Method and system for evaluating severity of ulcerative colitis

Abstract

The invention discloses an ulcerative colitis severity assessment method and system, and relates to the technical field of artificial intelligence, and the method comprises the following steps: data collection and ulcerative colitis symptom classification; preliminarily evaluating the severity of ulcerative colitis according to clinical data; performing correlation analysis based on the preliminary evaluation model and the severity score; performing intelligent diagnosis according to the combination of the severity score and the real-time physiological monitoring data; personalized treatment scheme recommendation is carried out according to the intelligent diagnosis result; evaluating the recovery progress of the patient by using the personalized treatment scheme; performing subsequent illness state prediction according to the recovery progress data; and a long-term monitoring scheme is provided for future disease course management based on a disease prediction result. According to the method, an improved regression analysis method is provided, and a nonlinear term, an interaction effect and a dynamic adjustment mechanism are introduced, so that the model can better capture a nonlinear complex relationship in clinical data, and the prediction precision is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and particularly to a method and system for evaluating the severity of ulcerative colitis. Background Art

[0002] Ulcerative Colitis (UC) is a common chronic inflammatory bowel disease, characterized by repeated inflammation of the colon and rectum. Clinically, the evaluation of the severity of UC is crucial for formulating personalized treatment plans, predicting the progression of the disease, and improving the quality of life of patients. Currently, the methods for evaluating the severity of ulcerative colitis mainly rely on the subjective assessment of clinical symptoms, the detection results of biomarkers, and imaging examinations. Traditional methods evaluate the disease activity through simple clinical scoring systems (such as the Mayo score, Harvey-Bradshaw index, etc.), but these methods have certain limitations, especially in terms of individual differences, disease complexity, and the accuracy of data processing. Ulcerative Colitis (UC) is a common chronic inflammatory bowel disease, characterized by repeated inflammation of the colon and rectum. Clinically, the evaluation of the severity of UC is crucial for formulating personalized treatment plans, predicting the progression of the disease, and improving the quality of life of patients. Currently, the methods for evaluating the severity of ulcerative colitis mainly rely on the subjective assessment of clinical symptoms, the detection results of biomarkers, and imaging examinations. Traditional methods evaluate the disease activity through simple clinical scoring systems (such as the Mayo score, Harvey-Bradshaw index, etc.), but these methods have certain limitations, especially in terms of individual differences, disease complexity, and the accuracy of data processing.

[0003] With the rapid development of computational medicine and artificial intelligence, more and more advanced methods have begun to be applied to disease evaluation. Under the existing technical framework, the multivariate regression analysis in step S2, the clinical course model in step S3, and the deep neural network in step S4 have become emerging research directions, which can comprehensively evaluate the patient's condition from multiple dimensions. However, although these methods have certain advantages over traditional scoring systems, they still face some important technical challenges and deficiencies.

[0004] In the prior art, multivariate regression analysis has limitations in evaluating the severity of ulcerative colitis in patients. The regression analysis method usually assumes a certain linear relationship between the patient's condition and multiple clinical variables (such as biomarkers, immune responses, etc.). Although this method can provide a relatively simple model, it fails to fully capture the complex non-linear relationships in clinical data. For example, the relationship between the severity of the disease and immune responses, inflammatory factors may be highly non-linear, which makes the regression analysis method based on linear assumptions have poor prediction performance in some cases. In addition, the performance of multivariate regression analysis is also easily limited by the amount of data. Especially when the sample data is insufficient, the generalization ability of the model is weak and overfitting is likely to occur. Summary of the Invention

[0005] In view of the deficiencies of the prior art, the present invention provides a method and system for evaluating the severity of ulcerative colitis to solve the problems raised in the above background art.

[0006] To achieve the above object, the present invention provides the following technical solutions: In the first aspect, an embodiment of the present invention provides a method for evaluating the severity of ulcerative colitis, including the following steps: S1. Data collection and classification of ulcerative colitis symptoms; Collect the clinical data of patients and the symptom data related to ulcerative colitis to obtain clinical data; S2. Preliminary evaluation of the severity of ulcerative colitis based on clinical data; Based on the clinical data, establish a preliminary evaluation model based on multivariate regression analysis and calculate the severity score through the regression analysis method; S3. Conduct correlation analysis based on the preliminary evaluation model and the severity score; Perform in-depth analysis through the clinical course model to obtain the analysis result; S4. Perform intelligent diagnosis by combining the severity score with real-time physiological monitoring data; Monitor the physiological data through a physiological monitor, and combine the physiological data with the severity score through a deep neural network to construct an intelligent diagnosis model to obtain an intelligent diagnosis result; S5. Recommend a personalized treatment plan based on the intelligent diagnosis result; The decision support system generates a personalized treatment plan; S6. Evaluate the patient's recovery progress using the personalized treatment plan; Track the patient's physiological data based on a wearable device to evaluate the patient's recovery progress and obtain recovery progress data; S7. Perform subsequent disease prediction based on the recovery progress data; Based on the recovery progress data, predict future complications or recurrence conditions using the Support Vector Machine (SVM) algorithm to obtain disease prediction results; S8. Provide a long-term monitoring plan for future disease course management based on the disease prediction results; Through Internet of Things technology, combine patients' self-monitoring with doctors' remote monitoring.

[0007] Further optimize this technical solution. The clinical data and symptom data related to ulcerative colitis in step S1 include: The patient's age, gender, medical history, Body Mass Index (BMI), intestinal inflammation area, bowel movement frequency, blood in stool condition, abdominal pain degree, fever degree, C-reactive protein, and white blood cell count.

[0008] Further optimize this technical solution. The multivariate regression analysis in step S2 includes: Multivariate regression model: ; Wherein, n: The type of clinical data, which is 9; : The severity score of the patient's ulcerative colitis; : The constant term, representing the basic bias of the model; : The th item of clinical data. This coefficient will be adaptively adjusted according to the patient's condition and dynamically updated through machine learning methods. With the feedback of clinical data, Optimize to make the impact of important symptoms and markers on the score more significant; : Represents the non-linear mapping of the th clinical symptom or biomarker. It is modeled through a high-order polynomial; : Represents the sum of the patient's basic biomarkers including C-reactive protein and white blood cell count; : The th non-linear effect coefficient of the variable, responsible for capturing the high-order effects of each variable, including square and cube, on the severity; : The non-linear effect of the rd variable, including C-reactive protein or white blood cell count; m: The number of, representing the types of clinical data related to colitis, including C-reactive protein and white blood cell count, which is 2.

[0009] To further optimize this technical solution, in step S2, the regression analysis method includes: First, calculate the initial ulcerative colitis severity score of the patient through multivariate regression analysis , and use the regression analysis method to construct the ulcerative colitis severity score of the patient and the transformation model of the historical ulcerative colitis severity score of this patient : ; Among them, : represents the severity score at time point ; : represents the severity score at time point ; : the initial severity score calculated by the multivariate regression model; : the coefficient that controls the rate of change of the score. This coefficient reflects the relationship between the clinical severity score and the course of the disease, and is fitted through historical data or clinical experience and set artificially; : the saturation point of the disease, indicating that after the condition reaches a certain critical value, further deterioration will be inhibited, and it is obtained through clinical experience; : represents the dynamic adjustment coefficient related to time, which controls the rate of change of the score in the time series and is obtained through an optimization algorithm; : is a dynamic adjustment function, which is calculated based on the patient's past severity score and the time step to simulate the evolution behavior of the disease at different time points and reflect the impact of changes in external factors, treatment interventions, or clinical data on the course of the disease.

[0010] To further optimize this technical solution, the clinical course model in step S3 includes: Clinical course model formula: ; Among them, : represents the disease severity at time i.e., the current course state; : represents the disease severity at time ; : the constant that affects the disease progression speed, which reflects the promoting effect of the severity score on the course of the disease; : A non-linear mapping function based on the severity score, which reflects how the severity affects the disease progression. A sigmoid function is used to fit its non-linear relationship; : The coefficient that controls the non-linear effect, which regulates the direct impact of the severity on the disease progression; : It reflects the influence of external environment or physiological factors on the disease progression, including immune response, treatment methods; a: The coefficient that regulates the self-aggravation of the disease; : The square term representing the self-aggravation effect of the disease, which captures the possible acceleration effect in the disease progression.

[0011] To further optimize this technical solution, the deep neural network in step S4 includes: Based on real-time physiological monitoring data including heart rate, blood pressure, and blood oxygen level, an intelligent diagnosis model is constructed. This model performs repetitive calculations on the physiological monitoring data until the association between the three data is found.

[0012] To further optimize this technical solution, the decision support system in step S5 includes: Combining the patient's severity score, clinical symptoms, medical history, and real-time physiological data, recommend drugs, treatment methods, and lifestyle to develop a personalized treatment plan.

[0013] To further optimize this technical solution, the physiological data in step S6 includes: The patient's body temperature, blood oxygen, and frequency of bloody stools.

[0014] To further optimize this technical solution, the support vector machine (SVM) model algorithm in step S7 includes: Combining the patient's recovery data, treatment response, and clinical characteristics, and performing non-linear operations to help predict the risks in the long-term disease progression, optimize the patient's treatment path; predict possible future complications or recurrence of the disease, allowing doctors to intervene in advance and adjust the treatment plan to reduce risks.

[0015] To further optimize this technical solution, the ulcerative colitis severity assessment system is used to implement the ulcerative colitis severity assessment method. The functional modules of this system include a data collection and preprocessing module, a multivariate regression analysis module, a clinical disease progression model module, a deep neural network prediction module, a personalized disease progression prediction module, a multi-task learning and optimization module, a model feedback and adjustment module, and a report generation and visualization module.

[0016] Second aspect, an embodiment of the present invention provides a computer device, including a memory and a processor, where the memory stores a computer program, and: when the computer program instructions are executed by the processor, the steps of a method and system for evaluating the severity of ulcerative colitis as described in the first aspect of the present invention are implemented.

[0017] Third aspect, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored, and: when the computer program instructions are executed by the processor, the steps of a method and system for evaluating the severity of ulcerative colitis as described in the first aspect of the present invention are implemented.

[0018] Compared with the prior art, the present invention provides a method and system for evaluating the severity of ulcerative colitis, having the following beneficial effects: The method and system for evaluating the severity of ulcerative colitis, by proposing an improved regression analysis method, introducing non-linear terms, interaction effects and a dynamic adjustment mechanism, enables the model to better capture the non-linear complex relationships in clinical data. For example, by combining regression analysis with a clinical course model, time dynamic factors can be added to the regression model to gradually update the model parameters and improve the adaptability to the course changes of individual patients. This improved method can overcome the neglect of non-linear relationships by traditional regression analysis models and significantly improve the prediction accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0020] Figure 1 It is a schematic flowchart of a method for evaluating the severity of ulcerative colitis proposed by the present invention; Figure 2 It is a schematic diagram of the intelligent diagnosis process in a method for evaluating the severity of ulcerative colitis proposed by the present invention; Figure 3 It is a schematic diagram of the Internet of Things technology of a method and system for evaluating the severity of ulcerative colitis proposed by the present invention; Figure 4 It is a schematic diagram of the modules of a system for evaluating the severity of ulcerative colitis proposed by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0021] To make the above objects, features and advantages of the present invention more obvious and understandable, the following will give a detailed description of the specific embodiments of the present invention with reference to the drawings in the specification.

[0022] In the following description, many specific details are set forth in order to provide a thorough understanding of the present invention. However, the present invention may be practiced in other ways different from those described herein. Those skilled in the art can make similar generalizations without departing from the spirit of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0023] Secondly, as used herein, an "embodiment" or "embodiments" refers to specific features, structures, or characteristics that may be included in at least one implementation of the present invention. The phrase "in an embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it an embodiment that is separate or mutually exclusive of other embodiments. Embodiment 1

[0024] Referring to Figures 1 to 3 , which is the first embodiment of the present invention, this embodiment provides a method for evaluating the severity of ulcerative colitis, including the following steps: S1. Data collection and classification of ulcerative colitis symptoms; Collect the clinical data of patients and the symptom data related to ulcerative colitis to obtain the clinical data. The clinical data and the symptom data related to ulcerative colitis include: The age, gender, medical history, body mass index (BMI), intestinal inflammation area, bowel movement frequency, blood in the stool condition, abdominal pain degree, fever degree, C-reactive protein, and white blood cell count of the patient.

[0025] In this embodiment, non-invasive in vitro diagnostic techniques (such as fecal occult blood test) are used to obtain the inflammatory markers (such as C-reactive protein and white blood cell count) of the patient. Data collection is the basis of the entire system, which provides data support for subsequent severity evaluation. All the collected data will be used as input to the evaluation model. S2. Preliminary evaluation of the severity of ulcerative colitis based on clinical data; Based on the clinical data, a preliminary evaluation model is established through multivariate regression analysis, and the severity score is calculated by the regression analysis method.

[0026] The multivariate regression analysis includes: Multivariate regression model: ; Wherein, n: the type of clinical data, which is 9; : the severity score of the patient's ulcerative colitis; : the constant term, representing the basic bias of the model; : the For a certain item of clinical data, this coefficient will be adaptively adjusted as the patient's condition changes, dynamically updated through machine learning methods, and with the feedback of clinical data, Optimized to make the impact of important symptoms and markers on the score more significant; : Represents the non - linear mapping of the th clinical symptom or biomarker. Modeled by a high - order polynomial; : Represents the sum of the patient's basic biomarkers including C - reactive protein and white blood cell count; : The th non - linear effect coefficient of the variable, responsible for capturing the high - order effects of each variable, including square and cube, on the severity; : The th variable including the non - linear effect of C - reactive protein or white blood cell count; m: The number of, representing the types of clinical data related to colitis including C - reactive protein and white blood cell count, which is 2.

[0027] In this embodiment, the usage of the formula is explained as: Dynamic weighting coefficient ( ): For each symptom or biomarker, the model dynamically adjusts the weight according to its impact on the severity of ulcerative colitis. For example, in the initial stage of the patient, the hematochezia is mild, which may have a relatively small impact on the disease severity score. However, as the condition deteriorates, the weight of hematochezia will increase. Through machine learning algorithms such as reinforcement learning or online learning models, will be continuously updated according to the patient's feedback information to optimize the scoring system.

[0028] Non - linear mapping ( ): Since the relationship between clinical symptoms and biomarkers is often not simply linear, this model performs non - linear transformation on each input through high - order polynomial regression or **Support Vector Regression (SVR)**. For example, the change in the concentration of C - reactive protein may show a more significant change near certain critical values in its impact on the disease condition. Traditional linear regression cannot capture this change, while using non - linear mapping can effectively model this complex relationship.

[0029] Non - linear effect term ( ): Changes in some symptoms or markers may exhibit higher-order non-linear effects. For example, the concentration of C-reactive protein has a relatively small impact on ulcerative colitis at low levels, but as the concentration increases, its impact on the disease becomes more significant and may even trigger severe pathological reactions. By introducing quadratic and cubic terms, such non-linear relationships can be effectively captured.

[0030] The formula is used as follows: Data input: In clinical practice, doctors or systems will use the clinical data input of patients (such as symptom data like bloody stools and diarrhea, and biomarker data like C-reactive protein and white blood cell count) as values.

[0031] Dynamic weight adjustment: Based on real-time clinical observations and symptom changes, the model will adjust , for example, if the bloody stools symptom of a patient becomes more severe, the weight of bloody stools will increase, thus affecting the severity score.

[0032] Non-linear mapping processing: Each input clinical symptom or biomarker will be mapped to an appropriate influence through a non-linear function (such as support vector regression), reflecting its true impact on the severity of ulcerative colitis.

[0033] Output score: The final model will give a comprehensive severity score , which takes into account the dynamic weighted influence of each symptom and biomarker and their non-linear relationships, helping doctors or decision support systems evaluate the current disease state of ulcerative colitis.

[0034] The regression analysis methods include: First, calculate the initial severity score of the patient's ulcerative colitis through multivariate regression analysis , and construct a conversion model of the patient's ulcerative colitis severity score with the patient's historical ulcerative colitis severity score : ; Among them, : represents the severity score at time point ; : represents the severity score at time point ; : the initial severity score calculated by the multivariate regression model; : The coefficient that controls the rate of change of the score. This coefficient reflects the relationship between the clinical severity score and the course of the disease, and is artificially set through fitting with historical data or clinical experience; Historical data regression analysis: By collecting a large amount of clinical data, especially the severity scores of patients at different time points (for example, clinical symptom scores, laboratory test results, etc.), regression analysis methods (such as the least squares method) can be used to estimate ; Clinical observation and expert experience: In the absence of sufficient data, it can be speculated through the observation of the course of the disease by clinical experts, combined with existing medical literature . In practical applications, the "online learning" method in machine learning can be used, that is, as more patient data accumulates, is adjusted in real time. For example, when new patient data indicates that the disease condition changes slowly, can be appropriately reduced.

[0035] : The saturation point of the disease, indicating that after the disease condition reaches a certain critical value, further deterioration will be inhibited, and it is obtained through clinical experience; Clinical data analysis: According to the course data of a large number of patients, observe that when the disease severity score reaches a certain specific value, the change of the course of the disease begins to slow down or tend to be stable. This value is the estimated value of . For example, in the disease deterioration stage, the scores of some patients may approach a stable value (such as the maximum tolerable level of the "high-risk" group). Logistic growth model: By applying the Logistic growth model to clinical data, the saturation point in the process of disease development can be fitted. The Logistic growth model is usually used to simulate the growth of population or disease, and its formula is: , where is the "carrying capacity" or "saturation point". By fitting the clinical data, the value of can be estimated. e is the base of the natural logarithm; represents the time offset, indicating the time point when the curve starts to grow or the midpoint of the curve, that is, the value of t when ; r is the growth rate, which determines the growth speed of the curve, that is, approaches the saturation point at a certain rate; represents the variable value of time t, which can represent the disease burden at time t or the level of a certain biomarker. In the absence of sufficient data, experts can speculate on the value of based on their understanding of the disease, combined with case studies in the literature.

[0036] : Represents a time-related dynamic adjustment coefficient that controls the rate of change of scores in a time series and is obtained through an optimization algorithm; Clinical data and treatment interventions: If the patient's condition improves or deteriorates during the treatment process, the coefficient will reflect the impact of these external interventions. For example, the improvement of the patient's condition after treatment may lead to an acceleration or deceleration of the rate of change of the score. By comparing the clinical data before and after treatment, the value of can be estimated; Dynamic adjustment and machine learning: In practical applications, reinforcement learning (RL) methods in machine learning can be used to dynamically adjust the value. By establishing a reward mechanism, the system can adjust in real time to adapt to the changes in the patient's condition. For example, in some stages, drug treatment or lifestyle changes may significantly accelerate the improvement of the condition, and the value of will increase in these stages; Model calibration: Similar to ,

[0037] can also be obtained through regression analysis or optimization algorithms. By repeatedly testing and adjusting the difference between the model output and the actual data, and the time step to calculate, simulating the evolution behavior of the disease at different time points, reflecting the impact of changes in external factors, treatment interventions or clinical data on the course of the disease.

[0038] In this embodiment, the analysis of the formula is as follows: Dynamic change of score: The first term in the formula reflects the change in the disease severity score from the previous time point to the current time point . This change is based on the difference between the initial score obtained through regression analysis and the historical score . The rate of this change process is controlled by the coefficient to ensure that the change conforms to the regular course of the disease clinically.

[0039] Nonlinear saturation effect: The second term introduces the saturation effect of the disease course, indicating that as time goes by and the severity increases, the condition will gradually enter a "stable" state. When approaches , the rate of deterioration of the condition will slow down, which is consistent with the phenomenon that some patients may reach a stable state in clinical practice.

[0040] Time-dependent adjustment (dynamic adjustment function): The third term Allows the model to adjust the score according to the change of time. The dynamic adjustment function here May take into account changes in external environment, patient treatment intervention, lifestyle and other factors. For example, if a patient receives drug treatment, the condition may be alleviated, resulting in a dynamic adjustment of the score. This function can simulate the non-linear changes of the condition at different time points and enhance the adaptability of the model.

[0041] S3. Conduct correlation analysis based on the preliminary evaluation model and the severity score; Conduct in-depth analysis through the clinical course model to obtain the analysis results.

[0042] The clinical course model includes: Clinical course model formula: ; Among them, : Represents the disease severity at time That is, the current disease course state; : Represents the disease severity at time ; : A constant affecting the disease progression speed, reflecting the promoting effect of the severity score on the disease course; : A non-linear mapping function based on the severity score, reflecting how the severity affects the disease course progression, and using the sigmoid function to fit its non-linear relationship; : A coefficient controlling the non-linear effect, adjusting the direct effect of the severity on the disease course; : Reflects the influence of external environment or physiological factors on the disease course, including immune response, treatment methods; a: A coefficient adjusting the self-aggravation of the disease; : Represents the square term of the disease self-aggravation effect, capturing the possible acceleration effect in the disease course. For example, the accumulation of certain symptoms may lead to a rapid deterioration of the condition after a certain critical point, and the quadratic term in the model can reflect this non-linear feedback.

[0043] In this embodiment, the interpretation of the formula is: Time recursion: This model is a dynamic system model, reflecting the evolution of the disease in the time dimension. The disease severity at each time step ( ) is determined by the condition at the previous moment ( ) and the current severity score ( Co - decision. This recursive relationship enables the progression of the disease to be dynamically adjusted over time and with the changes in the disease condition.

[0044] Feedback mechanism: The two - way feedback mechanism in the model is crucial. The current severity score ( ), which affects the rate of change of the disease course (controlled by and ), while the current disease - course state ( ), affects the self - exacerbation of the disease (controlled by the term). This feedback mechanism can reflect the non - linear effects in the disease course, i.e., the deterioration of the disease condition may accelerate, especially near certain critical points.

[0045] Non - linear effects ( and ): In this model, we adopted non - linear functions to capture the complex relationship between the severity score and the disease - course progression. The sigmoid function or higher - order polynomials can simulate the disease - course changes of ulcerative colitis at different severities. For example, when the severity score is higher than a certain critical value, the disease condition may exacerbate exponentially, while at low severity, it may remain stable.

[0046] Saturation effect of the disease course ( ): The

[0047] term reflects the "saturation" effect of the disease. As the disease course progresses, the rate of disease deterioration gradually slows down. When the severity score or disease progression approaches the saturation point, further deterioration of the disease condition will be inhibited. This is in line with clinical experience, that is, some patients may enter a relatively stable stage after experiencing an initial exacerbation of the disease. The actual application method of the formula is as follows: Predicting the disease course: By inputting the real - time severity score ( ), the clinical disease - course model can predict the evolution of the patient's disease course in the future for a certain period. For example, if the patient's severity score changes from mild blood in the stool ( lower) to moderate abdominal pain ( higher), the model will predict the exacerbation trend of the disease condition according to this change, helping doctors adjust the treatment plan.

[0048] Real - time adjustment: This model can be used in combination with real - time clinical data. As the patient's disease condition changes, the input data is continuously updated. In this way, doctors can keep track of the patient's disease progression at any time and adjust the treatment method according to the prediction results.

[0049] Personalized treatment optimization: By combining this clinical course model with the severity assessment in step S2, doctors can further optimize the patient's treatment path based on the output of the course model and personalized treatment plans. For example, when the model predicts a worsening trend in the condition, it may recommend increasing the drug dosage or introducing new treatment methods.

[0050] S4. Perform intelligent diagnosis by combining the severity score with real-time physiological monitoring data; Monitor physiological data through a physiological monitor, and combine the physiological data with the severity score through a deep neural network to construct an intelligent diagnosis model and obtain an intelligent diagnosis result.

[0051] The deep neural network includes: Based on real-time physiological monitoring data including heart rate, blood pressure, and blood oxygen level, construct an intelligent diagnosis model that repeatedly calculates the physiological monitoring data until the relationship between the three data is found.

[0052] In this embodiment, the formula of the deep neural network model is: .

[0053] Among them, represents the prediction result at time t, including multiple prediction tasks such as the severity of the disease, immune response, and treatment effect.

[0054] represents the clinical input feature vector at time t (such as biomarkers, imaging features, etc.).

[0055] are the weight matrices of the neural network, corresponding to different hidden layers.

[0056] are the bias vectors of the hidden layers of the neural network respectively.

[0057] is the rectified linear unit (ReLU) activation function, used to introduce non-linearity.

[0058] The activation function of the output layer is usually used to map the output of the model to the predicted value (such as Softmax for classification tasks and Sigmoid for regression tasks).

[0059] The explanation of the formula is: Fusion of the input layer and clinical data: The input layer includes two parts of information: one part is the clinical data characteristics of the patient (such as hematological indicators, immune markers, imaging features, etc.), and the other part is the dynamic severity score from step S3 , representing the patient's current condition.

[0060] These two parts of data are concatenated ( ) as the input to the network, forming a comprehensive feature vector to ensure that the model considers both the patient's clinical data and the dynamic evolution of the disease course.

[0061] Multi-layer non-linear feature extraction: In the hidden layer of the neural network, the ReLU activation function is used to perform non-linear mapping on the input, which can effectively capture complex non-linear relationships. The multi-layer structure enables the network to gradually extract higher-order features, helping the model learn deeper disease course and disease development features.

[0062] Dynamic feedback mechanism: The clinical course model in step S3 provides dynamic feedback input to the neural network through the given severity score , enabling the model to adjust the prediction of the patient's disease course in real time. For example, if the patient's condition worsens, the severity score output by the disease course model will be used to dynamically adjust the prediction of the neural network, thus accurately capturing the evolution of the disease course.

[0063] Multi-task learning: The output layer not only predicts the severity of the patient (i.e., predicts the current state of the disease), but can also simultaneously predict other tasks, such as immune response, treatment effect, etc. Through sharing the hidden layer representation, the information between these tasks can promote each other, improving the generalization ability and prediction accuracy of the model.

[0064] Optimization and loss function: Use loss functions with regularization (such as cross-entropy loss function, mean squared error loss function, etc.) to train the network. The loss function comprehensively considers the error between the model output and the true label, and optimizes the weights and biases of the neural network through the backpropagation algorithm, thereby minimizing the error and optimizing the prediction performance of the model.

[0065] An example of using the formula is as follows: Suppose there is a patient's clinical data at time and the severity score obtained through step S3 . These data can be input into the deep neural network, and the prediction result can be obtained through forward propagation calculation. This result may include the outputs of multiple prediction tasks, such as: Predicting severity: , indicating the current disease severity of the patient.

[0066] Immune response prediction: , indicating the current status or response of the patient's immune system.

[0067] Treatment effect prediction: , indicating the possible effects after treatment.

[0068] In this way, the deep neural network can dynamically combine the progress of the clinical course while processing complex clinical data, making more accurate and personalized course predictions.

[0069] S5. Recommend a personalized treatment plan based on the intelligent diagnosis results; The decision support system generates a personalized treatment plan.

[0070] The decision support system includes: Combining the patient's severity score, clinical symptoms, medical history, and real-time physiological data, recommend drugs, treatment methods, and lifestyle to develop a personalized treatment plan.

[0071] In this embodiment, the decision support system provides real-time diagnostic advice, treatment plan recommendations, and risk assessments for doctors by integrating information such as the patient's medical record data, medical knowledge base, and clinical guidelines. This includes but is not limited to drug selection, dose adjustment, surgical methods, lifestyle adjustments, and the use of immunosuppressive therapies or biologics. The system can provide customized diagnosis and treatment advice based on the patient's individual information, such as the condition, medical history, examination results, etc. This personalized treatment plan helps to improve the treatment effect, reduce adverse reactions, and also saves treatment costs and time for the patient. In addition, CDSS can also provide personalized advice in drug treatment, helping doctors select appropriate drugs and doses, predict drug metabolism and side effects, thereby optimizing the drug treatment plan.

[0072] S6. Evaluate the patient's recovery progress using the personalized treatment plan; Track the patient's physiological data based on the wearable device to evaluate the patient's recovery progress and obtain recovery progress data.

[0073] The physiological data includes: The patient's body temperature, blood oxygen, and frequency of bloody stools.

[0074] In this embodiment, a wearable device is used to evaluate the recovery progress of patients. By continuously tracking physiological data such as the patient's body temperature, blood oxygen, and frequency of bloody stools, the response to treatment is evaluated. Combining with the biomarker levels of the disease, the intelligent system can evaluate the patient's recovery progress and adjust the treatment plan according to the changes. Through this method, the situation of lagging treatment plan or insignificant treatment effect can be effectively avoided.

[0075] S7. Perform subsequent disease prediction based on the recovery progress data; Based on the recovery progress data, predict future complications or recurrence conditions based on the support vector machine (SVM) algorithm to obtain disease prediction results.

[0076] The support vector machine (SVM) model algorithm includes: Combining the patient's recovery data, treatment response, and clinical characteristics, and performing non-linear operations to help predict risks in the long-term course of the disease and optimize the patient's treatment path; predicting possible future complications or recurrence of the condition, allowing doctors to intervene in advance and adjust the treatment plan to reduce risks.

[0077] In this embodiment, the SVM maps the recovery data, treatment response, and clinical characteristics to a high-dimensional space through the kernel trick. In this space, the data may become easier to separate. The kernel function allows the SVM to implicitly perform operations in the high-dimensional space without explicitly calculating the coordinates in the high-dimensional space.

[0078] Commonly used kernel functions include linear kernel, polynomial kernel, radial basis function (RBF) kernel, and sigmoid kernel.

[0079] S8. Provide a long-term monitoring plan for future disease course management based on the disease prediction results; Through the Internet of Things technology, the combination of patient self-monitoring and doctor remote monitoring is realized.

[0080] In this embodiment, the Internet of Things-based intelligent device worn by the patient can transmit data to the cloud in real time. Through the intelligent system analysis, the health risks are predicted, and the doctor adjusts the treatment plan in a timely manner according to the data feedback, so as to realize more refined and personalized long-term disease course management. Embodiment Two

[0081] Please refer to Figure 4 , the functional modules of the ulcerative colitis severity assessment system include a data acquisition and preprocessing module, a multivariate regression analysis module, a clinical course model module, a deep neural network prediction module, a personalized disease course prediction module, a multi-task learning and optimization module, a model feedback and adjustment module, and a report generation and visualization module.

[0082] In this embodiment, the clinical data, historical medical records, laboratory test results, imaging data, etc. of the patient are collected by the data acquisition and preprocessing module and preprocessed; the multivariate regression analysis module uses the multivariate regression analysis method to analyze the clinical data of the patient and evaluate the current severity of the disease. By considering the relationships between multiple clinical factors (such as inflammatory factors, immune responses, etc.), a severity score is generated. ; through the clinical course model module, based on the time recurrence formula, the disease progression of the patient is modeled to predict the dynamic evolution of the disease. According to and time-related factors, the disease course prediction of the patient is dynamically adjusted; through the deep neural network prediction module, the input from multi-source data (such as patient clinical data, historical disease courses, laboratory test results, etc.) is processed to further predict the disease development of the patient; through the personalized disease course prediction module, the outputs of multivariate regression analysis, clinical course model, and deep neural network are combined to provide a personalized disease course prediction for each patient; through the multi-task learning and optimization module, which is based on multi-task learning and deep neural network methods to solve multiple related tasks simultaneously; through the model feedback and adjustment module, the model parameters of regression analysis, clinical course model, and deep neural network are adjusted and optimized in real time; through the report generation and visualization module, a comprehensive disease course assessment report is generated, including information such as disease severity score, immune response, treatment effect evaluation, etc., and an easy-to-understand visualization interface is provided, such as disease course progression chart, treatment effect curve, etc. Embodiment III

[0083] This embodiment also provides a computer device applicable to a method and system for evaluating the severity of ulcerative colitis, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement a method and system for evaluating the severity of ulcerative colitis as proposed in the above embodiment.

[0084] This embodiment also provides a storage medium, on which a computer program is stored. When the program is executed by the processor, it implements a method and system for evaluating the severity of ulcerative colitis as proposed in the above embodiment.

[0085] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a carrier network, NFC (Near Field Communication), or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or may also be a button, a trackball, or a touchpad provided on the housing of the computer device, or may also be an external keyboard, touchpad, or mouse, etc.

[0086] If a function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The foregoing storage medium includes: various media such as a USB flash drive, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk, or an optical disc that can store program codes.

[0087] The logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a definite sequence list of executable instructions for implementing a logical function, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch instructions from the instruction execution system, apparatus, or device and execute the instructions), or used in combination with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.

[0088] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection (electronic device) having one or more wirings, a portable computer diskette (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable media can even be paper or other suitable media on which a program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or otherwise processing it as appropriate, and then storing it in a computer memory.

[0089] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, the multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0090] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A method for assessing the severity of ulcerative colitis, characterized in that: The following steps are involved: S1. Data collection and classification of ulcerative colitis symptoms; Clinical data of patients and symptom data related to ulcerative colitis were collected to obtain clinical data; S2. Perform a preliminary assessment of the severity of ulcerative colitis based on clinical data; According to clinical data, a preliminary assessment model was established based on multivariate regression analysis, and the severity score was calculated by regression analysis method; S3, correlation analysis based on the preliminary assessment model and severity score; Conduct in-depth analysis through the clinical course model to obtain analysis results; S4, intelligent diagnosis based on severity score combined with real-time physiological monitoring data; Monitor physiological data through physiological monitors, and combine physiological data with severity scores through deep neural networks to build intelligent diagnosis models and obtain intelligent diagnosis results; S5. Recommend personalized treatment plans based on intelligent diagnosis results; Based on the intelligent diagnosis results, the decision support system generates personalized treatment plans; S6. Assess the patient's recovery progress using personalized treatment plans; Track the patient's physiological data based on wearable devices, evaluate the patient's recovery progress, and obtain recovery progress data; S7. Predict the subsequent condition based on the recovery progress data; According to the recovery progress data, the future complications or recurrences are predicted based on the support vector machine (SVM) algorithm to obtain the disease prediction results; S8. Provide long-term monitoring solutions for future disease management based on disease prediction results; Through the Internet of Things technology, the combination of patient self-monitoring and doctor remote monitoring can be achieved.

2. A method for evaluating the severity of ulcerative colitis according to claim 1, characterized in that: The clinical data and symptom data related to ulcerative colitis in step S1 include: The patient's age, gender, medical history, body mass index (BMI), intestinal inflammation area, bowel movement frequency, blood in the stool, abdominal pain severity, fever severity, C-reactive protein, and white blood cell count.

3. A method for evaluating the severity of ulcerative colitis according to claim 1, characterized in that: The multivariate regression analysis in step S2 includes: Multivariate regression model: ； in, n: type of clinical data, which is 9; : Patients' ulcerative colitis severity score; : Constant term, representing the basic bias of the model; : No. This coefficient will be adaptively adjusted as the patient's condition changes and dynamically updated through machine learning methods. Optimization, so that important symptoms and markers have a more significant impact on the score; :Indicates the The nonlinear mapping of clinical symptoms or biomarkers is modeled by high-order polynomials; : It represents the sum of the patient's basic biomarkers including C-reactive protein and white blood cell count; : No. The nonlinear effect coefficient of the item variable is responsible for capturing the high-order effects of each variable on severity, including square and cube; : No. The nonlinear effects of the variables included C-reactive protein or white blood cell count; m: The number, indicating the types of clinical data related to colitis including C-reactive protein and white blood cell count, is 2.

4. A method for evaluating the severity of ulcerative colitis according to claim 1, characterized in that: In step S2, the regression analysis method includes: The initial ulcerative colitis severity score was calculated by multivariate regression analysis. , using regression analysis to construct a severity score for ulcerative colitis in patients The severity score of ulcerative colitis in relation to the patient's history Conversion model: ； in, : Indicates a time point severity score; : Indicates a time point severity score; : Initial severity scores calculated by multivariate regression model; : The coefficient that controls the rate of change of the score. This coefficient reflects the relationship between the clinical severity score and the course of the disease. It is fitted through historical data or clinical experience and is set artificially. : The saturation point of the disease, which means that after the disease reaches a certain critical value, further deterioration will be suppressed, which is derived from clinical experience; : represents the time-related dynamic adjustment coefficient, which controls the rate of change of the score in the time series and is obtained through the optimization algorithm; : A dynamically adjusted function based on the patient's past severity scores and time step To calculate and simulate the evolutionary behavior of the disease at different time points, reflecting the impact of external factors, therapeutic interventions or changes in clinical data on the course of the disease.

5. A method for evaluating the severity of ulcerative colitis according to claim 1, characterized in that: The clinical course model in step S3 includes: Clinical course model formula: ； in, : Indicates the time The severity of the disease at a moment is the current disease course; : Indicates the time The severity of the disease at the moment; : A constant that affects the rate of disease progression, reflecting the role of severity score in promoting the course of the disease; : A nonlinear mapping function based on severity score reflects how severity affects disease progression, and the sigmoid function is used to fit its nonlinear relationship; : The coefficient that controls the nonlinear effect and adjusts the direct effect of severity on the course of the disease; : Reflects the impact of external environmental or physiological factors on the course of the disease, including immune response and treatment methods; a: coefficient regulating the self-aggrandizement of the disease; : represents the square term of the self-aggrandizing effect of the disease, capturing the possible acceleration effect in the course of the disease.

6. A method for evaluating the severity of ulcerative colitis according to claim 1, characterized in that: The deep neural network in step S4 includes: Based on real-time physiological monitoring data including heart rate, blood pressure, and blood oxygen level, an intelligent diagnosis model is constructed. The model performs repetitive calculations on the physiological monitoring data until the correlation between the three data is found.

7. A method for evaluating the severity of ulcerative colitis according to claim 1, characterized in that: The decision support system in step S5 includes: Based on the patient's severity score, clinical symptoms, medical history and real-time physiological data, we recommend medications, treatment methods and lifestyle to develop a personalized treatment plan.

8. A method for evaluating the severity of ulcerative colitis according to claim 1, characterized in that: The physiological data in step S6 includes: The patient's body temperature, blood oxygen, and frequency of blood in stool.

9. A method for evaluating the severity of ulcerative colitis according to claim 1, characterized in that: The support vector machine (SVM) model algorithm in step S7 includes: Combining the patient's recovery data, treatment response, and clinical characteristics, and performing nonlinear calculations, it helps predict risks in the long term and optimize the patient's treatment path. It also predicts possible future complications or recurrences, allowing doctors to intervene in advance and adjust treatment plans to reduce risks.

10. A system for evaluating the severity of ulcerative colitis, used to implement a method for evaluating the severity of ulcerative colitis according to any one of claims 1 to 9, characterized in that: The functional modules of the system include data acquisition and preprocessing module, multivariate regression analysis module, clinical course model module, deep neural network prediction module, personalized course prediction module, multi-task learning and optimization module, model feedback and adjustment module, and report generation and visualization module.

Images