sample respectively. The feature with high disease importance score is recorded as X′. X′ represents the important feature matrix of the disease: X′=[x1x2...x J ], x J According to I j The important features of the disease are selected, J represents the number of important features of the disease, and the formula is realized by reducing the dimension of the important feature matrix of the disease: x=W T X′, In the formula, x represents the important feature matrix of the disease after dimensionality reduction, W T Represents the transpose of the weight matrix.
7. The artificial intelligence-based myocardial disease auxiliary diagnosis system according to claim 1, characterized in that: The symptom assessment module scores the severity of the symptom based on the symptom prediction results of the auxiliary diagnosis module, with a threshold value between 0 and 10, where a score above 7 indicates severe symptoms and a score below 4 indicates mild symptoms.
8. The artificial intelligence-based myocardial disease auxiliary diagnosis system according to claim 1, characterized in that: The symptom analysis module performs analysis based on the evaluation results of the symptom assessment module, and provides detailed diagnosis basis and diagnosis suggestions based on the analysis results.
9. The artificial intelligence-based myocardial disease auxiliary diagnosis system according to claim 1, characterized in that: The personalized diagnosis and treatment module provides personalized diagnosis and treatment based on the diagnosis suggestions and diagnosis results provided by the symptom analysis module.
Citation Information
Cited By
AI-based customer behavior analysis and prediction system and method
CN119250889A
