A liver cirrhosis prognosis risk detection system

Abstract

The present application relates to the technical field of medical information processing, in particular to a liver cirrhosis prognosis risk detection system. After each detection of the patient's liver by the system, the image data can quantify the image liver cirrhosis risk from the morphological changes in the liver region. According to the difference between the actual detection data and the prediction data, and the prognosis response non-normality, the disease progression index is determined. The drug adjustment data is used to determine the drug adjustment degree, and then the liver cirrhosis progression index is obtained. The data lag and the liver cirrhosis progression index are quantified to adjust the detection frequency of the system. The present application enables the system to effectively respond to the patient's condition and generate an effective early warning signal to remind medical staff to perform liver detection when necessary.

Description

Technical Field

[0001] This invention relates to the field of medical information processing technology, specifically to a prognostic risk detection system for cirrhosis. Background Technology

[0002] Prognostic assessment of cirrhosis requires comprehensive consideration of multiple factors, including the patient's clinical presentation, laboratory results, imaging findings, and pathological changes. Existing cirrhosis prognostic management systems can assess and warn of patient risk through data collection, preprocessing, feature extraction, and dynamic monitoring and early warning. However, due to inappropriate settings for patient testing frequency, key data on liver disease may not be detected in a timely manner, leading to inaccurate assessments of abnormal prognostic risks and hindering efforts to improve patient outcomes and reduce the risk of complications. Summary of the Invention

[0003] To address the technical problem in existing liver cirrhosis prognostic risk assessment and early warning systems where inappropriate detection frequency settings lead to the failure to detect key liver disease data in a timely manner, resulting in incorrect assessments of patients' abnormal prognostic risk for liver cirrhosis, this invention aims to provide a liver cirrhosis prognostic risk detection system. The specific technical solution adopted is as follows:

[0004] This invention proposes a prognostic risk detection system for liver cirrhosis, the system comprising:

[0005] The liver imaging data analysis module is used to obtain the risk of liver cirrhosis in each detection process based on the changes in the area of ​​the liver region in the liver image between adjacent detection processes and the edge roughness.

[0006] The cirrhosis progression analysis module is used to acquire the patient's test data for each test; obtain the predicted data corresponding to the test data using a pre-trained predictive neural network; obtain the patient's state deviation index for each test based on the difference between the predicted data and the test data; obtain the complication risk for each test based on the state deviation index and the non-standard prognostic response statistically obtained in the prognostic follow-up system; and obtain the disease progression index for each test based on the imaging cirrhosis risk.

[0007] The drug adjustment impact analysis module is used to obtain the drug adjustment degree based on the difference in drug types before each test compared to the previous test; and to obtain the cirrhosis malignancy progression index under each test based on the drug adjustment degree and the disease progression index.

[0008] The system data lag analysis module is used to statistically analyze the changes in the cirrhosis progression index and the changes in the detection frequency between all adjacent historical detection processes in each detection process; the system data lag is obtained based on the asynchronous nature of the changes in the cirrhosis progression index and the changes in the detection frequency.

[0009] The system detection frequency correction module is used to adjust the system detection frequency based on the liver cirrhosis progression index and the data lag during each detection process.

[0010] Further, the detection data includes physiological data sequences and physiological health scores; the prediction data includes predicted physiological data sequences and predicted physiological health scores; the Euclidean distance between the physiological data sequences and the predicted physiological data sequences is obtained; the score difference between the predicted physiological health score and the physiological health score is obtained, and the product of the score difference and the Euclidean distance is used as the state deviation index.

[0011] Furthermore, the method for obtaining the risk of liver cirrhosis via imaging includes:

[0012] For the current detection process, the area of ​​the liver region in the current detection process is used as the denominator, and the area of ​​the liver region in the previous detection process is used as the numerator to obtain the area comparison feature; the edge roughness of the liver region in the current detection process is used as the numerator, and the edge roughness of the liver region in the previous detection process is used as the denominator to obtain the edge roughness comparison feature; the imaging cirrhosis risk is obtained based on the area comparison feature and the edge roughness comparison feature.

[0013] Furthermore, the method for obtaining the drug adjustment degree includes:

[0014] The drug adjustment degree is obtained by taking the number of drug types in the current testing process as the numerator and the number of drug types that are the same in the current testing process and the previous testing process as the denominator.

[0015] Furthermore, the method for obtaining the liver cirrhosis malignancy progression index includes:

[0016] The difference in the disease progression index between each test and the previous test is obtained. After normalizing the difference in the disease progression index, the sum of the difference and the normalized drug adjustment degree is used as the cirrhosis malignancy progression index.

[0017] Furthermore, the method for obtaining the data lag includes:

[0018] Select a preset number of consecutive historical detection processes from the current detection process; for each group of adjacent historical detection processes, obtain the difference in the cirrhosis malignancy progression index and the difference in detection frequency between the two historical detection processes;

[0019] The difference in the average cirrhosis progression index between adjacent historical detection processes of all groups is used as the numerator, and the sum of the average detection frequency difference and the preset hyperparameter is used as the denominator. The ratio obtained is the data lag.

[0020] Furthermore, adjusting the system detection frequency based on the cirrhosis progression index during each detection process and the data lag includes:

[0021] The correction coefficient is obtained by multiplying the cirrhosis progression index under each test process by the data lag and then normalizing the result.

[0022] If the correction coefficient is less than a preset threshold, the detection frequency is not adjusted; otherwise, an increase coefficient is obtained based on the correction coefficient, and the detection frequency is increased using the increase coefficient.

[0023] Further, the step of obtaining the growth coefficient based on the correction coefficient and increasing the detection frequency using the growth coefficient includes:

[0024] The sum of the positive integer 1 and the correction coefficient is used as the growth coefficient, and the product of the growth coefficient and the detection frequency is used as the increased detection frequency.

[0025] Furthermore, the method for obtaining the prognostic response non-standardity includes:

[0026] The patient's behavior execution record sequence is statistically analyzed in the follow-up system during each testing process; each element in the behavior execution record sequence represents the patient's degree of execution of the will, and the element value includes effective execution and ineffective execution. The number of ineffective executions is counted as the degree of non-standard prognostic response.

[0027] Furthermore, the method for obtaining the disease progression index includes:

[0028] The relative difference between the imaging risk of cirrhosis in each testing process and the historical average imaging risk of cirrhosis is obtained, and the product of the relative difference and the risk of complications is used as the disease progression index.

[0029] The present invention has the following beneficial effects:

[0030] This invention quantifies the risk of liver cirrhosis based on morphological changes in the liver region using imaging data after each liver examination. A state deviation index is determined by the difference between the actual test data and the expected predicted data. This is further combined with the prognostic response non-standardization in the prognostic follow-up system to quantify the patient's data on disease prognosis and behavioral abnormalities during each test. Considering that medication adjustments are crucial information during patient prognosis, the degree of medication adjustment is determined using medication adjustment data, thereby obtaining a liver cirrhosis progression index. This liver cirrhosis progression index can assess the patient's liver cirrhosis risk after each test, providing a reference for adjusting the system's testing frequency. Furthermore, considering the potential for system lag, a slower testing frequency update indicates greater lag, as significant changes in the liver cirrhosis progression index occur during the testing process. Therefore, the system's testing frequency is adjusted by quantifying data lag and the liver cirrhosis progression index together, enabling the system to effectively respond to the patient's condition and generate effective early warning signals to remind medical staff to perform liver tests when necessary. Attached Figure Description

[0031] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 This is a block diagram of a liver cirrhosis prognostic risk detection system provided in one embodiment of the present invention. Detailed Implementation

[0033] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a liver cirrhosis prognostic risk detection system proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0034] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0035] The specific scheme of the liver cirrhosis prognostic risk detection system provided by the present invention will be described in detail below with reference to the accompanying drawings.

[0036] Please see Figure 1 The diagram illustrates a block diagram of a liver cirrhosis prognostic risk detection system according to an embodiment of the present invention. The system includes: a liver imaging data analysis module 101, a liver cirrhosis progression analysis module 102, a drug adjustment impact analysis module 103, a system data lag analysis module 104, and a system detection frequency correction module 105.

[0037] It should be noted that this embodiment of the invention aims to adjust the system's detection frequency based on the liver data from each test. Therefore, other modules in the system, such as data acquisition and data transmission, will not be described further. Existing data acquisition modules can be used to directly access medical data obtained from each patient's test in the hospital, and specific details will not be elaborated upon. This embodiment of the invention mainly analyzes bilirubin levels, transaminase levels, and liver imaging; therefore, these three types of data can be directly accessed for processing within the system of this embodiment.

[0038] Imaging examinations can non-invasively provide information on the anatomical structure and pathological changes of the liver, including liver size, surface smoothness, vascularity, and potential liver nodules or complications. The liver imaging data analysis module 101 analyzes the morphology of liver regions in liver imaging data, selecting liver region area and edge roughness as imaging features for analysis. Based on changes in liver region area and edge roughness between adjacent examinations, the imaging risk of cirrhosis can be obtained for each examination. In other words, the imaging risk of cirrhosis represents the degree of risk of morphological changes in the liver region. If, compared to the previous examination, the liver is more atrophied and exhibits irregular, jagged edges, it indicates a higher degree of malignant progression of cirrhosis in the patient during this examination, and thus a greater imaging risk of cirrhosis.

[0039] Preliminary assessment results regarding changes in liver structure and function were obtained through the liver imaging data analysis module 101. These imaging features, such as reduced liver volume, irregular surface, and portal vein dilation, all indicate the progression of cirrhosis. However, the progression of liver disease is not only affected by its own pathological changes but also significantly interfered with by the drug treatments received by the patient, which may lead to "false positives" in the assessment of cirrhosis progression.

[0040] This invention further considers that the prognosis management system should focus on the orderly changes in the patient's condition. If the liver disease shows significant malignant progression and irregular fluctuations during the prognosis process, it also indicates that the system's detection frequency needs to be increased, and more attention needs to be paid to the patient. Therefore, this invention sets up a liver cirrhosis progression analysis module 102 to acquire the patient's detection data for each detection process and use a pre-trained predictive neural network to obtain the predicted data corresponding to the detection data. Based on the difference between the predicted data and the detection data, the patient's state deviation index for each detection process is obtained. That is, the greater the difference between the two data, the more the patient's condition exceeds expectations in the actual process, the greater the state deviation index, and the more attention the patient needs to receive at a higher detection frequency. Further, the prognosis response non-standardization degree in the prognosis follow-up system is statistically analyzed. The prognosis response non-standardization degree is the patient's adherence to the will during the prognosis process. The greater the prognosis response non-standardization degree, the less effectively the patient follows the doctor's instructions, and the greater the risk. Therefore, the risk of complications in each detection process can be further statistically analyzed. Combined with the imaging liver cirrhosis risk, the disease progression index for each detection process can be obtained. The greater the disease progression index, the more frequently the system needs to monitor the patient.

[0041] During a patient's prognosis process, medications are typically adjusted based on test results. The degree of medication adjustment reflects the patient's current state of stability. Greater adjustments require a faster response time from the testing system. Therefore, the medication adjustment impact analysis module 103 calculates the degree of medication adjustment based on the difference in medication types between the previous and current testing processes. This is further combined with the disease progression index to obtain the cirrhosis malignancy progression index for each testing process. In other words, the cirrhosis malignancy progression index simultaneously represents the risk of liver excretion function, liver morphology risk, and the impact of medication adjustments. A higher cirrhosis malignancy progression index indicates a greater need for the testing system to increase the testing frequency for that patient.

[0042] In practical applications, the system may suffer from response lag. This lag effect may prevent changes in a patient's condition from being reflected in the early warning system in a timely manner, thus affecting clinical decision-making and timely intervention. For example, when a patient's liver function changes due to medication adjustments, the system may fail to capture these changes immediately, leading to a delay in the early warning signal. Furthermore, a patient's physiological state may fluctuate rapidly; if the system cannot adapt quickly to these changes, it may result in inaccurate and untimely risk assessment. Therefore, the system data lag analysis module 104 statistically analyzes the changes in the cirrhosis progression index and the changes in the detection frequency between all adjacent historical detection processes for each test. The greater the asynchrony between these two changes—that is, the less effectively the detection frequency responds to large changes in the cirrhosis progression index—the greater the data lag in that detection process. Therefore, further quantification of data lag allows the system detection frequency correction module 105 to adjust the system detection frequency based on the cirrhosis progression index and data lag in each detection process. A reasonable and effective detection frequency can avoid missed detections of the patient's liver data and improve prognosis.

[0043] In summary, this invention quantifies the risk of liver cirrhosis based on morphological changes in the liver region using imaging data after each liver examination. A disease progression index is determined based on the difference between actual and predicted data, as well as the degree of non-standard prognostic response. The degree of drug adjustment is determined using medication adjustment data, thereby obtaining a cirrhosis malignancy progression index. The system's detection frequency is adjusted in conjunction with the quantified data lag and the cirrhosis malignancy progression index. This invention enables the system to effectively respond to the patient's condition, generating effective early warning signals to remind medical staff to perform liver examinations when necessary.

[0044] Preferably, in this embodiment of the invention, the detection data includes a physiological data sequence and a physiological health score; the prediction data includes a predicted physiological data sequence and a predicted physiological health score; the Euclidean distance between the physiological data sequence and the predicted physiological data sequence is obtained; the score difference between the predicted physiological health score and the physiological health score is obtained, and the product of the score difference and the Euclidean distance is used as the state deviation index.

[0045] As a specific example, the physiological data sequence in this embodiment of the invention may include various biochemical indicators of the liver, laboratory test data, medical procedures, clinical time, etc., such as bilirubin, transaminase levels, heart rate, blood pressure, etc. Specific settings can be configured according to system requirements and will not be elaborated here. It should be noted that the physiological health score is also obtained by the system using existing scoring systems or by using a doctor's judgment. The higher the physiological health score, the better the patient's liver function recovery. These scoring systems can quantify liver function status, help identify high-risk patients, assist doctors in assessing patient prognosis, and thus provide more reliable data support for treatment or intervention plans. Specific details are also well known to those skilled in the art and will not be elaborated or limited here.

[0046] The neural network used in this embodiment of the invention is an LSTM neural network. In other embodiments of the invention, a Transformer network model may also be used. The specific model training methods are well known to those skilled in the art and will not be elaborated here. The input to the prediction neural network is the physiological data sequence and physiological health score obtained in one detection process, and the output is the predicted physiological data sequence and the predicted physiological health score.

[0047] As a specific example, the state deviation index in this embodiment of the invention is expressed by the formula:

[0048] Where, d a Let M' be the state deviation index during the a-th detection process. a M is the predicted physiological health score for the a-th testing process. a H' represents the physiological health score for the a-th testing process. a H represents the predicted physiological data sequence during the a-th detection process. a Let |H' be the physiological data sequence of the a-th detection process. a -H a || represents the Euclidean distance between two sequences. In this formula, the difference in scores is represented by a ratio. The larger the ratio, the more significantly the predicted physical health score is greater than the actual physical health score, indicating that the patient's postoperative condition has not met expectations, and thus the greater the condition deviation index. Similarly, a larger Euclidean distance also indicates that the patient's postoperative condition has not met expectations, and thus the greater the condition deviation index.

[0049] Preferably, in this embodiment of the invention, the method for obtaining the risk of liver cirrhosis via imaging includes:

[0050] For the current testing process, the liver area in the current testing process is used as the denominator, and the liver area in the previous testing process is used as the numerator to obtain the area contrast feature. The larger the area contrast feature, the smaller the area in the current testing process is compared to the previous testing process, indicating that the liver has shrunk, and the greater the risk of cirrhosis.

[0051] The edge roughness of the liver region in the current detection process is used as the numerator, and the edge roughness of the liver region in the previous detection process is used as the denominator to obtain the edge roughness contrast feature. The larger the edge roughness contrast feature, the rougher the edge roughness of the liver region in the current detection process is compared with the previous detection process, and the greater the risk of cirrhosis should be.

[0052] The risk of liver cirrhosis on imaging is obtained based on area contrast features and edge roughness contrast features. In this embodiment of the invention, the product of the area contrast features and the edge roughness contrast features is used as the risk of liver cirrhosis on imaging.

[0053] Preferably, in this embodiment of the invention, the method for obtaining edge roughness includes:

[0054] The curvature of each edge point on the edge of the liver region is obtained, and the variance of the curvature is used as the edge roughness. Variance represents the uniformity of data distribution; therefore, the larger the variance, the more uneven the curvature distribution, the more obvious the jagged features, and the greater the edge roughness.

[0055] Preferably, in this embodiment of the invention, the method for obtaining the prognostic response non-standardity includes:

[0056] The patient's behavior execution record sequence is statistically analyzed in the follow-up system during each testing process; each element in the behavior execution record sequence represents the patient's degree of execution of the will, and the element value includes effective execution and ineffective execution. The number of ineffective executions is counted as the degree of non-standard prognostic response.

[0057] As a specific example, in this embodiment of the invention, the execution record sequence is obtained statistically by the follow-up system. The follow-up system can record various data during the patient's follow-up process, such as: during the current patient's prognosis monitoring (including historical data), the time sequence of medical order execution (e.g., prescription issuance time, medication attendance records) can be obtained from a smart pillbox or follow-up platform; the execution rate of biochemical tests (whether necessary tests are completed on schedule) can be obtained from the LIS testing system or the data interface of the health checkup center; vital sign measurement records (whether weight, ascites, and blood pressure are measured regularly) can be uploaded from home devices or the follow-up platform; and the delay in doctor-patient interaction response (whether the patient responds promptly after the doctor issues a reminder / consultation) can be obtained from the corresponding logs of the doctor-patient system. These records are then converted into a binary sequence to obtain the execution record sequence. In this embodiment of the invention, the element value of valid execution is set to 1, and the element value of invalid execution is set to 0. That is, the prognostic response non-standardization degree is the number of elements with a value of 0 in the execution record sequence.

[0058] In this embodiment of the invention, the product of the state deviation index and the prognostic response non-standardity can be directly used as the complication risk for each testing process.

[0059] Preferably, in one embodiment of the present invention, the method for obtaining the disease progression index includes:

[0060] The relative difference between the imaging risk of cirrhosis at each testing session and the historical average imaging risk of cirrhosis is obtained. The product of the relative difference and the risk of complications is used as the disease progression index. The historical average imaging risk of cirrhosis is the average imaging risk of cirrhosis in all other historical testing sessions of the patient, excluding the current testing session.

[0061] It should be noted that the relative differences in the embodiments of this invention can also be quantified in the form of a ratio, where the numerator is the imaging risk of liver cirrhosis during the current testing process, and the denominator is the historical average imaging risk of liver cirrhosis. A larger relative difference indicates a greater change in the patient's liver morphology during the current testing process, suggesting faster disease progression, and thus a higher disease progression index. Similarly, a higher risk of complications indicates a higher disease progression index; therefore, a product method is used to fuse and quantify the two indicators.

[0062] Preferably, in one embodiment of the present invention, the method for obtaining the drug adjustment degree includes:

[0063] The drug adjustment degree is calculated by dividing the number of drug types in the current testing process by the number of drug types that are the same in the current testing process and the number of drug types that are the same in the previous testing process by the number of drug types in the current testing process. The drug adjustment degree is a ratio; the smaller the denominator, the greater the drug adjustment degree, indicating that the patient has undergone a greater drug adjustment in the current testing process compared to the previous one. It should be noted that this embodiment of the invention focuses on drug types, not drug brands. In the prognosis of cirrhosis, certain types of drugs are essential drugs, therefore the denominator of the drug adjustment degree will not be 0.

[0064] Preferably, the method for obtaining the liver cirrhosis malignancy progression index in this embodiment of the invention includes:

[0065] The difference in the disease progression index between each testing process and the previous testing process is obtained. After normalizing the difference in the disease progression index, the sum of this difference and the normalized drug adjustment degree is used as the cirrhosis malignancy progression index. It should be noted that, in this embodiment of the invention, the difference in the disease progression index between each testing process and the previous testing process can also be quantified using a ratio, i.e., the numerator is the disease progression index of the current testing process, and the denominator is the disease progression index of the previous testing process.

[0066] In the embodiments of the present invention, the normalization operation can be implemented by range standardization through the range of data values ​​in the data dimension, or by function mapping methods such as hyperbolic tangent function mapping, which will not be elaborated further.

[0067] Preferably, in this embodiment of the invention, the method for obtaining data lag includes:

[0068] A predetermined number of consecutive historical detection processes are selected prior to the current detection process. For each group of adjacent historical detection processes, the difference in the cirrhosis progression index and the difference in detection frequency between the two historical detection processes are obtained. It should be noted that the number of historical detection processes selected here is the same as that selected in the first cirrhosis risk acquisition process, which is 3.

[0069] The difference in the average cirrhosis progression index between adjacent historical detection processes of all groups is used as the numerator, and the sum of the average detection frequency difference and the preset hyperparameter is used as the denominator. The ratio obtained is the data lag. That is, if the numerator is larger and the denominator is smaller, it indicates that the cirrhosis progression index has changed significantly within a certain time range, but the detection frequency has not been significantly adjusted. The hyperparameter is a positive integer of 1 to prevent the denominator from being 0.

[0070] Preferably, in this embodiment of the invention, adjusting the system detection frequency based on the cirrhosis progression index and the data lag during each detection process includes:

[0071] The correction coefficient is obtained by multiplying the cirrhosis progression index for each test by the data lag and then normalizing the results. It should be noted that the normalization method here is the same as that described above, both employing range standardization; further details will not be elaborated upon.

[0072] If the correction coefficient is less than a preset threshold, it indicates a high data update frequency, meaning the system can obtain new data promptly, reflecting the latest patient status. The system can frequently update data and identify changes in patient status in a timely manner, indicating that no adjustment to the detection frequency is needed. Otherwise, it means the system's data update frequency is low, and historical data may no longer be highly effective in the current situation. This is because a low update frequency means lagging clinical data, which may lead to risks not being identified in a timely manner. Therefore, it is necessary to obtain an increase coefficient based on the correction coefficient and use this increase coefficient to increase the detection frequency.

[0073] In this embodiment of the invention, considering that the correction coefficient has been normalized, the threshold is set to 0.6, and the growth coefficient is a positive integer 1 plus the correction coefficient. The growth coefficient can be increased by multiplying it by the original detection frequency.

[0074] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0075] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A liver cirrhosis prognosis risk detection system characterized by, The system comprises: a liver image data analysis module, configured to obtain an image cirrhosis risk in each detection process according to the area change of the liver region in the liver image between adjacent detection processes and the edge roughness; a liver cirrhosis progression analysis module, configured to obtain detection data of the patient in each detection process, obtain prediction data corresponding to the detection data by using a pre-trained prediction neural network, obtain a state deviation index of the patient in each detection process according to the difference between the prediction data and the detection data, obtain a complication risk in each detection process according to the state deviation index and a prognosis response irregularity statistically obtained in a prognosis follow-up system, and obtain a disease progression index in each detection process in combination with the image cirrhosis risk; a drug adjustment influence analysis module, configured to obtain a drug adjustment degree according to the difference between the drug types before each detection process and the drug types in the last detection process, and obtain a liver cirrhosis deterioration progression index in each detection process according to the drug adjustment degree and the disease progression index; a system data lag analysis module, configured to, in each detection process, statistically obtain the liver cirrhosis deterioration progression index change between all adjacent historical detection processes and the detection frequency change between adjacent historical detection processes, and obtain data lag of the system according to the asynchronization between the liver cirrhosis deterioration progression index change and the detection frequency change; a system detection frequency correction module, configured to adjust the system detection frequency according to the liver cirrhosis deterioration progression index in each detection process and the data lag.

2. The liver cirrhosis prognosis risk detection system according to claim 1, wherein The detection data comprises a physiological data sequence and a physiological data health score; and the prediction data comprises a predicted physiological data sequence and a predicted physiological data health score. An Euclidean distance between the physiological data sequence and the predicted physiological data sequence is obtained, a score difference between the predicted physiological data health score and the physiological data health score is obtained, and the product of the score difference and the Euclidean distance is taken as the state deviation index.

3. The liver cirrhosis prognosis risk detection system according to claim 1, wherein The method for obtaining the image cirrhosis risk comprises: for the current detection process, taking the liver region area in the current detection process as the denominator and the liver region area in the last detection process as the numerator to obtain an area comparison feature, and taking the edge roughness of the liver region in the current detection process as the numerator and the edge roughness of the liver region in the last detection process as the denominator to obtain an edge roughness comparison feature, and obtaining the image cirrhosis risk according to the area comparison feature and the edge roughness comparison feature.

4. The liver cirrhosis prognosis risk detection system according to claim 1, wherein The method for obtaining the drug adjustment degree comprises: taking the number of drug types in the current detection process as the numerator and the number of the same drug types between the current detection process and the last detection process as the denominator to obtain the drug adjustment degree.

5. The liver cirrhosis prognosis risk detection system according to claim 1, wherein The method for obtaining the liver cirrhosis deterioration progression index comprises: obtaining the difference between the disease progression index between each detection process and the last detection process, normalizing the difference between the disease progression indexes, and taking the sum of the normalized disease progression index and the normalized drug adjustment degree as the liver cirrhosis deterioration progression index.

6. The liver cirrhosis prognosis risk detection system according to claim 1, wherein The method for obtaining the data lag comprises: selecting a preset number of continuous historical detection processes in sequence from the current detection process; for each set of adjacent historical detection processes, obtaining a difference in liver cirrhosis progression index between the two historical detection processes and a difference in detection frequency; taking the average difference in liver cirrhosis progression index of all sets of adjacent historical detection processes as the numerator, and the sum of the average difference in detection frequency and a preset hyperparameter as the denominator, and obtaining a ratio value as the data lag.

7. The liver cirrhosis prognosis risk detection system according to claim 1, wherein adjusting the system detection frequency according to the liver cirrhosis progression index in each detection process and the data lag, including: multiplying the liver cirrhosis progression index in each detection process and the data lag to obtain a correction coefficient, and performing normalization processing on the correction coefficient; if the correction coefficient is less than a preset threshold, the detection frequency is not adjusted; otherwise, obtaining a growth coefficient according to the correction coefficient, and increasing the detection frequency by using the growth coefficient.

8. The liver cirrhosis prognosis risk detection system according to claim 7, wherein obtaining a growth coefficient according to the correction coefficient, and increasing the detection frequency by using the growth coefficient, including: taking the sum of a positive integer 1 and the correction coefficient as the growth coefficient, and taking the product of the growth coefficient and the detection frequency as the increased detection frequency.

9. The liver cirrhosis prognosis risk detection system according to claim 1, wherein the method for obtaining the prognosis response irregularity includes: counting the behavior execution record sequence of the patient in the follow-up system in each detection process; each element in the behavior execution record sequence represents the execution degree of the patient on the will behavior, and the element value includes valid execution and invalid execution, and the number of invalid executions is counted as the prognosis response irregularity.

10. The liver cirrhosis prognosis risk detection system according to claim 1, wherein the method for obtaining the disease progression index includes: obtaining the relative difference between the image liver cirrhosis risk of each detection process and the historical average image liver cirrhosis risk, and taking the product of the relative difference and the complication risk as the disease progression index.

Images