System and method for dynamically monitoring sepsis blood perfusion in combination with blood oxygen and blood flow parameters

Abstract

The invention discloses a sepsis blood perfusion dynamic monitoring method combining blood oxygen and blood flow parameters, and belongs to the technical field of medical monitoring. The method specifically comprises the following steps: S1, constructing a sepsis patient grading evaluation diagnosis and treatment model, and obtaining sepsis patient information, index data, sepsis severity at that time, a treatment scheme and post-treatment index data to train the model; s2, information and index data of current sepsis patients are collected, blood flow and blood oxygen are measured, the current sepsis patients are matched through the sepsis patient grading diagnosis and treatment model, the severity degree of sepsis of the current patients is judged, and the information, the index data and the severity degree of sepsis of the current patients are displayed. By acquiring basic information, physiological indexes, severity, treatment schemes and post-treatment data of sepsis patients, a database covering a whole diagnosis and treatment process is constructed, so that the model can learn association between different patient features and diagnosis and treatment results, and comprehensiveness of evaluation and prediction is improved.

Description

Technical Field

[0001] The present invention relates to the field of medical monitoring technology, and in particular to a sepsis blood perfusion dynamic monitoring system and method combining blood oxygen and blood flow parameters. Background Art

[0002] Sepsis is a life-threatening disease that can turn into septic shock, which can have a 40% mortality rate, without early intervention. Sepsis is caused by the body's intense response to infection and typically occurs in three stages. In the first stage, the infection reaches the bloodstream and causes inflammation. In the second stage, the infection reaches other organs. The final stage is septic shock, which leads to organ failure. Sepsis treated in the early stages has a very high survival rate. The stage at which a patient is diagnosed is crucial, and each stage has a different mortality rate. For example, the mortality rate of severe sepsis and septic shock is very high within 28 days. The main causes of sepsis are typically pneumonia, urinary tract infections, digestive system infections, and bloodstream infections. Septic infections are more common in the very elderly and very young, as well as in people with compromised immune systems or pre-existing conditions (such as diabetes). Sepsis is also common in patients undergoing major surgery or procedures that damage the patient's immune system. According to the CDC, four physical findings are used to diagnose sepsis: fever, low blood pressure, increased heart rate, and difficulty breathing. Although additional laboratory tests may be needed to confirm sepsis, these physical measurements can help provide initial indicators for recommended interventions. Since approximately 80% of sepsis cases begin outside of a hospital setting, appropriately timed intervention can save many lives. Sepsis involves numerous physiological parameters and blood biochemical indices, and their relationships are complex, requiring specialized physician interpretation. In clinical practice, inaccurate interpretations or omissions of important information are common. Summary of the Invention

[0003] Purpose of the invention: The purpose of the present invention is to provide a system and method for dynamic monitoring of sepsis blood perfusion that combines blood oxygen and blood flow parameters; it can solve the problem of inaccurate interpretation or omission of important information in clinical practice.

[0004] Technical solution: To solve the above technical problems, according to one aspect of the present invention, more specifically a system and method for dynamic monitoring of sepsis blood perfusion combining blood oxygen and blood flow parameters, specifically comprising the following steps: S1. Build a hierarchical assessment and treatment model for sepsis patients, obtain sepsis patient information, indicator data, the severity of sepsis at the time, treatment plan, and post-treatment indicator data to train the model; S2. Collect information and indicator data of the current sepsis patient, including measuring blood flow and blood oxygen, match the current sepsis patient with the sepsis patient hierarchical diagnosis and treatment model, determine the severity of the current sepsis patient, and display the current sepsis patient information, indicator data, and sepsis severity; S3. Based on the current patient information and indicator data, provide a recommended treatment plan through the hierarchical diagnosis and treatment model for sepsis patients; S4. Determine a selected treatment plan, record adjustments to the selected treatment plan, analyze the adjustments to the selected treatment plan using an intelligent single-chip computer to determine an adoption and application index for the treatment plan, determine treatment measures for the treatment plan based on the adoption and application index, and update the treatment plan to the hierarchical diagnosis and treatment model for sepsis patients based on the treatment measures for the treatment plan; S5. Input the current patient's information, indicator data, confirmed severity of sepsis, treatment plan, and indicator data after treatment into the hierarchical diagnosis and treatment model for sepsis patients to improve the model training.

[0005] Furthermore, the patient information includes the patient's age, height, and weight; the index data includes the patient's basic physiological data, the patient's blood oxygenation, and blood flow parameters required for dynamic monitoring of sepsis blood perfusion; and the post-treatment index data includes the patient's basic physiological data after treatment, the patient's blood oxygenation and blood flow parameters required for dynamic monitoring of sepsis blood perfusion, and the corresponding timestamp.

[0006] Furthermore, the step S1 specifically includes the following steps: S11. Build a hierarchical assessment and treatment model for sepsis patients to obtain information on sepsis patients, indicator data, the severity of sepsis at the time, treatment plans, and post-treatment indicator data; S12. Preprocess the acquired sepsis patient information, indicator data, severity of sepsis at the time, treatment plan, and post-treatment indicator data to filter out invalid data and extract feature vectors of the patient information, indicator data, and post-treatment indicator data; S13. Bundle the extracted patient information indicator data and the characteristic vector of the post-treatment indicator data with the treatment plan corresponding to the patient at that time to obtain the big data and input it into the sepsis patient grading assessment and diagnosis and treatment model to train the sepsis patient grading assessment and diagnosis and treatment model.

[0007] Furthermore, the step S2 specifically includes the following steps: S21. Collect information and indicator data of current sepsis patients, including measuring blood flow and blood oxygen, pre-process the real-time collected data, remove noise from the data, and perform spatiotemporal alignment of various data; S22. Extracting characteristic vectors of the current sepsis patient information and indicator data from the preprocessed data, transmitting the extracted characteristic vectors to a sepsis patient grading assessment and diagnosis and treatment model for matching, and obtaining the sepsis severity of the sepsis patient in the sepsis patient grading assessment and diagnosis and treatment model that has the highest matching degree with the characteristic vector of the current sepsis patient; S23. Display the current sepsis patient information and indicator data in real time, and push alerts on the severity of sepsis in the current sepsis patient.

[0008] Furthermore, in step S3, when giving a recommended treatment plan, the extracted characteristic vector of the current sepsis patient will be matched through the sepsis patient graded assessment and diagnosis model. If multiple treatment plans are matched, they will be sorted according to the patient's post-treatment indicators of each treatment plan. The treatment plan with more obvious recovery of the patient's post-treatment indicator data and faster recovery time will be ranked higher.

[0009] Furthermore, the specific steps of step S4 are: S41. Select a treatment plan to be implemented from the suggested treatment plans as an initial treatment plan template; S42: Determine whether to adjust the initial treatment plan template by analyzing the initial treatment plan template, record the adjustment data of the initial treatment plan template, obtain an implementation treatment plan after the analysis, treat the current sepsis patient according to the implementation treatment plan, and record the treatment data of the current sepsis patient in real time; S43. Analyze and process the adjustment data of the recorded initial treatment plan template to obtain the adoption and application index of the treatment plan. If the adoption and application index is higher than the preset threshold, the treatment plan is retained. Otherwise, the treatment plan is marked as a treatment plan that needs to be evaluated. The evaluation of the treatment plan includes adjustment and deletion operations.

[0010] Furthermore, in step S43, when analyzing and processing the adjustment data of the recorded initial treatment plan template to obtain the adoption and application index of the treatment plan, the adoption and application index of the treatment plan is obtained by analyzing the number of times the treatment plan is selected as the initial treatment plan, the number of times the initial treatment plan is used after adjustment, the total number of treatment parameters of the initial treatment plan, and the number of treatment parameters adjusted each time for the initial treatment plan: 、 in, is the adoption index of the treatment plan, is the number of times the treatment regimen was selected as the initial treatment regimen, is the number of times the initial treatment regimen was adjusted. is the total number of treatment parameters of the initial treatment plan, The number of treatment parameters that are adjusted each time for this initial treatment plan.

[0011] According to another aspect of the present invention, a system for dynamic monitoring of sepsis blood perfusion by combining blood oxygen and blood flow parameters is provided. The system is used to implement the above-mentioned method for dynamic monitoring of sepsis blood perfusion by combining blood oxygen and blood flow parameters, and includes: a data acquisition module, a data processing module, a graded evaluation and diagnosis module, a data storage module, an analysis and processing module, and a display and interaction module; Data acquisition module: used to collect various data during the use of the monitoring system; Data processing module: used to remove noise and perform spatiotemporal alignment preprocessing on the patient's relevant data, extract the feature vector of the preprocessed data, and input the extracted feature vector into the hierarchical evaluation and treatment module for analysis and processing; Grading Assessment and Treatment Module: This module is used to analyze and process the input feature vectors using the constructed grading assessment and treatment model for sepsis patients, assess the patient's sepsis verification level, and provide recommended treatment plans. The grading assessment and treatment model for sepsis patients is trained based on the patient's relevant data. Data storage module: used to store various data collected, processed and analyzed; Analysis and processing module: used to analyze and process the adjustment data of the treatment plan template to obtain the adoption and application index of the treatment plan; Display and interaction module: used to display various data and treatment plans, adjust treatment plans, and push reminders on the severity of sepsis in patients. Beneficial effects

[0012] 1. By obtaining the basic information, physiological indicators, severity, treatment plan and post-treatment data of sepsis patients, a database covering the entire diagnosis and treatment process is constructed, so that the model can learn the relationship between different patient characteristics and diagnosis and treatment results, improve the comprehensiveness of evaluation and prediction, and pre-process the original data by filtering invalid values, extracting feature vectors, etc., to eliminate noise interference, make the data input to the model more representative, avoid misjudgment due to data quality, and enhance the reliability of model training. By bundling feature vectors and treatment plans for model training, the system has intelligent analysis capabilities based on historical experience, providing data support for subsequent real-time evaluation and treatment recommendations for patients, and reducing dependence on the experience of a single doctor.

[0013] 2. Real-time collection of key indicators such as patient blood oxygen and blood flow parameters, combined with individual information such as age and weight, allows for rapid determination of sepsis severity through model matching, shortening diagnosis time. This is particularly suitable for the early and rapid intervention of sepsis. Real-time data is denoised and time-space aligned to ensure comparability of data from different sources and at different times, avoiding analytical biases caused by data heterogeneity and improving the accuracy of severity assessments. Patient data and severity are displayed in real time, with alerts pushed, helping medical staff quickly understand the condition and avoiding manual interpretation of data that could lead to omission of key information. This is particularly suitable for emergency response in intensive care settings.

[0014] 3. By matching the patient's feature vector with historical cases in the model, targeted treatment plans are generated to avoid "one-size-fits-all" treatment. This is especially suitable for patients with underlying diseases or the personalized needs of special populations. When multiple treatment plans are matched, they are sorted by the recovery effect of post-treatment indicators, and plans with fast onset and significant effects are recommended first. This assists doctors in quickly developing the optimal treatment path, reducing trial and error costs, and integrating a large number of historical treatment cases so that the system can learn from successful experiences, especially providing cross-case treatment ideas for rare or complex cases, making up for the lack of clinical experience of young doctors.

[0015] 4. Objectively evaluate the effectiveness of the plan through the "Adoption and Application Index" to avoid subjective judgment bias. For example, if a plan is frequently selected and adjusted few times, it indicates that it is highly applicable and can be retained first; otherwise, it is marked as requiring optimization to improve the overall quality of the plan library, record the details of the plan adjustment and post-treatment data, and update the plan library in the model in real time so that the system can adapt to individual differences in patients and dynamic changes in their condition, avoiding the lag of fixed plans. Ensure the timeliness of the recommended plan, and through the closed-loop process of "select → adjust → record → analyze → update", convert clinical experience into quantifiable model parameters, promote the transformation of sepsis diagnosis and treatment from experience-based to data-driven, and improve the consistency of diagnosis and treatment among different medical institutions.

[0016] 5. Input the complete diagnosis and treatment data of new patients into the model so that the model can continuously learn new case characteristics, enhance the ability to evaluate rare or variant cases, and avoid model failure due to outdated data. As sepsis treatment guidelines are updated or new therapies emerge, the system can automatically optimize the recommendation logic by incorporating new data to ensure that treatment recommendations are synchronized with the latest medical advances. By continuously accumulating data, the model can identify early subtle changes in indicators and provide early warning of the risk of sepsis worsening. Combined with historical case comparisons, it can reduce misdiagnosis or missed diagnosis caused by incomplete interpretation of indicators. It is especially suitable for early screening scenarios by non-critical care specialists. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a flow chart of the method. DETAILED DESCRIPTION

[0018] In order to make the technical solution of the present invention clearer, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments. Example 1

[0019] The first step is to build a hierarchical assessment and treatment model for sepsis patients through the hierarchical assessment and treatment module. The model is trained by obtaining sepsis patient information, indicator data, the severity of sepsis at the time, treatment plan, and post-treatment indicator data. The specific steps include: 1. Build a hierarchical assessment and treatment model for sepsis patients to obtain information on sepsis patients, indicator data, the severity of sepsis at the time, treatment plans, and post-treatment indicators; 2. Pre-process the acquired sepsis patient information, indicator data, severity of sepsis at the time, treatment plan, and post-treatment indicator data to filter out invalid data and extract the feature vectors of patient information, indicator data, and post-treatment indicator data; 3. The extracted patient information indicator data and the characteristic vector of the post-treatment indicator data are bundled with the corresponding treatment plan of the patient at that time to obtain the big data input into the sepsis patient grading assessment and treatment model, and the sepsis patient grading assessment and treatment model is trained.

[0020] Patient information includes age, height, and weight. Index data includes basic physiological data, blood oxygenation, and blood flow parameters required for dynamic monitoring of sepsis blood perfusion. Post-treatment index data includes basic physiological data, blood oxygenation, and blood flow parameters required for dynamic monitoring of sepsis blood perfusion, as well as corresponding timestamps. By obtaining basic information, physiological indicators, severity, treatment plan, and post-treatment data of sepsis patients, a database covering the entire diagnosis and treatment process is constructed, enabling the model to learn the relationship between different patient characteristics and diagnosis and treatment results, improving the comprehensiveness of evaluation and prediction. The raw data is preprocessed by filtering invalid values ​​and extracting feature vectors to eliminate noise interference, making the data input to the model more representative, avoiding misjudgments due to data quality, and enhancing the reliability of model training. By bundling feature vectors with treatment plans for model training, the system has intelligent analysis capabilities based on historical experience, providing data support for real-time evaluation and treatment recommendations for subsequent patients, and reducing reliance on the experience of a single doctor.

[0021] The second step is to collect the current sepsis patient information and indicator data through the data acquisition module, including measuring blood flow and blood oxygen, match the current sepsis patient with the sepsis patient hierarchical diagnosis and treatment model, determine the severity of sepsis in the current patient, and display the current sepsis patient information, indicator data and sepsis severity through the display interaction module. The specific steps include: 1. Collect information and indicator data of current sepsis patients, including measuring blood flow and blood oxygen, pre-process the real-time collected data, remove noise from the data, and align the data in time and space; 2. Extract the characteristic vectors of the current sepsis patient information and indicator data from the preprocessed data, transfer the extracted characteristic vectors to the sepsis patient grading assessment and diagnosis and treatment model for matching, and obtain the sepsis severity of the sepsis patient in the sepsis patient grading assessment and diagnosis and treatment model with the highest matching degree with the characteristic vector of the current sepsis patient; 3. Display the current information and indicator data of sepsis patients in real time, and push alerts on the severity of sepsis in current sepsis patients.

[0022] Real-time collection of key indicators such as patient blood oxygen and blood flow parameters, combined with individual information such as age and weight, allows for rapid determination of sepsis severity through model matching, shortening diagnosis time. This is particularly suitable for the early and rapid intervention of sepsis. Real-time data is denoised and time-space aligned to ensure comparability of data from different sources and at different times, avoiding analytical biases caused by data heterogeneity and improving the accuracy of severity assessments. Patient data and severity are displayed in real time, with alerts pushed, helping medical staff quickly grasp the condition and avoiding manual interpretation of data that could lead to omission of key information. This is particularly suitable for emergency response in intensive care settings.

[0023] The third step is to use the tiered evaluation and treatment module to recommend treatment options based on the current patient's information and indicator data, using the tiered evaluation and treatment model for sepsis patients. When recommending treatment options, the model matches the extracted feature vector of the current sepsis patient with the tiered evaluation and treatment model. If multiple treatment options are matched, they are sorted based on the patient's post-treatment indicators for each treatment option. The treatment options with the most significant recovery in the patient's post-treatment indicators and the fastest recovery time are ranked higher. By matching the patient's feature vector with historical case studies in the model, targeted treatment options are generated, avoiding a "one-size-fits-all" approach. This approach is particularly suitable for the personalized needs of patients with comorbidities or special populations. When multiple treatment options are matched, they are sorted by the effect of post-treatment indicator recovery, prioritizing options with rapid onset and significant results. This helps doctors quickly develop the optimal treatment path, reducing trial-and-error costs. By integrating a large number of historical treatment cases, the system can draw on successful experiences, providing cross-case treatment strategies, especially for rare or complex cases, to address the lack of clinical experience among young doctors.

[0024] The fourth step is to determine the selected treatment plan through the display interaction module, record the adjustment of the selected treatment plan through the data acquisition module, analyze the adjustment of the selected treatment plan through the analysis and processing module to obtain the adoption and application index of the treatment plan, determine the treatment measures for the treatment plan based on the adoption and application index of the treatment plan, and update the treatment plan to the hierarchical diagnosis and treatment model for sepsis patients based on the treatment measures of the treatment plan. The specific steps are as follows: 1. Select the proposed treatment plan from the recommended treatment plans as the initial treatment plan template; 2. Analyze and judge the initial treatment plan template, determine whether to adjust the initial treatment plan template, record the adjustment data of the initial treatment plan template, obtain the implementation treatment plan after evaluation, treat the current sepsis patient according to the implementation treatment plan, and record the treatment data of the current sepsis patient in real time; 3. Analyze and process the adjustment data of the recorded initial treatment plan template to obtain the adoption and application index of the treatment plan. If the adoption and application index is higher than the preset threshold, the treatment plan will be retained. Otherwise, the treatment plan will be marked as a treatment plan that needs to be evaluated, and the evaluation of the treatment plan will include adjustment and deletion operations.

[0025] When analyzing and processing the adjustment data of the recorded initial treatment plan template to obtain the adoption and application index of the treatment plan, the adoption and application index of the treatment plan is obtained by analyzing the number of times the treatment plan is selected as the initial treatment plan, the number of times the initial treatment plan is used after adjustment, the total number of treatment parameters of the initial treatment plan, and the number of treatment parameters adjusted each time of the initial treatment plan: 、 in, is the adoption index of the treatment plan, is the number of times the treatment regimen was selected as the initial treatment regimen, is the number of times the initial treatment regimen was adjusted. is the total number of treatment parameters of the initial treatment plan, The number of treatment parameters that are adjusted each time for this initial treatment plan. The bigger, The smaller, The larger the value of , the more frequently the solution is selected as the initial solution and the less adjustments are made. The larger the value, the greater the adoption index. The smaller it is, the larger the molecule is, which means that the original parameters of the treatment plan are more in line with the treatment needs and the larger the adoption and application index is.

[0026] The "Adoption Index" is used to objectively evaluate the effectiveness of the plan and avoid subjective judgment bias. For example, if a plan is frequently selected and adjusted rarely, it indicates that it is highly applicable and can be retained first; otherwise, it is marked as requiring optimization to improve the overall quality of the plan library, record plan adjustment details and post-treatment data, and update the plan library in the model in real time so that the system can adapt to individual patient differences and dynamic changes in the condition, avoiding the lag of fixed plans. To ensure the timeliness of the recommended plan, through the closed-loop process of "select → adjust → record → analyze → update", clinical experience is converted into quantifiable model parameters, promoting the transformation of sepsis diagnosis and treatment from experience-based to data-driven, and improving the consistency of diagnosis and treatment between different medical institutions.

[0027] The fifth step is to input the current patient's information, indicator data, confirmed sepsis severity, treatment plan, and post-treatment indicator data into the sepsis patient tiered diagnosis and treatment model. The tiered assessment and treatment module then refines and trains the model. Inputting the model's complete diagnosis and treatment data for new patients allows it to continuously learn new case characteristics, enhancing its ability to assess rare or variant cases and avoiding model failure due to outdated data. As sepsis treatment guidelines are updated or new therapies emerge, the system can automatically optimize its recommendation logic by incorporating new data, ensuring that treatment recommendations are synchronized with the latest medical advances. Through continuous data accumulation, the model can identify subtle early indicator changes, providing early warning of the risk of sepsis worsening. Combined with historical case comparisons, this reduces misdiagnoses or missed diagnoses caused by incomplete interpretation of indicators. This is particularly suitable for early screening scenarios by non-critical care specialists. Example 2

[0028] When calculating the adoption index, When: , at this time, if If the preset threshold is exceeded, the treatment plan is retained; if If the preset threshold is exceeded, the treatment plan will be marked as a treatment plan that needs to be evaluated, and the evaluation of the treatment plan will include adjustment and deletion.

[0029] The above-described embodiments merely illustrate several implementations of the present invention, and while their descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

Claims

1. A method for dynamic monitoring of sepsis blood perfusion by combining blood oxygen and blood flow parameters, characterized in that: The specific steps include: S1. Build a hierarchical assessment and treatment model for sepsis patients, obtain sepsis patient information, indicator data, the severity of sepsis at the time, treatment plan, and post-treatment indicator data to train the model; S2. Collect information and indicator data of the current sepsis patient, including measuring blood flow and blood oxygen, match the current sepsis patient with the sepsis patient hierarchical diagnosis and treatment model, determine the severity of the current sepsis patient, and display the current sepsis patient information, indicator data, and sepsis severity; S3. Based on the current patient information and indicator data, provide a recommended treatment plan through the hierarchical diagnosis and treatment model for sepsis patients; S4. Determine a selected treatment plan, record adjustments to the selected treatment plan, analyze the adjustments to the selected treatment plan using an intelligent single-chip computer to determine an adoption and application index for the treatment plan, determine treatment measures for the treatment plan based on the adoption and application index, and update the treatment plan to the hierarchical diagnosis and treatment model for sepsis patients based on the treatment measures for the treatment plan; S5. Input the current patient's information, indicator data, confirmed severity of sepsis, treatment plan, and indicator data after treatment into the hierarchical diagnosis and treatment model for sepsis patients to improve the model training.

2. The method for dynamic monitoring of sepsis blood perfusion by combining blood oxygen and blood flow parameters according to claim 1, characterized in that: The patient information includes the patient's age, height, and weight. The index data includes the patient's basic physiological data, blood oxygenation, and blood flow parameters required for dynamic monitoring of sepsis blood perfusion. The post-treatment index data includes the patient's basic physiological data after treatment, blood oxygenation, and blood flow parameters required for dynamic monitoring of sepsis blood perfusion, as well as the corresponding timestamp.

3. The method for dynamic monitoring of sepsis blood perfusion by combining blood oxygen and blood flow parameters according to claim 1, characterized in that: The step S1 specifically includes the following steps: S11. Build a hierarchical assessment and treatment model for sepsis patients to obtain information on sepsis patients, indicator data, the severity of sepsis at the time, treatment plans, and post-treatment indicator data; S12. Preprocess the acquired sepsis patient information, indicator data, severity of sepsis at the time, treatment plan, and post-treatment indicator data to filter out invalid data and extract feature vectors of the patient information, indicator data, and post-treatment indicator data; S13. Bundle the extracted patient information indicator data and the characteristic vector of the post-treatment indicator data with the treatment plan corresponding to the patient at that time to obtain the big data and input it into the sepsis patient grading assessment and diagnosis and treatment model to train the sepsis patient grading assessment and diagnosis and treatment model.

4. The method for dynamic monitoring of sepsis blood perfusion by combining blood oxygen and blood flow parameters according to claim 1, characterized in that: The step S2 specifically includes the following steps: S21. Collect information and indicator data of current sepsis patients, including measuring blood flow and blood oxygen, pre-process the real-time collected data, remove noise from the data, and perform spatiotemporal alignment of various data; S22. Extracting characteristic vectors of the current sepsis patient information and indicator data from the preprocessed data, transmitting the extracted characteristic vectors to a sepsis patient grading assessment and diagnosis and treatment model for matching, and obtaining the sepsis severity of the sepsis patient in the sepsis patient grading assessment and diagnosis and treatment model that has the highest matching degree with the characteristic vector of the current sepsis patient; S23. Display the current sepsis patient information and indicator data in real time, and push alerts on the severity of sepsis in the current sepsis patient.

5. The method for dynamic monitoring of sepsis blood perfusion by combining blood oxygen and blood flow parameters according to claim 4, characterized in that: In step S3, when giving a recommended treatment plan, a match will be made based on the extracted feature vector of the current sepsis patient using the sepsis patient hierarchical assessment and treatment model. If multiple treatment plans are matched, they will be sorted according to the patient's post-treatment indicators for each treatment plan. The treatment plan with more obvious recovery of the patient's post-treatment indicator data and faster recovery time will be ranked higher.

6. The method for dynamic monitoring of sepsis blood perfusion by combining blood oxygen and blood flow parameters according to claim 5, characterized in that: The specific steps of step S4 are: S41. Select a treatment plan to be implemented from the suggested treatment plans as an initial treatment plan template; S42: Determine whether to adjust the initial treatment plan template by analyzing the initial treatment plan template, record the adjustment data of the initial treatment plan template, obtain an implementation treatment plan after the analysis, treat the current sepsis patient according to the implementation treatment plan, and record the treatment data of the current sepsis patient in real time; S43. Analyze and process the adjustment data of the recorded initial treatment plan template to obtain the adoption and application index of the treatment plan. If the adoption and application index is higher than the preset threshold, the treatment plan is retained. Otherwise, the treatment plan is marked as a treatment plan that needs to be evaluated. The evaluation of the treatment plan includes adjustment and deletion operations.

7. The method for dynamic monitoring of sepsis blood perfusion by combining blood oxygen and blood flow parameters according to claim 6, characterized in that: In step S43, when analyzing and processing the adjustment data of the recorded initial treatment plan template to obtain the adoption and application index of the treatment plan, the adoption and application index of the treatment plan is obtained by analyzing the number of times the treatment plan is selected as the initial treatment plan, the number of times the initial treatment plan is used after adjustment, the total number of treatment parameters of the initial treatment plan, and the number of treatment parameters adjusted each time for the initial treatment plan: 、 in, is the adoption index of the treatment plan, is the number of times the treatment regimen was selected as the initial treatment regimen, is the number of times the initial treatment regimen was adjusted. is the total number of treatment parameters of the initial treatment plan, The number of treatment parameters that are adjusted each time for this initial treatment plan.

8. A sepsis blood perfusion dynamic monitoring system combining blood oxygen and blood flow parameters, characterized by: The system is used to implement the method for dynamic monitoring of sepsis blood perfusion in combination with blood oxygen and blood flow parameters as described in any one of claims 1 to 7, comprising: a data acquisition module, a data processing module, a graded evaluation and diagnosis module, a data storage module, an analysis and processing module, and a display and interaction module; Data acquisition module: used to collect various data during the use of the monitoring system; Data processing module: used to remove noise and perform spatiotemporal alignment preprocessing on the patient's relevant data, extract the feature vector of the preprocessed data, and input the extracted feature vector into the hierarchical evaluation and treatment module for analysis and processing; Grading Assessment and Treatment Module: This module is used to analyze and process the input feature vectors using the constructed grading assessment and treatment model for sepsis patients, assess the patient's sepsis verification level, and provide recommended treatment plans. The grading assessment and treatment model for sepsis patients is trained based on the patient's relevant data. Data storage module: used to store various data collected, processed and analyzed; Analysis and processing module: used to analyze and process the adjustment data of the treatment plan template to obtain the adoption and application index of the treatment plan; Display and interaction module: used to display various data and treatment plans, adjust treatment plans, and push reminders on the severity of sepsis in patients.

Images