A symptom intelligent assessment system for a decompensated cirrhosis patient

Abstract

The application discloses a kind of symptoms intelligent evaluation systems of decompensated cirrhosis patient, it is related to medical information processing technical field;System is based on the construction of unpleasant symptom theory, specifically includes: data acquisition module, symptom group analysis module, evaluation grading module, intervention suggestion module.The application quantifies symptom experience from three dimensions of frequency, severity and distress degree, avoids the deviation of only focusing on symptom intensity and ignoring frequency or distress degree, and more accurately reflects the subjective feelings of patients;Based on the unpleasant symptom theory, through correlation analysis and factor analysis, ascites-related, encephalopathy with pain, psychological symptom group and other clusters are automatically identified, revealing the synergistic effect between symptoms and common pathophysiological mechanisms, providing cluster intervention basis for clinical, with higher intervention efficiency.

Description

Technical Field

[0001] This invention relates to the field of medical information processing technology, and in particular to an intelligent symptom assessment system for patients in the decompensated stage of cirrhosis. Background Technology

[0002] Cirrhosis is the end stage of many chronic liver diseases, characterized by diffuse liver fibrosis, regenerative nodules, and pseudolobule formation. It is a leading cause of death from liver-related diseases worldwide. Epidemiological data shows that cirrhosis and related chronic liver diseases cause approximately one million deaths annually, accounting for 2% of all deaths globally, ranking 11th among the leading causes of death worldwide. In my country, the main causes of cirrhosis include hepatitis B virus infection, alcoholic liver disease, and non-alcoholic fatty liver disease. Cirrhosis not only reduces the average life expectancy of patients by about 0.10 years, but its morbidity and mortality rates are also increasing year by year with population aging. In 1990, there were approximately 7 million patients in my country, which increased to nearly 12 million in 2016, imposing a heavy medical and economic burden on society and patients.

[0003] The course of cirrhosis is divided into compensated and decompensated stages. Compensated patients often experience mild or no typical symptoms, while decompensated patients, due to decreased liver function and portal hypertension, develop complications such as ascites, esophageal and gastric variceal bleeding, hepatic encephalopathy, and infections, with a mortality risk nearly 10 times higher than in the compensated stage. In addition to typical complications, patients often experience multiple symptoms including fatigue, abdominal distension, sleep disturbances, and pain, which are characterized by prolonged duration and recurrent episodes, further exacerbating the symptom burden and suffering. Studies show that 92.5% of decompensated patients experience fatigue, and 75.3% experience sleep disturbances. Furthermore, symptoms exhibit clustering characteristics (such as encephalopathy with pain symptom clusters, ascites-related symptom clusters, and psychological symptom clusters). Physiological symptoms interact with psychological symptoms such as anxiety and depression, creating a vicious cycle that significantly reduces patients' quality of life and ability to cope with the disease.

[0004] Current clinical symptom management models are centered on single symptoms, making it difficult to reflect the true experience of multiple coexisting and synergistic symptoms. Symptom assessment tools mainly fall into two categories: single-symptom assessment scales (such as the Numerical Rating Scale for Pain and the Pittsburgh Sleep Quality Index) can only measure specific symptoms and cannot comprehensively assess multiple related symptoms; while general multi-symptom assessment scales (such as the Memory Symptom Assessment Scale and the Anderson Symptom Assessment Scale) cover multiple dimensions but lack disease-specific items (e.g., they do not include muscle cramps and weakness, which are common in cirrhosis, but include non-specific symptoms such as sweating and dysphagia), resulting in insufficient sensitivity and specificity in assessment. In addition, existing tools often ignore the multidimensional differences in the frequency, severity, and distress of symptoms (some patients have low symptom frequency but high distress), and do not provide individualized intervention suggestions based on factors such as patient demographics, disease characteristics, and self-management behaviors, thus limiting the improvement of symptom management efficiency.

[0005] The theory of unpleasant symptoms (TOUS) originated from Pugh and Gift's research on fatigue symptoms during childbirth and dyspnea symptoms in patients with chronic obstructive pulmonary disease. Through clinical observation, expert discussion, and literature review, the concept was first proposed in 1995 by Lenz's team, a Fellow of the American Academy of Nursing, and revised in 1997. It is one of the most commonly used theories in symptom research, and many studies on symptom management in cancer and chronic disease patients use it as a theoretical framework. This theory assesses the physical and psychological reactions patients may experience when faced with uncomfortable or unpleasant symptoms, including three core concepts: symptoms, influencing factors, and symptom outcomes.

[0006] Symptoms refer to the uncomfortable feelings that patients subjectively experience or perceive. Common symptoms include shortness of breath, fatigue, and pain, which are signals of changes in bodily function and threats to health. This theory emphasizes that symptoms include four dimensions: intensity, time, stress, and quality. The intensity dimension reflects the severity of symptoms and is a frequently observed dimension; the time dimension reflects the frequency, duration, and pattern of symptom occurrence; the stress dimension reflects the degree of unpleasant experience or distress caused by symptoms, and this dimension is closely related to quality of life and the patient's psychological state; the quality dimension is the patient's description of their symptom experience, reflecting the specific characteristics or nature of the symptoms. This study aims to construct a symptom cluster assessment scale for patients in the decompensated stage of cirrhosis. Therefore, dimensions that can directly reflect the symptom burden are prioritized, while the quality dimension mainly describes the nature of symptoms and is not easy to quantify. Therefore, the frequency, severity, and distress levels were ultimately selected. Furthermore, the theory indicates that symptoms often do not occur singly but have synergistic effects. The adverse effects they cause are not simply additive but multiplicative, exhibiting simultaneity and additive nature. The occurrence and manifestation of symptoms are influenced by three categories of factors: physiological (nutritional status, complications, etc.), psychological (emotions, cognition, and attitudes, etc.), and situational (lifestyle, social support, etc.). These factors interact with each other, collectively affecting symptoms. Therefore, even if different patients experience the same symptom, their symptom experience and outcome may vary greatly due to individual differences. Symptom outcome refers to the result or prognosis after a patient experiences symptoms, mainly including functional status and cognitive function. The former is reflected in daily life and social activities, while the latter is reflected in logical thinking and mental state, and is related to the number and severity of symptoms. The three core concepts of the unpleasant symptom theory are dynamically interconnected, systematically explaining the complete process of a patient's symptom experience.

[0007] The unpleasant symptom theory elucidates the mechanism of interaction between symptoms, emphasizes the importance of multidimensional symptom assessment, and considers influencing factors and symptom outcomes. Exploring its core concepts allows for a complete reflection of patients' symptom characteristics and overall experience, providing important theoretical support for related symptom research. Other common symptom theories, such as symptom management theory and symptom experience theory, primarily guide symptom management behaviors and strategies, while the latter tends to describe patients' perception and reactions to symptoms. Therefore, based on the purpose and significance of this study, the unpleasant symptom theory was chosen as the theoretical framework. The theoretical dimensions were transformed into three dimensions for symptom assessment in this scale: frequency of occurrence, severity, and degree of distress. Symptoms in patients with decompensated cirrhosis were assessed from these three dimensions, corresponding one-to-one with the three dimensions of time, intensity, and stress, respectively. A scientific and standardized scale was developed, and the influencing factors related to symptom clusters in patients with decompensated cirrhosis were explored.

[0008] In summary, there is an urgent need for an intelligent assessment system for patients with decompensated cirrhosis. This system should assess symptom burden from multiple dimensions, identify the intrinsic correlations of symptom clusters, and provide tiered management recommendations by combining the influence of multiple factors. This would compensate for the shortcomings of existing technologies and achieve precise symptom management. Summary of the Invention

[0009] The purpose of this invention is to address the shortcomings of existing technologies by proposing an intelligent symptom assessment system for patients in the decompensated stage of cirrhosis.

[0010] To achieve the above objectives, the present invention adopts the following technical solution: A symptom assessment system for patients with decompensated cirrhosis includes: The data acquisition module is used to collect multidimensional symptom data and basic characteristic data of patients with decompensated cirrhosis. The multidimensional symptom data is generated based on a preset symptom cluster assessment scale for patients with decompensated cirrhosis, and the basic characteristic data includes demographic and disease characteristics. The symptom cluster analysis module performs item analysis and reliability and validity tests on the multidimensional symptom data, and identifies clustering characteristics of symptoms by calculating the correlation coefficient between symptoms, and divides the symptoms into different symptom clusters. The assessment grading module is used to assess the patient's symptom burden level based on the weighted sum of the scores of each item in the symptom cluster, combined with the associated validity of the criterion scale. The intervention recommendation module is used to output corresponding stratified management recommendations based on the patient's symptom burden level, demographic and sociological characteristics, and disease characteristics, using regression analysis results.

[0011] Preferably, the data acquisition module is configured to perform the following specific operations: For each symptom item in the aforementioned symptom cluster assessment scale for patients with decompensated cirrhosis, quantitative scores were collected across three dimensions: In terms of frequency of occurrence, a discrete rank rating scale is used to map symptom occurrence to a first rating set. ; For the severity dimension, a discrete rating scale is used to map the severity of symptoms to a second rating set. ; For the degree of distress dimension, a discrete rating scale is used to map the degree of symptom interference to a third rating set. ; The final score for each symptom. The basic feature data is determined by the weighted calculation results of the first, second and third scoring sets; the demographic sociological features in the basic feature data include age, gender, marital status and education level, and the disease features include disease course, number of complications and disease severity classification.

[0012] Preferably, the symptom cluster analysis module is configured as follows when performing correlation analysis and clustering: Set correlation coefficient threshold When the correlation coefficient between any two symptoms Greater than When the symptoms are statistically significant, they are considered to have a synergistic effect and co-occurrence. Project Analysis: Using the critical ratio method, the total score of the scale is distributed according to a preset proportion. Divide into high and low groups, if a certain item The value is less than the preset threshold. or Value greater than significance level Delete it; calculate the correlation coefficient between the score of each item and the total score. If the correlation coefficient is less than the preset correlation threshold... If the Cronbach's α coefficient increases after deleting an entry, the system recommends deleting that entry. Principal component analysis and orthogonal rotation with maximum variance were used to reduce the dimensionality of correlated symptoms, and the extracted eigenvalues ​​were greater than the preset values. Common factors; Based on the extracted common factors, the symptoms were divided into ascites-related symptom cluster, encephalopathy with pain symptom cluster, and psychological symptom cluster. Among them, the ascites-related symptom group includes symptoms such as abdominal distension, lower extremity edema, and dyspnea; the encephalopathy with pain symptom group includes manifestations of hepatic encephalopathy, abdominal pain, and muscle spasms; and the psychological symptom group includes symptoms such as anxiety, depression, and sleep disorders.

[0013] Preferably, the data acquisition module is also used to collect patients' self-management behavior data and self-efficacy data. The self-management behavior data were obtained through the Self-Management Behavior Scale for Patients with Cirrhosis. This scale includes dietary management, daily life management, medication management, and disease monitoring management, and was mapped to a fourth score set using a discrete ordinal rating scale. ; The self-efficacy data were obtained through the Chronic Disease Self-Efficacy Scale, which contains multiple items and was mapped to a fifth rating set using a discrete rank rating method. .

[0014] Preferably, the intervention suggestion module is further configured to construct an analysis model based on three dimensions: physiological, psychological, and contextual. The physiological dimension is associated with the patient's disease characteristics and complication data to analyze the impact of pathophysiological mechanisms on symptoms. The psychological dimension is associated with the patient's self-efficacy score and the psychological symptom cluster data identified by the symptom cluster analysis module to analyze the amplification effect of psychological state on symptom perception. The contextual dimension is associated with the patient's demographic sociological characteristics and self-management behavior score to analyze the impact of social environment and management ability on symptom burden. The intervention suggestion module is specifically used to implement the following tiered strategy: If the patient's age is greater than a preset age threshold And the self-efficacy score is lower than the preset efficacy threshold. Then, the first type of intervention recommendations are generated, which mainly include simplifying medication regimens and guidance on family participation in care. If the patient belongs to the ascites-related symptom cluster and the course of the disease exceeds the preset duration Then, a second type of intervention recommendation is generated, which mainly includes dietary control prescriptions and vital sign monitoring records; If the patient belongs to a psychological symptom cluster and their educational level meets the preset conditions. This will generate a third type of intervention suggestion, which mainly includes links to psychological adjustment resources and online consultation access.

[0015] Preferably, the system further includes a scale verification module to ensure the scientific validity of the assessment tool; this module is configured to perform: Construct validity testing: calculation The value and Bartlett's test of sphericity result, when the KMO value is greater than the preset value And Bartlett's test Value less than Then, confirm that the data is suitable for factor analysis; Reliability test: The system calculates Cronbach's α coefficient and split-half reliability. At the same time, the system retests at least 10% of the samples after two weeks. The overall Cronbach's α coefficient, split-half reliability and test-retest reliability of the scale are required to be >0.70, which shows that the scale has good stability and internal consistency.

[0016] Criterion-related validity: The correlation coefficient between the total score of this scale and the total score of the CLDQ was calculated systematically; a significant negative correlation is expected. The specific configuration of the assessment and grading module is as follows: The patient's total symptom score The symptom burden level was determined by comparing the data with a reference dataset. Based on the results of univariate analysis, variables with statistical significance in influencing symptom burden were selected as reference factors for grading adjustment.

[0017] Preferably, the system further includes a prediction model module, which is configured as follows: Basic characteristic data, multidimensional symptom data, self-management behavior data, and self-efficacy data are used as independent variables X, and the total symptom score is used as the dependent variable Y. A symptom burden prediction model was established using regression analysis: ; Based on standardized regression coefficients The absolute value of the value can be used to identify independent risk factors that contribute to an increased symptom burden.

[0018] Preferably, the symptom cluster analysis module also integrates a dynamic tracking unit. Set fixed follow-up time points ; Compare the changes in patients' symptom cluster scores at different time points and calculate the rate of change. ; When the score of a specific symptom cluster rises above a preset warning threshold At that time, a high-risk warning signal is automatically generated and sent to the medical staff terminal.

[0019] A method for intelligent symptom assessment in patients with decompensated cirrhosis, using the aforementioned system, includes the following steps: S1: Obtain general patient information, including demographic and disease characteristics, through a human-computer interaction interface; S2: Display the symptom item pool through the interactive interface and collect discrete level scores for each item across multiple dimensions; S3: Call the symptom cluster analysis module to calculate the correlation coefficient matrix between each symptom, extract common factors to form symptom clusters, and calculate the total symptom score; S4: Invoke the intervention suggestion module, match the independent risk factors obtained from regression analysis with the pre-stored clinical pathways, and generate a personalized stratified management report; In step S4, the generation of the individualized hierarchical management report specifically includes: Based on the results of the univariate analysis, samples with p-values ​​less than [value missing] were selected. Influencing factors; The selected factors are imported into statistical software for regression analysis to obtain the regression coefficients of each factor. The report highlights the top N factors with positive regression coefficients and the largest absolute values, and generates targeted intervention measures.

[0020] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements an intelligent symptom assessment method for patients with decompensated cirrhosis, and the storage medium is used to store an item pool and expert consultation weight data of the symptom cluster assessment scale for patients with decompensated cirrhosis.

[0021] The beneficial effects of this invention are as follows: 1. This invention comprehensively quantifies symptom experience from three dimensions: frequency of occurrence, severity, and degree of distress, avoiding the bias of focusing only on symptom intensity while ignoring frequency or degree of distress, and more accurately reflecting the patient's subjective feelings; based on the theory of unpleasant symptoms, it automatically identifies clusters such as ascites-related symptoms, encephalopathy with pain, and psychological symptom clusters through correlation analysis and factor analysis, revealing the synergistic effect and common pathophysiological mechanism among symptoms, providing a basis for cluster-based intervention in clinical practice, and making intervention more efficient.

[0022] 2. This invention combines three-dimensional factors: physiological (disease characteristics, complications), psychological (self-efficacy, psychological symptom clusters), and situational (demographic sociological characteristics, self-management behavior), and outputs stratified suggestions through a multiple linear regression model; for example, it simplifies medication regimens and introduces family support for elderly patients with low self-efficacy, and strengthens sodium restriction diets and abdominal circumference monitoring for patients with ascites, thereby achieving precise management.

[0023] 3. This invention sets fixed follow-up points, compares changes in symptom cluster scores, monitors the progression of high-risk symptom clusters such as encephalopathy with pain in real time, and automatically pushes early warnings to medical terminals to intervene in advance and reduce the incidence of complications.

[0024] 4. This invention ensures the scientific validity of the scale through item analysis and reliability and validity testing, avoiding subjective bias; at the same time, it is compatible with standardized scales, achieving a unified assessment of symptom burden, quality of life, and self-efficacy. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of the TOUS theoretical model. Detailed Implementation

[0026] The technical solution of the present invention will be further described in detail below with reference to specific embodiments.

[0027] Example 1: A symptom intelligent assessment system and method for patients with decompensated cirrhosis, such as... Figure 1As shown, this system is built on the theory of unpleasant symptoms and is applied to clinical scenarios in hospital infectious disease departments or gastroenterology departments.

[0028] 1. System initialization and database construction

[0029] Before the system goes live, developers need to embed the core data from their research into the system database: Entry pool initialization: The system loads an entry pool containing initial symptom entries.

[0030] Criterion scale data: The system has a built-in Chinese version of the Chronic Liver Disease Questionnaire (CLDQ), whose Cronbach's α coefficient is known to be 0.84~0.90, which will be used for subsequent validity verification.

[0031] 2. Data Acquisition Phase

[0032] 2.1 Enrollment and Screening

[0033] The system obtains the patient list through the hospital's HIS system interface and automatically filters patients with decompensated cirrhosis who meet the diagnostic criteria of the "Guidelines for the Diagnosis and Treatment of Liver Cirrhosis" (2019). The system automatically excludes patients with serious physical diseases such as heart, lung, and kidney diseases or malignant tumors.

[0034] 2.2 Multidimensional Symptom Data Collection

[0035] The system's front end (such as a tablet or webpage) displays the "Symptom Cluster Assessment Scale for Patients with Decompensated Liver Cirrhosis" to the patient. For each symptom (such as "fatigue" or "abdominal distension"), the system collects data from three dimensions: Frequency of occurrence: Using a Likert 5-point rating scale, 0 points = never, 1 point = rarely, 2 points = sometimes, 3 points = often, 4 points = always.

[0036] Severity: Using a Likert 5-point scale, 0 points = not serious at all, 1 point = not serious, 2 points = average, 3 points = serious, 4 points = very serious.

[0037] Distress level: Using a Likert 5-point rating system, 0 points = not bothered at all, 1 point = not bothered much, 2 points = neutral, 3 points = somewhat bothered, 4 points = very bothered.

[0038] The system automatically calculates the final score for each symptom in the background: Score S =(F+I+B) / 3.

[0039] 2.3 Auxiliary Scale Collection

[0040] Self-management behavior: The system used the Self-Management Behavior Scale for Patients with Cirrhosis (compiled by Wang Qian et al.), which contains 24 items (7 items on diet, 7 items on daily life, 5 items on medication, and 5 items on monitoring), and adopted the Likert 4-point scale (1~4 points), with a total score range of 24~96 points.

[0041] Self-efficacy: The system calls the Self-Efficacy Scale for Chronic Diseases (SEMCD), which contains 6 items and uses a rating scale of 1 to 10. The higher the score, the stronger the confidence.

[0042] 3. Symptom cluster analysis and scale validation (server-side processing)

[0043] 3.1 Project Analysis (Cleanup Data)

[0044] The system automatically cleans the collected questionnaires, with a sample size of at least 149.

[0045] Critical Ratio (CR) method: The system categorizes the top 27% of total scores as the high-scoring group and the bottom 27% as the low-scoring group. If an item has a CR value < 3.0 and P > 0.05, the system automatically marks that item as a low-discrimination item.

[0046] Correlation coefficient method: If the correlation coefficient between the score of an item and the total score is <0.40, the system marks the item as a low-correlation item.

[0047] Internal consistency: If deleting an entry increases the Cronbach's α coefficient, the system recommends deleting that entry.

[0048] 3.2 Construct validity and cluster analysis

[0049] The system uses the following algorithm to identify symptom clusters: Suitability test: Calculate the KMO value and Bartlett's test of sphericity. The system determines that the data is suitable for factor analysis only when KMO > 0.70 and Bartlett's test p < 0.05.

[0050] Factor extraction: Principal component analysis (PCA) and orthogonal rotation with maximum variance were used. Common factors with eigenvalues ​​>1 were extracted, and each factor contained at least 3 entries with factor loadings >0.40.

[0051] Clustering results: The system pre-identifies three core symptom clusters: Ascites-related symptom cluster: including abdominal distension, lower extremity edema, and difficulty breathing.

[0052] Encephalopathy with pain symptoms: including hepatic encephalopathy symptoms, abdominal pain, muscle spasms, etc.

[0053] Psychological symptom clusters include anxiety, depression, sleep disorders, etc.

[0054] 3.3 Reliability and Criterion-Related Validity

[0055] Reliability test: The system calculates Cronbach's α coefficient and split-half reliability. At the same time, the system retests at least 10% of the samples after two weeks. The overall Cronbach's α coefficient, split-half reliability and test-retest reliability of the scale are required to be >0.70, which shows that the scale has good stability and internal consistency.

[0056] Criterion correlation: The system calculates the correlation coefficient between the total score of this scale and the total score of the CLDQ. A significant negative correlation is expected (i.e., the more severe the symptoms, the lower the quality of life).

[0057] 4. Assessment and Grading and Analysis of Influencing Factors

[0058] 4.1 Symptom burden classification

[0059] The system calculates the patient's total symptom score and grades it according to the symptom experience (characteristics of inconsistency in frequency, severity, and level of distress): Light burden: Total score x falls within the range of 0 ≤ x < 25%y; Moderate burden: Total score x falls within the range of 25%y ≤ x < 75%y; Heavy burden: Total score x falls within the range of 75%y≤x≤100%y; Where y is the upper limit of the total score.

[0060] 4.2 Regression Analysis of Influencing Factors

[0061] The system imports the 21 potential influencing factors collected into a multiple linear regression model: Y(SymptomBurden)=β0+β1(Age)+β2(Gender)+…+β n (Self-efficacy).

[0062] The system filters out variables with P < 0.10 and includes them in the regression equation. For example, the analysis results may show that "self-efficacy (β = -0.45)" and "disease duration (β = 0.32)" are independent risk factors affecting symptom burden.

[0063] 5. Output of individualized intervention recommendations

[0064] The system generates a report based on the physiological-psychological-situational three-dimensional model and the above analysis results: Contextual dimension: For example, if the patient is a farmer and has a self-efficacy score of <6, the system outputs: "It is recommended to start a family support plan, simplify the medication guidance program, and focus on educating patients on the early identification of complications."

[0065] Physiological dimension: For example, if a patient belongs to the ascites-related symptom cluster and has a Child-Pugh classification of C, the system outputs: "It is recommended to strictly restrict sodium intake, monitor weight and abdominal circumference daily, and be alert for spontaneous bacterial peritonitis."

[0066] Psychological dimension: For example, if a patient belongs to a psychological symptom cluster and has sleep disorders, the system outputs: "Cognitive behavioral therapy intervention is recommended, and links to mindfulness-based stress reduction training resources are provided."

[0067] 6. Dynamic tracking and early warning

[0068] The system sets up follow-up nodes.

[0069] The system compares patient data at T0 (admission) and T1 (2 weeks after discharge).

[0070] If the score for "Encephalopathy with Pain Symptom Cluster" increases by more than a preset threshold (e.g., 20%), the system will automatically send an alert to the medical staff: "Patient [Name] has an increased risk of hepatic encephalopathy. Please intervene promptly."

[0071] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A symptom assessment system for patients in the decompensated stage of cirrhosis, characterized in that, include: The data acquisition module is used to collect multidimensional symptom data and basic characteristic data of patients with decompensated cirrhosis. The multidimensional symptom data is generated based on a preset symptom cluster assessment scale for patients with decompensated cirrhosis, and the basic characteristic data includes demographic and disease characteristics. The symptom cluster analysis module performs item analysis and reliability and validity tests on the multidimensional symptom data, and identifies clustering characteristics of symptoms by calculating the correlation coefficient between symptoms, and divides the symptoms into different symptom clusters. The assessment grading module is used to assess the patient's symptom burden level based on the weighted sum of the scores of each item in the symptom cluster, combined with the associated validity of the criterion scale. The intervention recommendation module is used to output corresponding stratified management recommendations based on the patient's symptom burden level, demographic and sociological characteristics, and disease characteristics, using regression analysis results.

2. The intelligent symptom assessment system for patients in the decompensated stage of cirrhosis according to claim 1, characterized in that, The data acquisition module is configured to perform the following specific operations: For each symptom item in the aforementioned symptom cluster assessment scale for patients with decompensated cirrhosis, quantitative scores were collected across three dimensions: In terms of frequency of occurrence, a discrete rank rating scale is used to map symptom occurrence to a first rating set. ; For the severity dimension, a discrete rating scale is used to map the severity of symptoms to a second rating set. ; For the degree of distress dimension, a discrete rating scale is used to map the degree of symptom interference to a third rating set. ; The final score for each symptom. The basic feature data is determined by the weighted calculation results of the first, second and third scoring sets; the demographic sociological features in the basic feature data include age, gender, marital status and education level, and the disease features include disease course, number of complications and disease severity classification.

3. The intelligent symptom assessment system for patients in the decompensated stage of cirrhosis according to claim 1, characterized in that, The symptom cluster analysis module is specifically configured as follows when performing correlation analysis and clustering: Set correlation coefficient threshold When the correlation coefficient between any two symptoms Greater than When the symptoms are statistically significant, they are considered to have a synergistic effect and co-occurrence. Project Analysis: Using the critical ratio method, the total score of the scale is distributed according to a preset proportion. Divide into high and low groups, if a certain item The value is less than the preset threshold. or Value greater than significance level Delete it; calculate the correlation coefficient between the score of each item and the total score. If the correlation coefficient is less than the preset correlation threshold... If the Cronbach's α coefficient increases after deleting an entry, the system recommends deleting that entry. Principal component analysis and orthogonal rotation with maximum variance were used to reduce the dimensionality of the screened symptoms, and the extracted eigenvalues ​​were greater than the preset values. Common factors; Based on the extracted common factors, the symptoms were divided into ascites-related symptom cluster, encephalopathy with pain symptom cluster, and psychological symptom cluster.

4. The intelligent symptom assessment system for patients in the decompensated stage of cirrhosis according to claim 1, characterized in that, The data acquisition module is also used to collect patients' self-management behavior data and self-efficacy data: The self-management behavior data were obtained through the Self-Management Behavior Scale for Patients with Cirrhosis. This scale includes dietary management, daily life management, medication management, and disease monitoring management, and was mapped to a fourth score set using a discrete ordinal rating scale. ; The self-efficacy data were obtained through the Chronic Disease Self-Efficacy Scale, which contains multiple items and was mapped to a fifth rating set using a discrete rank rating method. .

5. The intelligent symptom assessment system for patients in the decompensated stage of cirrhosis according to claim 4, characterized in that, The intervention suggestion module is further configured to construct an analysis model based on three dimensions: physiological, psychological, and contextual. The physiological dimension is associated with the patient's disease characteristics and complication data to analyze the impact of pathophysiological mechanisms on symptoms. The psychological dimension is associated with the patient's self-efficacy score and the psychological symptom cluster data identified by the symptom cluster analysis module to analyze the amplification effect of psychological state on symptom perception. The contextual dimension is associated with the patient's demographic sociological characteristics and self-management behavior score to analyze the impact of social environment and management ability on symptom burden. The intervention suggestion module is specifically used to implement the following tiered strategy: If the patient's age is greater than a preset age threshold And the self-efficacy score is lower than the preset efficacy threshold. Then, the first type of intervention recommendations are generated, which mainly include simplifying medication regimens and guidance on family participation in care. If the patient belongs to the ascites-related symptom cluster and the course of the disease exceeds the preset duration Then, a second type of intervention recommendation is generated, which mainly includes dietary control prescriptions and vital sign monitoring records; If the patient belongs to a psychological symptom cluster and their educational level meets the preset conditions. This will generate a third type of intervention suggestion, which mainly includes links to psychological adjustment resources and online consultation access.

6. The intelligent symptom assessment system for patients in the decompensated stage of cirrhosis according to claim 1, characterized in that, The system also includes a scale verification module to ensure the scientific validity of the assessment tools. This module is configured to perform: Construct validity testing: calculation The value and Bartlett's test of sphericity result, when the KMO value is greater than the preset value And Bartlett's test Value less than Then, confirm that the data is suitable for factor analysis; Reliability test: The system calculates Cronbach's α coefficient and split-half reliability. At the same time, the system retests at least 10% of the samples after two weeks. The overall Cronbach's α coefficient, split-half reliability and test-retest reliability of the scale are required to be >0.70, which shows that the scale has good stability and internal consistency. Criterion-related validity: The correlation coefficient between the total score of this scale and the total score of the CLDQ was calculated systematically; a significant negative correlation is expected. The specific configuration of the assessment and grading module is as follows: The patient's total symptom score The symptom burden levels were determined by comparing the data with a reference dataset. Based on the results of univariate analysis, variables with statistical significance in influencing symptom burden were selected as reference factors for grading adjustment.

7. The intelligent symptom assessment system for patients in the decompensated stage of cirrhosis according to claim 1, characterized in that, The system also includes a prediction model module, which is configured as follows: Basic characteristic data, multidimensional symptom data, self-management behavior data, and self-efficacy data are used as independent variables X, and the total symptom score is used as the dependent variable Y. A symptom burden prediction model was established using regression analysis: ; Based on standardized regression coefficients The absolute value of the value can be used to identify independent risk factors that contribute to increased symptom burden.

8. The intelligent symptom assessment system for patients in the decompensated stage of cirrhosis according to claim 1, characterized in that, The symptom cluster analysis module also integrates a dynamic tracking unit: Set fixed follow-up time points ; Compare the changes in patients' symptom cluster scores at different time points and calculate the rate of change. ; When the score of a specific symptom cluster rises above a preset warning threshold At that time, a high-risk warning signal is automatically generated and sent to the medical staff terminal.

9. A method for intelligent assessment of symptoms in patients with decompensated cirrhosis, characterized in that, The system described in any one of claims 1 to 8 includes the following steps: S1: Obtain general patient information, including demographic and disease characteristics, through a human-computer interaction interface; S2: Display the symptom item pool through the interactive interface and collect discrete level scores for each item across multiple dimensions; S3: Call the symptom cluster analysis module to calculate the correlation coefficient matrix between each symptom, extract common factors to form symptom clusters, and calculate the total symptom score; S4: Invoke the intervention suggestion module, match the independent risk factors obtained from regression analysis with the pre-stored clinical pathways, and generate a personalized stratified management report; In step S4, the generation of the individualized hierarchical management report specifically includes: Based on the results of the univariate analysis, samples with p-values ​​less than [value missing] were selected. Influencing factors; The selected factors are imported into statistical software for regression analysis to obtain the regression coefficients of each factor. The report highlights the top N factors with positive regression coefficients and the largest absolute values, and generates targeted intervention measures.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the intelligent symptom assessment method for patients in the decompensated stage of cirrhosis as described in claim 9, and the storage medium is used to store the item pool of the symptom cluster assessment scale for patients in the decompensated stage of cirrhosis.

Images