A predictive model for IgA nephropathy and its construction method

Abstract

This invention discloses an IgA nephropathy prediction model and its construction method, relating to the field of medical technology. By constructing a feature set, it measures whether the renal mesangial area is in a state of pathological enhancement, thereby assessing the potential risk of IgA nephropathy. Next, based on the measurement results in S1, it analyzes the occult risk of the lesions and further assesses whether the patient has a high risk of disease progression by calculating an occult progression factor. The calculation of the occult progression factor helps identify high-risk individuals among patients with mild symptoms. If the occult progression factor exceeds a set threshold, the system automatically triggers a prediction mechanism to guide subsequent treatment strategies. Based on this, through training the prediction model using the patient's historical physical examination data, it further outputs a comprehensive IgA prediction value and generates a risk level. This process achieves a comprehensive analysis from historical data to current physical examination results, ensuring the accuracy of the prediction results.

Description

Technical Field

[0001] This invention relates to the field of medical technology, specifically to an IgA nephropathy prediction model and its construction method. Background Technology

[0002] IgA nephropathy, a common primary glomerular disease, has become a key focus of nephrologists' daily diagnosis and treatment. Early diagnosis and monitoring of IgA nephropathy are crucial for improving patient prognosis and optimizing treatment plans. Research on disease prediction and analysis methods for IgA nephropathy is a means of quantifying disease progression trends based on clinical and experimental data, thereby providing doctors with more accurate decision-making support.

[0003] Although current medical research offers several clinical diagnostic methods for IgA nephropathy, such as urine tests, blood immune markers, and kidney biopsies, these traditional methods still have significant limitations in predicting disease progression and early intervention. Firstly, urine and blood analyses cannot fully reflect the dynamic changes in pathological deposits, and most methods rely on the appearance of clinical symptoms, easily missing the early stages of the disease. These shortcomings mainly stem from the limitations of traditional diagnostic methods, which often depend on relatively basic and static clinical indicators. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an IgA nephropathy prediction model and its construction method, thus solving the problems mentioned in the background section.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an IgA nephropathy prediction model and its construction method, comprising the following steps:

[0006] S1: Through physical examination, obtain relevant symptom information of the patient to be tested, and after preprocessing, obtain feature set. Based on feature set, measure whether the deposition is in a pathological enhancement state, and obtain historical data through the physical examination results of the patient to be tested in historical time periods.

[0007] S2: Based on the measurement results in S1, the risk of occult lesions in the patients to be tested is analyzed again to calculate and obtain the occult progression factor LPI. If the occult progression factor LPI exceeds the set threshold, the prediction mechanism is triggered, and the prediction model is trained in combination with the content of S1.

[0008] S3: Activate the prediction mechanism, output the IgA comprehensive prediction value IGX from the output of the prediction model, and generate the risk level;

[0009] S4: Based on the risk level, reverse the intervention recommendations to assess the patient's response to improvement.

[0010] Preferably, step S1 specifically includes:

[0011] S11: Various physical examinations are conducted to obtain the results of each examination. The results of each examination are cleaned and duplicate values ​​are removed to obtain the feature set of the patient to be tested. The feature set includes the homogeneity index of urinary red blood cell morphology (URR), proteinuria fluctuation frequency (PRF), serum Gd-IgA1 concentration density value (IGD), and mesangial deposition grayscale recognition factor (MES) of the patient to be tested in this physical examination. Among them, the various physical examinations include urine test, blood test, and renal biopsy.

[0012] Preferably, step S1 further includes:

[0013] S12: The specific method for obtaining the urinary erythrocyte morphology homogeneity index (URR) is as follows:

[0014] Fresh urine samples were collected from patients to be tested, centrifuged, and urine sediment slides were prepared. The contours of red blood cells were extracted by microscopic imaging to obtain the morphological characteristics of red blood cells. Each red blood cell was represented as a morphological feature vector, and the consistency of red blood cell morphological characteristics was evaluated by clustering across the entire sample. Finally, the urinary red blood cell morphology homogeneity index (URR) was constructed.

[0015] Preferably, step S1 further includes:

[0016] S13: Based on the feature set and after dimensionless processing, the deposition response coefficient RSD of the patient in this physical examination is calculated to measure whether the deposition is in a pathological enhancement state. The specific calculation method is as follows:

[0017] ;

[0018] Wherein, URR represents the homogeneity index of urinary erythrocyte morphology, PRF represents the frequency of proteinuria fluctuation, IGD represents the serum Gd-IgA1 concentration density value, MES represents the mesangial deposition gray scale recognition factor, and e represents the Euler number, with a value of approximately 2.71828.

[0019] S14: Obtain the deposition response coefficient from the last physical examination result in the historical data by retrieving the data. And the deposition response coefficient from the last physical examination result in the historical data. The deposition response coefficient RSD of the patient's current physical examination is compared numerically. If the RSD of the patient's current physical examination exceeds the deposition response coefficient of the last physical examination result in the historical data, the result is considered negative. If the current patient is in a pathological enhancement state, a disease latency analysis command will be issued. If the deposition response coefficient RSD of the current physical examination of the patient does not exceed the deposition response coefficient of the last physical examination result in the historical data, then the analysis will be performed. If the patient is not in a state of pathological enhancement, then it is determined that the patient being tested is not currently in a state of enhanced pathology.

[0020] Preferably, step S2 specifically includes:

[0021] S21: The prediction model is trained by dividing historical data into training and validation sets. The model is validated by measuring the degree of abnormal fluctuation in training stability. Upon receiving the disease latency analysis instruction, the occult disease risk of the patient under test is analyzed again to obtain the occult progression factor LPI of the patient under test in this physical examination. The specific formula is as follows:

[0022] ;

[0023] in, This is to increase the weight of deposition intensity on the rate of advancement. It nonlinearly amplifies abnormal red blood cell behavior while ensuring a limited value range.

[0024] Preferably, step S2 further includes:

[0025] S22: A pre-set threshold is used to determine the occultation of lesions in the patient being tested by comparing the occult progression factor LPI with the pre-set threshold. The specific details are as follows:

[0026] When the occult progression factor LPI exceeds a pre-set threshold, it indicates that the lesions in the patient being tested are occult, and the prediction mechanism will be triggered at this time.

[0027] When the occult progression factor LPI does not exceed the preset threshold, it indicates that the lesions of the patient being tested do not currently exhibit occultity, and the prediction mechanism is not triggered at this time.

[0028] Preferably, step S3 specifically includes:

[0029] S31: Pre-collect the physical examination results of the patient to be tested in historical time periods to obtain historical data. The historical data includes the urinary erythrocyte morphology homogeneity index (URR), proteinuria fluctuation frequency (PRF), serum Gd-IgA1 concentration density value (IGD), mesangial deposition grayscale recognition factor (MES), and deposition response coefficient (RSD) and latent progression factor (LPI) obtained from each physical examination in the patient's historical time periods. Combine the deposition response coefficient (RSD) and latent progression factor (LPI) of the patient's current physical examination to obtain an input feature set. By inputting the input feature set into a convolutional neural network model and training it, the trained convolutional neural network model is used as a prediction model to output the comprehensive IgA prediction value (IGX) from the output of the prediction model. The comprehensive IgA prediction value (IGX) is calculated according to the following formula:

[0030] ;

[0031] in, , where is the Sigmoid activation function used for probability normalization, and x in this formula refers to , This represents the mean of all depositional response coefficients in the input feature set. This represents the mean of all hidden progression factors in the input feature set. and They represent the mean values ​​of all depositional response coefficients in the input feature set, respectively. and the mean of all hidden progression factors in the input feature set The weight.

[0032] Preferably, step S3 further includes:

[0033] S32: By inputting a feature set, a risk range is set. The risk range is then compared numerically with the IgA comprehensive prediction value IGX to generate a corresponding risk level. Specifically:

[0034] If the IgA comprehensive predictive value IGX exceeds the risk range, it indicates that the patient to be tested has IgA nephropathy and is at a high risk level. At this time, the first intervention instruction will be issued.

[0035] If the comprehensive IgA predictive value IGX falls within the risk range, it indicates that the patient being tested has IgA nephropathy at a medium risk level, and a second intervention instruction will be issued at this time.

[0036] If the IgA comprehensive predictive value IGX does not exceed the risk range, it indicates that the patient under test has IgA nephropathy at a low risk level, and intervention instruction number three will be issued.

[0037] Preferably, step S4 specifically includes:

[0038] S41: When the No. 1 and No. 2 intervention instructions are received, the intervention recommendations will be deduced based on the risk level, and the reversible risk assessment factor RVI will be constructed.

[0039] S42: When the reversible risk assessment factor RVI exceeds the pre-set assessment threshold, it means that the deposits and immune response are in the initial stage. Treatment intervention is carried out by taking immunosuppressive drugs and regularly following up to adjust the medication in a timely manner. If the reversible risk assessment factor RVI does not exceed the pre-set assessment threshold, it means that the deposits have begun to appear. At this time, dialysis treatment will be carried out, combined with renal biopsy, to further clarify the pathological type, assess the degree of fibrosis and residual renal function.

[0040] S43: Based on the intervention recommendations in S42, perform treatment intervention on the patient. After the treatment intervention, repeat the content of S1 to S3 to reassess the corresponding risk level until the third intervention instruction is issued. Adjust the medication in a timely manner by taking immunosuppressive drugs to perform treatment intervention on the patient.

[0041] Preferably, an IgA nephropathy prediction model is constructed using the aforementioned IgA nephropathy prediction construction method.

[0042] This invention provides an IgA nephropathy prediction model and its construction method, which has the following beneficial effects:

[0043] (1) First, relevant symptoms of patients are obtained through physical examination, and the data are preprocessed to construct a feature set. Based on these features, the pathological enhancement state of the renal mesangial area is measured, thereby assessing the potential risk of IgA nephropathy. This process utilizes multiple key indicators, such as the homogeneity index of urinary erythrocyte morphology, the frequency of proteinuria fluctuations, the serum Gd-IgA1 concentration density value, and the mesangial deposition grayscale identification factor, to achieve a comprehensive quantification of deposition activity. Compared with traditional single-indicator methods, this fusion of multi-dimensional data provides a more accurate and reliable assessment of deposition response. Next, based on the measurement results in S1, the occult risk of lesions is analyzed. By calculating the occult progression factor LPI, the risk of disease progression is further assessed. The calculation of the occult progression factor LPI helps identify high-risk individuals among patients with mild symptoms, which is of great significance for early intervention. If the occult progression factor LPI exceeds the set threshold, the system will automatically trigger a prediction mechanism to provide guidance for subsequent treatment strategies. Based on this, by training the predictive model with the patient's historical physical examination data, the system further outputs a comprehensive IgA predictive value (IGX) and generates a risk level. This process achieves a comprehensive analysis from historical data to current physical examination results, ensuring the accuracy of the prediction results. Based on the generated risk level, the system can automatically deduce intervention recommendations, thereby enabling the formulation of personalized treatment plans. Especially in cases where the patient's condition is mild but the potential risk of progression is high, it can provide timely treatment intervention suggestions to minimize the risk of further deterioration. In summary, the IgA nephropathy prediction method of this invention, through the fusion of multi-dimensional data, accurate assessment of occult progression risk, and recommendation of personalized intervention strategies, significantly improves the early diagnosis, treatment predictability, and individualized treatment level of IgA nephropathy, and has important clinical significance for reducing kidney function damage and improving patients' quality of life.

[0044] (2) In step S21, the prediction model is repeatedly trained and validated by dividing historical data into training and validation sets. By measuring the stability and abnormal fluctuations of the model during training, the model's generalization ability under different data scenarios is ensured. This process enables the model to adapt to the clinical manifestations and physical examination data of different patients, enhancing its ability to judge occult progression. Especially in patients with mild symptoms, it can effectively identify high-risk individuals with mild surface symptoms but active deposition, thereby enabling early intervention. By calculating the occult progression factor LPI and combining it with factors such as deposition intensity, abnormal erythrocyte behavior, and proteinuria fluctuations, the model can quantify the potential progression risk of the patient's condition.

[0045] (3) Step S4 automatically generates intervention recommendations based on the patient's risk level, and then adjusts the treatment strategy based on the RVI value, providing strong decision support for clinical treatment. In step S41, the reversible risk assessment factor RVI is further constructed by the intervention instructions generated based on the patient's risk level. This factor combines the comprehensive predictive value of IgA IGX and provides a treatment window by assessing the patient's deposition status and immune response. Attached Figure Description

[0046] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0047] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0048] Example 1

[0049] Please see Figure 1 This invention provides an IgA nephropathy prediction model and its construction method, comprising the following steps:

[0050] S1: Through physical examination, obtain relevant symptom information of the patient to be tested, and after preprocessing, obtain feature set. Based on feature set, measure whether the deposition is in a pathological enhancement state, and obtain historical data through the physical examination results of the patient to be tested in historical time periods.

[0051] S2: Based on the measurement results in S1, the risk of occult lesions in the patients to be tested is analyzed again to calculate and obtain the occult progression factor LPI. If the occult progression factor LPI exceeds the set threshold, the prediction mechanism is triggered, and the prediction model is trained in combination with the content of S1.

[0052] S3: Activate the prediction mechanism, output the IgA comprehensive prediction value IGX from the output of the prediction model, and generate the risk level;

[0053] S4: Based on the risk level, reverse the intervention recommendations to assess the patient's response to improvement.

[0054] In this embodiment of the invention, the method not only overcomes the limitations of traditional detection methods but also integrates multi-dimensional data (such as urine, serum, and imaging) to automatically generate predictions of disease progression and intervention suggestions. Specifically, step S1 collects relevant symptoms from patients through physical examinations, such as urinary red blood cell morphology, proteinuria fluctuation frequency, and serum Gd-IgA1 concentration. After preprocessing, a feature set is formed, which then measures whether the deposition is in a pathological enhancement state, providing important basis for subsequent analysis.

[0055] Step S2 analyzes the latent progression factor LPI. If the LPI exceeds a set threshold, a prediction mechanism is triggered to indicate potential risk. Step S3 combines historical data with physical examination results to train a prediction model and generate a comprehensive IgA prediction value (IGX) and risk level, thereby providing early disease warning.

[0056] For example, assuming a patient presents with mild proteinuria and occult hematuria during a physical examination, this method can accurately measure the pathological state of their kidney deposits. If the occult progression factor LPI value is high and the IgA composite predictive value IGX indicates that the patient is at intermediate to high risk, the system will suggest that the patient may have a strong risk of occult disease progression and recommend further clinical examination and early intervention, such as immunosuppressive therapy or renal biopsy. Ultimately, through this intelligent predictive method, patients can receive timely intervention before their condition significantly worsens, avoiding missing the optimal treatment window, reducing the incidence of end-stage renal disease, and improving patients' quality of life and treatment outcomes.

[0057] Furthermore, doctors can use the risk levels generated by this method to develop personalized treatment plans for patients at different risk levels, improving the accuracy and effectiveness of treatment. In summary, this invention provides strong support for early screening and treatment decisions in IgA nephropathy, reducing the risk of misdiagnosis and missed diagnosis, and improving patient survival rates and treatment outcomes.

[0058] Example 2

[0059] Please refer to Figure 1 Specifically: S1 includes the following steps:

[0060] S11: Various physical examinations are conducted to obtain the results of each examination. The results of each examination are cleaned and duplicate values ​​are removed to obtain the feature set of the patient to be tested. The feature set includes the homogeneity index of urinary red blood cell morphology (URR), proteinuria fluctuation frequency (PRF), serum Gd-IgA1 concentration density value (IGD), and mesangial deposition grayscale recognition factor (MES) of the patient to be tested in this physical examination. Among them, the various physical examinations include urine test, blood test, and renal biopsy.

[0061] The specific steps in S1 also include:

[0062] S12: The specific method for obtaining the urinary erythrocyte morphology homogeneity index (URR) is as follows:

[0063] Fresh urine samples were collected from patients to be tested, centrifuged, and urine sediment slides were prepared. The contours of red blood cells were extracted by microscopic imaging to obtain the morphological characteristics of red blood cells. Each red blood cell was represented as a morphological feature vector. Clustering was used to assess the consistency of red blood cell morphological characteristics in the entire sample. Finally, the Urine Red Blood Cell Morphology Homogeneity Index (URR) was constructed to characterize pathological hematuria.

[0064] Among them, proteinuria fluctuation frequency (PRF) represents the number of cycles of abnormal fluctuations in urinary protein levels per unit time, and is used to reflect the stage-by-stage recurrence of the disease course, because IgA nephropathy often has "intermittent mild to moderate proteinuria", which can fluctuate with infection, stress, etc.

[0065] Serum Gd-IgA1 concentration density (IGD) represents its aggregation concentration and density distribution in serum, reflecting the intensity of abnormal immune accumulation in the body. Gd-IgA1 concentration is detected by collecting blood samples and using the ELISA specific antibody method.

[0066] The mesangial deposition grayscale identification factor (MES) reflects the image grayscale characteristics of immunodeposit in the mesangial region of renal biopsy tissue pathological sections. Higher grayscale indicates greater deposition density and more concentrated IgA complexes. Pathological sections are obtained through renal biopsy (IgA immunohistochemical staining), and the localization (segmentation) and average grayscale values ​​of the mesangial region are extracted to output image grayscale normalization parameters, specifically: ;in, The grayscale mean is... and These represent the lower and upper limits of the grayscale value, respectively.

[0067] The specific steps in S1 also include:

[0068] S13: Based on the feature set and after dimensionless processing, the deposition response coefficient RSD of the patient in this physical examination is calculated to measure whether the deposition is in a pathological enhancement state. The specific calculation method is as follows:

[0069] ;

[0070] Wherein, URR represents the homogeneity index of urinary erythrocyte morphology, PRF represents the frequency of proteinuria fluctuation, IGD represents the serum Gd-IgA1 concentration density value, MES represents the mesangial deposition gray scale recognition factor; e represents the Euler number, with a value of approximately 2.71828.

[0071] In the formula, the outer square root is used to standardize the sedimentation activity to a positive continuous space, which facilitates subsequent processing;

[0072] The first item combines morphological hematuria abnormalities with fluctuations in proteinuria; It emphasizes the dominant role of severe pathological hematuria in the deposition reaction. Logarithmic compression transformation of proteinuria frequency is performed to suppress the amplification effect of high-frequency values;

[0073] The second term fuses the abnormal IgA immune concentration with the image deposition grayscale, and the denominator... This is a typical Sigmoid transform, used to normalize the mesangial deposition grayscale identification factor (MES) to (0, 1), emphasizing that the larger the grayscale (the heavier the deposition), the closer the Sigmoid value is to 1; this part as a whole plays a role in enhancing and adjusting the deposition image.

[0074] S14: Obtain the deposition response coefficient from the last physical examination result in the historical data by retrieving the data. And the deposition response coefficient from the last physical examination result in the historical data. The deposition response coefficient RSD of the patient's current physical examination is compared numerically. If the RSD of the patient's current physical examination exceeds the deposition response coefficient of the last physical examination result in the historical data, the result is considered negative. If the current patient is in a pathological enhancement state, a disease latency analysis command will be issued. If the deposition response coefficient RSD of the current physical examination of the patient does not exceed the deposition response coefficient of the last physical examination result in the historical data, then the analysis will be performed. If the patient is not in a state of pathological enhancement, then it is determined that the patient being tested is not currently in a state of enhanced pathology.

[0075] In this embodiment of the invention, in step S1, relevant symptoms of the patient to be tested are obtained through physical examination, and a feature set is constructed by cleaning and deduplicating the data. This feature set includes the urinary red blood cell morphology homogeneity index (URR), proteinuria fluctuation frequency (PRF), serum Gd-IgA1 concentration density value (IGD), and mesangial deposition grayscale recognition factor (MES). This feature set covers urine test, blood test, and renal biopsy data, and comprehensively considers the patient's multi-dimensional clinical manifestations, helping to more comprehensively assess the pathological enhancement status and avoid the risk of misdiagnosis caused by a single indicator.

[0076] For example, suppose a patient presents with mild proteinuria during a physical examination, but the morphology of red blood cells in the urine shows consistent abnormalities and the concentration of Gd-IgA1 in the serum is high. By analyzing these characteristics, the system can quantify the risk of deposition and detect potential IgA nephropathy in a timely manner.

[0077] In step S13, the deposition response coefficient (RSD) is calculated based on the feature set to measure whether the deposition is in a pathological enhancement state. The calculation of RSD comprehensively considers the grayscale information of hematuria, proteinuria, immune abnormalities, and mesangial deposition. Dimensionless processing enables RSD to effectively measure the activity level of deposition, providing a scientific basis for assessing the risk of disease progression. In summary, the IgA nephropathy prediction method of this invention, by comprehensively collecting patient physical examination data and combining it with historical information, accurately assesses the patient's disease risk and makes timely treatment intervention recommendations, which helps to improve the early diagnosis and personalized treatment of IgA nephropathy, significantly improving patient prognosis.

[0078] Example 3

[0079] Please refer to Figure 1 Specifically: The specific steps of S2 include:

[0080] S21: The prediction model is trained by dividing historical data into training and validation sets. The model is validated by measuring abnormal fluctuations in training stability. Upon receiving a disease latency analysis instruction, the occult disease risk of the patient under test is analyzed again to enhance the model's ability to identify individuals with mild clinical manifestations but highly active disease deposition. This yields the occult progression factor (LPI) of the patient under test in this physical examination, specifically obtained through the following formula:

[0081] ;

[0082] in, This is to increase the weight of deposition intensity on the rate of advancement. It nonlinearly amplifies abnormal red blood cell behavior while ensuring a limited value range (the value in radians is the upper limit π / 2).

[0083] Adding +0.5 to the denominator is to avoid the denominator being 0, and at the same time to compress and regulate the stability of low-frequency proteinuria.

[0084] This is to avoid drastic changes in the prediction model values ​​and to enhance the convergence of the results when the exponent is too high.

[0085] The specific steps in S2 also include:

[0086] S22: A pre-set threshold is used to determine the occultation of lesions in the patient being tested by comparing the occult progression factor LPI with the pre-set threshold. The specific details are as follows:

[0087] When the occult progression factor LPI exceeds a pre-set threshold, it indicates that the lesions of the patient being tested are occult and have a strong occult progression trend. The higher the potential risk of progression, the more likely the prediction mechanism will be triggered.

[0088] When the occult progression factor LPI does not exceed the preset threshold, it indicates that the lesions of the patient being tested do not currently exhibit occultity, and the prediction mechanism is not triggered at this time.

[0089] In this embodiment of the invention, the IgA nephropathy prediction method provided by the present invention significantly improves the ability to identify the early stages of the disease through the calculation and analysis of the occult progression factor LPI, especially in patients with mild clinical manifestations but active deposition, and can accurately predict the risk of disease progression.

[0090] Specifically, step S2 divides historical data into training and validation sets for training the predictive model. By measuring the degree of abnormal fluctuations during training, the stability and accuracy of the model are ensured. This process effectively eliminates data bias and uncertainty, enabling the model to adapt to the clinical characteristics of different patients in practical applications. Through the calculation of the occult progression factor LPI, step S2 further enhances the predictive model's ability to identify occult progression, especially in patients with subtle clinical symptoms but significant deposition. The LPI calculation formula integrates factors such as deposition intensity, abnormal erythrocyte behavior, and proteinuria variability, thereby quantifying the risk of occult lesions and improving predictive accuracy. Particularly during the pathological enhancement phase, an increase in the occult progression factor LPI triggers the predictive mechanism, providing a basis for subsequent treatment and intervention. This mechanism allows the predictive model to not only identify obvious lesions but also accurately identify patients with active deposition but mild clinical manifestations, providing a scientific basis for early intervention in these high-risk individuals.

[0091] Furthermore, the threshold setting and risk assessment mechanism in step S2 further enhances the model's practical application value. By comparing the occult progression factor LPI with a preset threshold, the degree of occult progression of lesions can be accurately distinguished. When the occult progression factor LPI exceeds the set threshold, the system will automatically trigger the prediction mechanism, reminding doctors to take more proactive monitoring and intervention measures; conversely, when the occult progression factor LPI does not exceed the threshold, the patient's condition is relatively stable, and the system will not trigger unnecessary prediction instructions. This mechanism not only effectively avoids over-intervention but also ensures the rational use of clinical resources. In summary, this invention, through precise calculation and dynamic monitoring of occult progression factors, provides a new solution for the early prediction and personalized treatment of IgA nephropathy, significantly improving the clinical application value of the prediction model, enabling timely identification of patients with high disease potential, and allowing for precise intervention.

[0092] Example 4

[0093] Please refer to Figure 1 Specifically: The specific steps of S3 include:

[0094] S31: Pre-collect the physical examination results of the patient to be tested in historical time periods to obtain historical data. The historical data includes the urinary erythrocyte morphology homogeneity index (URR), proteinuria fluctuation frequency (PRF), serum Gd-IgA1 concentration density value (IGD), mesangial deposition grayscale recognition factor (MES), and deposition response coefficient (RSD) and latent progression factor (LPI) obtained from each physical examination in the patient's historical time periods. Combine the deposition response coefficient (RSD) and latent progression factor (LPI) of the patient's current physical examination to obtain an input feature set. By inputting the input feature set into a convolutional neural network model and training it, the trained convolutional neural network model is used as a prediction model to output the comprehensive IgA prediction value (IGX) from the output of the prediction model. The comprehensive IgA prediction value (IGX) is calculated according to the following formula:

[0095] ;

[0096] in, , where is the Sigmoid activation function used for probability normalization, and x in this formula refers to , This represents the mean of all depositional response coefficients in the input feature set. This represents the mean of all hidden progression factors in the input feature set. and They represent the mean values ​​of all depositional response coefficients in the input feature set, respectively. and the mean of all hidden progression factors in the input feature set The weights (which can be obtained through model training);

[0097] Specifically, the training process includes feature extraction, recognition error assessment, and minimization of training error.

[0098] The trained convolutional neural network model is used as the prediction model. The prediction model mainly includes the following structural components: convolutional layers are used to extract features, activation layers are used to increase non-linear expressive power, pooling layers are used to reduce dimensionality and enhance position invariance, flattening layers are used to convert vectors into inputs to fully connected layers, fully connected layers are used to combine high-dimensional features, and output layers are used to output the final result.

[0099] By employing weighted fusion and nonlinear activation, a structurally stable probabilistic index is output.

[0100] The Sigmoid activation function is used to make the IgA composite prediction value IGX range from 0 to 1, which has medical probabilistic interpretability.

[0101] The specific steps in S3 also include:

[0102] S32: By inputting a feature set, a risk range is set. The risk range is then compared numerically with the IgA comprehensive prediction value IGX to generate a corresponding risk level. Specifically:

[0103] If the IgA comprehensive predictive value IGX exceeds the risk range, it indicates that the patient to be tested has IgA nephropathy and is at a high risk level. At this time, the first intervention instruction will be issued.

[0104] If the comprehensive IgA predictive value IGX falls within the risk range, it indicates that the patient being tested has IgA nephropathy at a medium risk level, and a second intervention instruction will be issued at this time.

[0105] If the IgA comprehensive predictive value IGX does not exceed the risk range, it indicates that the patient under test has IgA nephropathy at a low risk level, and intervention instruction number three will be issued.

[0106] In this embodiment of the invention, by combining the patient's historical data with the current physical examination results, a convolutional neural network (CNN) model is used for accurate prediction, providing effective support for the early diagnosis and personalized treatment of IgA nephropathy. In step S31, the patient's historical physical examination data (such as the urinary erythrocyte morphology homogeneity index URR, proteinuria fluctuation frequency PRF, serum Gd-IgA1 concentration density value IGD, mesangial deposition grayscale recognition factor MES, deposition response coefficient RSD, and occult progression factor LPI) are collected in advance and combined with the current physical examination results to construct a comprehensive feature set. By inputting these features into the convolutional neural network model, the model performs deep learning on the data during training and generates a comprehensive IgA prediction value IGX. This value ranges from 0 to 1 and is mapped to a probability value through the sigmoid activation function, reflecting the probability that the patient has IgA nephropathy.

[0107] In step S32, based on the IgA comprehensive predictive value (IGX), the system sets different risk levels. When the IgA comprehensive predictive value (IGX) exceeds the high-risk threshold, the system issues intervention command number one, indicating that the patient is in a high-risk state and requires immediate treatment intervention; when the IGX value falls into the intermediate-risk range, intervention command number two is issued, prompting regular monitoring and treatment adjustment; and when the IGX value is below the low-risk threshold, intervention command number three is issued, indicating that the patient is in a low-risk state and no major intervention is required. Through this multi-dimensional data fusion and intelligent prediction method, this invention significantly improves the early prediction accuracy of IgA nephropathy and the personalized judgment of treatment response. It not only provides doctors with accurate risk assessments but also helps to develop more precise intervention strategies, thereby improving treatment outcomes and patients' quality of life.

[0108] Example 5

[0109] Please refer to Figure 1 Specifically: The specific steps of S4 include:

[0110] S41: Upon receiving Intervention Directive 1 and Intervention Directive 2, intervention recommendations will be derived based on the risk level. Specific details include:

[0111] The reversible risk assessment factor RVI is constructed as follows:

[0112] ;

[0113] in, This represents the overall predicted value of IgA. This indicates that the current risk level has not reached the high-risk threshold, and an intervention window is available.

[0114] The higher the reversibility risk assessment factor (RVI), the stronger the reversibility, and the more likely it is to be treated with drugs such as RAS blockers.

[0115] The Reversible Risk Assessment Factor (RVI) is used to assess whether an individual has an intervention window or potential reversibility.

[0116] The larger the overall value, the higher the reversibility (for non-deposition-dominated individuals).

[0117] S42: When the reversible risk assessment factor RVI exceeds the pre-set assessment threshold, it means that the deposits and immune response are in the initial stage. Treatment intervention is carried out by taking immunosuppressive drugs, such as glucocorticoids (e.g., prednisone), cyclophosphamide, etc., and by regular follow-up, assessing changes in renal function and urine indicators every 3-6 months, and adjusting the medication in a timely manner. If the reversible risk assessment factor RVI does not exceed the pre-set assessment threshold, it means that the deposits have begun to appear and the damage may have spread to an irreversible stage. At this time, dialysis treatment will be carried out, and alternative treatments such as hemodialysis or peritoneal dialysis can be considered. Combined with renal biopsy, further clarify the pathological type, assess the degree of fibrosis and residual renal function.

[0118] S43: Based on the intervention recommendations in S42, perform treatment intervention on the patient. After the treatment intervention, repeat the content of S1 to S3 to reassess the corresponding risk level until the third intervention instruction is issued. Adjust the medication in a timely manner by taking immunosuppressive drugs to perform treatment intervention on the patient.

[0119] In this embodiment of the invention, the IgA nephropathy prediction method significantly improves the accuracy of treatment intervention decisions for IgA nephropathy in step S4 by constructing a reversible risk assessment factor (RVI) and combining it with the risk level generated by the prediction model. This step enables timely monitoring of the patient's condition and the development of personalized treatment plans, allowing for effective intervention at an early stage of the disease and reducing the risk of further deterioration.

[0120] In step S41, upon receiving intervention instructions one and two, the system derives the reversible risk assessment factor (RVI) based on the patient's comprehensive IgA predictive value (IGX) and risk level. This factor is adjusted according to the IgA predictive value and a set high-risk threshold. If the patient's risk is within the interventionable stage, a high RVI value indicates that the patient's condition is still reversible and can be treated with medications such as RAS blockers. An increase in the RVI helps the system accurately determine whether the patient can slow disease progression and achieve an interventional effect through medication. In step S42, when the RVI exceeds a preset threshold, it indicates that deposits and immune responses are in the early stages, making immunosuppressive drug therapy appropriate. For example, glucocorticoids (such as prednisone) and cyclophosphamide can effectively slow disease progression. Simultaneously, the patient's renal function and urine parameters should be assessed every 3-6 months to adjust the medication regimen promptly for optimal treatment results.

[0121] When the reversible risk assessment factor (RVI) does not exceed the threshold, it means that the patient's deposits have begun to appear and the damage is relatively severe. At this time, dialysis treatment is necessary, along with a renal biopsy to clarify the pathological type and degree of fibrosis, to ensure timely treatment. In step S43, the system repeats steps S1 to S3 based on the follow-up results after treatment to re-examine and reassess the patient's condition to determine whether further adjustments to the treatment plan are needed, ensuring the patient receives the most appropriate intervention. For example, if patient A has a high IGX value and a high RVI at the first physical examination, indicating that their condition is in an early reversible state, the system recommends using a RAS blocker and regular monitoring of renal function. In the follow-up 3 months later, patient A's renal function remained stable, urine indicators were normal, and the RVI value was still high, indicating that the drug treatment was effective, and maintenance treatment should continue. Patient B, although having a low RVI value in the initial diagnosis and with visible deposits, was deemed by the system to have an irreversible condition. Dialysis treatment was recommended in conjunction with a renal biopsy for further evaluation to ensure the accuracy of the treatment. Through this precise treatment intervention mechanism, this invention further enhances the early diagnosis and personalized treatment capabilities for patients with IgA nephropathy, effectively improving treatment outcomes and reducing kidney function damage.

[0122] Specifically, all the aforementioned parameters have been dimensionally processed to eliminate units.

[0123] Specifically: an IgA nephropathy prediction model, constructed using the aforementioned IgA nephropathy prediction construction method.

[0124] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An IgA nephropathy prediction data processing system, characterized by, Comprise the following modules: Data processing module: the related symptom condition of the patient to be detected obtained by physical examination is preprocessed to obtain a feature set, based on the feature set, whether the deposition is in a pathological enhancement state is measured, and the historical data is obtained through the physical examination results of the patient to be detected in the historical period; Concealment risk analysis module: based on the measurement results in the data processing module, the risk of concealing the patient's lesion is analyzed again to calculate and obtain the concealment progression factor LPI, if the concealment progression factor LPI exceeds the set threshold, the prediction model is started to calculate, and the prediction model is trained in combination with the data processing module content; Prediction module: the physical examination results of the patient to be detected in the historical period are collected in advance to obtain historical data, and the input feature set is obtained in combination with the deposition response coefficient RSD of the patient to be detected this time and the concealment progression factor LPI of the patient to be detected this time; The comprehensive prediction value IGX representing the risk of IgA nephropathy is output from the output end of the prediction model; Reversible risk assessment module: based on the comprehensive prediction value IGX, the reversible risk assessment factor RVI for assessing the potential of disease improvement is calculated; The specific execution logic of the data processing module comprises: S13: based on the feature set and after dimensionless processing, whether the deposition is in a pathological enhancement state is measured to calculate and obtain the deposition response coefficient RSD of the patient to be detected this time, and the specific calculation and obtaining method is as follows: ； Wherein, URR represents the urine red blood cell morphology homogeneity index, PRF represents the proteinuria fluctuation frequency, IGD represents the serum Gd-IgA1 concentration density value, MES represents the mesangial deposition gray recognition factor, e represents Euler number, and the numerical value is about 2.71828; S14: obtaining the deposition response coefficient in the last physical examination result in the historical data by calling the last physical examination result in the historical data , and comparing the deposition response coefficient in the last physical examination result in the historical data with the deposition response coefficient RSD of the current physical examination of the patient to be detected, if the deposition response coefficient RSD of the current physical examination of the patient to be detected exceeds the deposition response coefficient in the last physical examination result in the historical data , it is judged that the patient to be detected is in a pathological enhancement state, and at this time, a disease latency analysis instruction is issued; if the deposition response coefficient RSD of the current physical examination of the patient to be detected does not exceed the deposition response coefficient in the last physical examination result in the historical data , it is judged that the patient to be detected is not in a pathological enhancement state; The specific execution logic of the concealment risk analysis module comprises: S21: the historical data is divided into a training set and a validation set to train the prediction model, and the prediction model is verified by measuring the abnormal fluctuation degree of the training stability, and after receiving the disease latency analysis instruction, the risk of concealing the patient's lesion is analyzed again to obtain the concealment progression factor LPI of the patient to be detected this time, which is obtained by the following formula: ； wherein, is to enhance the weight of the progress of the deposition strength, is to nonlinearly amplify the abnormal behavior of red blood cells while ensuring the value range is limited; The input feature set is input into the convolutional neural network model, the trained convolutional neural network model is used as the prediction model, the IgA comprehensive prediction value IGX is output from the output end of the prediction model, and the comprehensive prediction value IGX is calculated and obtained according to the following formula: ； wherein, is a sigmoid activation function for probability normalization, wherein x refers to , denotes the mean of all deposition response coefficients in the input feature set, denotes the mean of all hidden progression factors in the input feature set, and denote the weights of the mean of all deposition response coefficients in the input feature set and the mean of all hidden progression factors in the input feature set respectively.

2. The IgA nephropathy prediction data processing system according to claim 1, characterized in that: The specific execution logic of the data processing module further comprises: S11: the feature set of the patient to be detected is obtained by data cleaning and repeated value elimination of each physical examination result, the feature set comprises the urine red blood cell morphology homogeneity index URR, the proteinuria fluctuation frequency PRF, the serum Gd-IgA1 concentration density value IGD and the mesangial deposition gray recognition factor MES of the patient to be detected this time, and each physical examination comprises urine detection, blood detection and kidney biopsy.

3. The IgA nephropathy prediction data processing system according to claim 2, characterized in that: The specific execution logic of the data processing module further includes: S12: the obtaining method of the urine red blood cell morphology homogeneity index URR is specifically: collecting a fresh urine sample of a patient to be detected, then centrifuging, preparing a urine sediment glass slide, and collecting a red blood cell contour through microscopic image collection to obtain red blood cell morphology characteristics, and expressing each red blood cell as a morphology characteristic vector to evaluate the consistency degree of the red blood cell morphology characteristics in the whole sample, and finally constructing the urine red blood cell morphology homogeneity index URR; the specific obtaining method of the proteinuria fluctuation frequency PRF is: ; wherein, Bs is the fluctuation frequency, and T is the monitoring period; the serum Gd-IgA1 concentration density value IGD is obtained by using an ELISA specific antibody method for detection; the mesangial deposition gray recognition factor MES is obtained by kidney puncture to obtain a pathological section, extracting a mesangial region positioning and a deposition gray average value to output an image gray normalization parameter, which is specifically: ; wherein, is a gray mean value, and are lower and upper limits of the gray value respectively.

4. The IgA nephropathy prediction data processing system according to claim 3, characterized in that: The specific execution logic of the hidden risk analysis module further includes: S22: A threshold is preset, and the hidden progression factor LPI is compared with the preset threshold to determine the concealment of the lesion of the patient to be detected, and the specific content is as follows: When the hidden progression factor LPI exceeds the preset threshold, it indicates that the lesion of the patient to be detected has concealment, and at this time the prediction mechanism will be triggered; When the hidden progression factor LPI does not exceed the preset threshold, it indicates that the lesion of the patient to be detected does not have concealment at present, and at this time the prediction mechanism is not triggered.

5. The IgA nephropathy prediction data processing system according to claim 4, characterized in that: The specific execution logic of the prediction module includes: S31: The medical examination results of the patient to be detected in the historical period are collected in advance to obtain historical data, and the historical data includes the urine red blood cell morphology homogeneity index URR, the proteinuria fluctuation frequency PRF, the serum Gd-IgA1 concentration density value IGD, the mesangial deposition gray recognition factor MES, the deposition response coefficient RSD and the hidden progression factor LPI obtained in each medical examination in the historical period of the patient to be detected, and the deposition response coefficient RSD and the hidden progression factor LPI obtained in this medical examination of the patient to be detected are combined to obtain an input feature set.

Images