Data processing method and device based on hemodynamic indexes and machine learning

Abstract

The present application relates to a kind of data processing method and device based on hemodynamic index and machine learning, method includes the following steps: S1, state data is obtained;S2, the VIS value is calculated;S3, the BPRI value is calculated;S4, evaluation result is calculated: feature vector is input into the machine learning model of preliminary training, the feature vector includes the BPRI value, the VIS value and state feature, by the machine learning model output target object in the state evaluation index of preset period.In one aspect, by introducing the BPRI that is made of MAP and VIS value compound, and combining VIS value, blood pressure level and circulation support intensity are simultaneously included in quantization, the responsiveness of target object to vasoactive drug and / or positive inotropic drug is comprehensively reflected, the effective integration of hemodynamic responsiveness dimension is realized;In another aspect, BPRI value and VIS value can be directly calculated, reduce the dependence of machine learning model on laboratory complex test data.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of medical data processing and clinical decision support technology, and more specifically, to a data processing method, apparatus, electronic device and readable storage medium based on hemodynamic indicators and machine learning. Background Technology

[0002] Cardiac arrest (CA) has an extremely high mortality rate. Although some individuals can achieve spontaneous circulation recovery (ROSC) after intervention, the ensuing ischemia-reperfusion injury and systemic inflammatory response often lead to persistent hemodynamic instability. Therefore, objective, early, and rapid quantitative assessment of the risk status of individuals is of great significance for supporting systemic decision-making and the rational allocation of critical care resources.

[0003] Current mainstream data assessment methods mostly use absolute blood pressure values ​​such as mean arterial pressure (MAP) as the sole data input feature. However, in real intensive care settings, the required dosage of vasoactive drugs or positive inotropic drugs (i.e., circulatory support intensity) varies significantly among individuals to maintain the same blood pressure level. Existing methods fail to integrate "absolute blood pressure values" with "drug support intensity" in modeling, and relying solely on blood pressure values ​​cannot accurately reflect the actual response capacity of the body's cardiovascular system to drugs.

[0004] Furthermore, because existing risk assessment models do not incorporate quantitative records of circulatory support intensity into their core feature space, they struggle to accurately characterize the nonlinear pathological changes in organisms under different intervention strategies. This lack of feature information often leads traditional assessment models to overestimate or underestimate the actual risk to the target population, failing to provide precise risk stratification and status warnings within the very early post-resuscitation time window.

[0005] In conclusion, there is an urgent need in this field to develop a completely new data processing method. Summary of the Invention

[0006] One objective of this invention is to provide a new technical solution for data processing methods, devices, electronic devices, and readable storage media based on hemodynamic indicators and machine learning, which can deeply integrate blood pressure levels with the intensity of drug circulation support and achieve earlier, more objective, and more accurate status assessment by extracting more representative hemodynamic response features.

[0007] According to a first aspect of the present invention, a data processing method based on hemodynamic parameters and machine learning is provided, comprising the following steps: S1, acquiring status data: acquiring status data within a preset time window after emergency treatment and sudden cardiac arrest in an inpatient intensive care unit from a medical database, wherein the status data includes at least mean arterial pressure and historical medication dosage data, the historical medication dosage data corresponding to vasoactive drugs and / or positive inotropic drugs; S2, calculating VIS value: converting the historical medication dosage data into a VIS value according to a preset vasoactive-positive inotropic scoring formula; S3, calculating BPRI value: calculating the BPRI value based on the mean arterial pressure and the VIS value, wherein BPRI = MAP / S4. Calculate the evaluation result: Input the feature vector into the pre-trained machine learning model. The feature vector includes the BPRI value, the VIS value, and the state feature. The machine learning model outputs the state evaluation index of the target object within a preset time period. The state feature consists of several core features selected from the candidate feature set by the feature selection algorithm. The core features include at least one of the following: laboratory test data, physiological monitoring data, clinical scale evaluation data, and basic attributes and medical history data.

[0008] Optionally, the machine learning model is an ensemble learning model based on gradient boosting decision trees.

[0009] Optionally, the machine learning model adopts the XGBoost model, and the feature selection algorithm adopts the Boruta algorithm.

[0010] Optionally, the core features include: anion gap, pH, sodium, SOFA score, creatinine, body temperature, age, and whether congestive heart failure is present.

[0011] Optionally, the data processing method based on hemodynamic indicators and machine learning further includes the following step: using the SHAP method to perform interpretability analysis on the state assessment index output by the machine learning model.

[0012] Optionally, the data processing method based on hemodynamic indicators and machine learning further includes the following steps: comparing the BPRI value with a first preset threshold, and / or comparing the VIS value with a second preset threshold; classifying the target object according to the comparison results, wherein when the BPRI value is lower than the first preset threshold and / or the VIS value is higher than the second preset threshold, a high-risk warning message is output.

[0013] Optionally, the status data includes one or more of the following: demographic attribute data, physiological characteristics data, past medical history records, laboratory test indicators, and organ function score data.

[0014] According to a second aspect of the present invention, a data processing apparatus based on hemodynamic indicators and machine learning is provided, comprising: a data acquisition module configured to acquire status data of patients in an inpatient intensive care unit within a preset time window after emergency treatment and immediate arrest from a medical database, the status data including at least mean arterial pressure and historical medication dosage data, the historical medication dosage data corresponding to vasoactive drugs and / or positive inotropic drugs; a calculation module configured to convert the historical medication dosage data into VIS values ​​according to a preset vasoactive-positive inotropic scoring formula, and to calculate a BPRI value based on the mean arterial pressure and the VIS value, wherein BPRI = MAP / VIS; and an evaluation module configured to input the BPRI value, the VIS value, and a feature vector into a pre-trained machine learning model, the machine learning model outputting a status evaluation index of the target object within a preset time period, the feature vector consisting of several core features selected from a candidate feature set by a feature selection algorithm, the core features including a combination of at least two types of data: laboratory test data, physiological monitoring data, clinical scale evaluation data, and basic attribute and medical history data.

[0015] According to a third aspect of the present invention, an electronic device is provided, comprising: a processor and a memory, wherein computer program instructions are stored in the memory, wherein when the computer program instructions are executed by the processor, the processor causes the processor to perform any of the methods described above.

[0016] According to a fourth aspect of the present invention, a computer-readable storage medium is provided, characterized in that the computer-readable storage medium stores a computer program, which, when executed by a processor, causes the processor to perform any of the methods described above.

[0017] One beneficial effect of this invention is that, according to the data processing method based on hemodynamic indicators and machine learning according to this invention, on the one hand, by introducing BPRI, which is composed of MAP and VIS values, and combining it with VIS values, blood pressure level and circulatory support strength are simultaneously incorporated into the quantification, comprehensively reflecting the target subject's responsiveness to vasoactive drugs and / or positive inotropic drugs, thus making up for the shortcomings of existing methods that only focus on absolute blood pressure values, and achieving effective integration of hemodynamic responsiveness dimensions; on the other hand, BPRI and VIS values ​​can be directly calculated, reducing the dependence of machine learning models on complex laboratory test data, making it suitable for low-latency near real-time assessment in the early time window after the target subject's final vital signs recover (or data is accessed by the system), effectively overcoming the defects of traditional data assessment models caused by waiting and running electrophysiological data or alarm markers.

[0018] Other features and advantages of the present invention will become clear from the following detailed description of exemplary embodiments of the invention with reference to the accompanying drawings. Attached Figure Description

[0019] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments of the invention and, together with their description, serve to explain the principles of these embodiments.

[0020] Figure 1 A flowchart illustrating a data processing method based on hemodynamic parameters and machine learning according to some embodiments is shown; Figure 2 A schematic diagram of the hardware structure of an electronic device according to some embodiments is shown.

[0021] Figure Labels Electronic device 100; processor 110; memory 120. Detailed Implementation

[0022] Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that, unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps set forth in these embodiments do not limit the scope of the invention.

[0023] The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the invention or its application or use.

[0024] Techniques, methods, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and equipment should be considered part of the specification.

[0025] In all the examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0026] It should be noted that similar labels and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be discussed further in subsequent figures.

[0027] The following is a flowchart illustrating the data processing method based on hemodynamic indicators and machine learning according to an embodiment of this application, with reference to the accompanying drawings.

[0028] like Figure 1 As shown, the data processing method based on hemodynamic parameters and machine learning according to an embodiment of the present invention includes the following steps: S1. Obtain Status Data: Obtain status data of patients who have experienced sudden cardiac arrest within a preset time window after admission to the intensive care unit (ICU) from a medical database. The status data includes at least mean arterial pressure (MAP) and historical medication dosage data, which corresponds to vasoactive drugs and / or positive inotropic drugs. In other words, historical status data of a target subject (e.g., an individual with a history of sudden cardiac arrest) within a preset time window (e.g., the first 24 hours after admission) after admission to the ICU can be obtained from a preset medical database (such as an electronic medical record system (EMR) or a critical care medicine information system (MIMIC)) via computer equipment or a server. In embodiments of this disclosure, the status data includes at least the recorded monitoring of mean arterial pressure (MAP) and historical medication dosage data for various vasoactive drugs and / or positive inotropic drugs.

[0029] Optionally, the status data includes one or more of the following: demographic attribute data, physiological characteristics data, past medical history records, laboratory test indicators, and organ function score data (such as sequencing organ estimation score SOFA). This enables the machine learning model to deeply explore and learn the high-order correlations between multi-source data, improve the accuracy of subsequent data processing and multi-dimensional evaluation of the model, and thus improve the accuracy and reliability of the final output status evaluation index.

[0030] S2. Calculate VIS value: Based on the preset vasoactivity-positive inotropic score (VIS) calculation formula, convert historical medication dosage data into VIS value. In other words, the VIS value is calculated based on historical medication dosage data according to the preset vasoactivity-positive inotropic score calculation formula.

[0031] S3. Calculate the BPRI value: Based on the mean arterial pressure (MAP) and VIS value, calculate the BPRI (Blood Pressure Response Index), where BPRI = MAP / VIS. BPRI is a composite index that combines MAP, which reflects blood pressure level, with VIS, which reflects circulatory support strength. It assesses the blood pressure responsiveness of a target subject to vasoactive drug therapy, primarily reflecting hemodynamic responsiveness under vasoactive drug support, rather than simply the drug dose or absolute blood pressure value.

[0032] S4. Calculate the evaluation results: Input the feature vector into the pre-trained machine learning model. The feature vector includes BPRI value, VIS value, and state features. The machine learning model outputs a state evaluation index for the target object within a preset time period. The state features consist of several core features selected from a candidate feature set by a feature selection algorithm. These core features include at least one of the following: laboratory test data, physiological monitoring data, clinical scale assessment data, and basic attributes and medical history data. In other words, the BPRI value, VIS value, and state features are input into the pre-trained machine learning model, which performs feature mapping and calculation, outputting a state evaluation index for the target object within a preset time period (e.g., 28 days). The state evaluation index is a statistical probability value ranging from 0 to 1, representing the objective probability of the target object experiencing an adverse prognosis (such as severe deterioration or death) within the preset time period. This index serves only as reference information for the computer output data, to facilitate the subsequent system in triggering appropriate levels of early warning mechanisms or assisting in the scheduling and planning of medical resources, rather than directly making a clinical diagnosis.

[0033] According to one embodiment of the present invention, the machine learning model is an ensemble learning model based on gradient boosting decision trees. It can not only utilize its built-in sparse expansion algorithm to supplement the historical state data in intensive care scenarios with manual imputation, avoiding the introduction of noise and significantly improving the generalization ability of the machine learning model in real medical environments; but also effectively preserve the original distribution characteristics and nonlinear interaction mapping of hemodynamic indicators, improving the accuracy of the state assessment index; and can output the dynamic contribution of each core feature to the state assessment index in real time with low computing power consumption, providing a transparent and reliable data source for decision-making and early warning.

[0034] Preferably, the machine learning model adopts the XGBoost model. Specifically, the Boruta algorithm is used for feature selection to obtain core features, and the XGBoost model is used to construct the machine learning model. This achieves good discriminative ability on the training set, validation set, and external test set (independent source data) (e.g., AUROC approximately 0.806 on the training set, approximately 0.783 on the validation set, and approximately 0.758 on the test set), demonstrating advantages such as improved accuracy of the state evaluation index and strong generalization ability.

[0035] In some specific embodiments of the present invention, the core features include: anion gap, pH, sodium, SOFA score, creatinine, body temperature, age, and whether congestive heart failure (CHF) is present. The selected core features comprehensively characterize the state of the target object from multiple dimensions such as baseline status, internal environment homeostasis, organ function and circulatory status, thereby improving the accuracy of the state assessment index.

[0036] In this embodiment, a multidimensional feature space was constructed covering laboratory test data (anion gap, pH, sodium, creatinine), physiological monitoring data (body temperature), clinical scale assessment data (SOFA score), and basic attribute and medical history data (age, congestive heart failure). By using multidimensional feature cross-mapping, the machine learning model can more comprehensively capture the latent data patterns of the target object in the final state, which greatly improves the accuracy of the state assessment index.

[0037] According to one embodiment of the present invention, the data processing method based on hemodynamic indicators and machine learning further includes the following steps: using the SHAP (SHapley Additive exPlanations) method to perform interpretability analysis on the state assessment index output by the machine learning model, and outputting the contribution and direction of each feature to the state assessment index. In this embodiment, the SHAP method can perform interpretability analysis on the output results of the machine learning model, clearly presenting the contribution and direction of each feature to the state assessment index (such as decreased BPRI value, increased VIS value, increased anion gap, decreased pH, etc.), enabling users (such as medical staff or system administrators) to quickly identify which abnormal data indicators (such as excessively low pH or a sudden increase in VIS score) caused the warning, thereby achieving accurate data tracing and corrective action.

[0038] In some specific embodiments of the present invention, the data processing method based on hemodynamic indicators and machine learning further includes the following steps: comparing the BPRI value with a first preset threshold, and / or comparing the VIS value with a second preset threshold; classifying the target object according to the comparison results, wherein when the BPRI value is lower than the first preset threshold and / or the VIS value is higher than the second preset threshold, a high-risk warning is output. For example, the first preset threshold is 4.75, and the second preset threshold is 15.2. Risk stratification of the target object is performed based on the first and second preset thresholds. When the BPRI value is lower than the first preset threshold and / or the VIS value is higher than the second preset threshold, a high-risk warning is output, which can provide decision support information and enable rapid and objective identification of the target object.

[0039] The present invention also discloses a data processing device based on hemodynamic indicators and machine learning, comprising: a data acquisition module, a calculation module and an evaluation module.

[0040] Specifically, the data acquisition module is configured to retrieve status data of patients requiring emergency care and immediate medical attention within a preset time window from a medical database. This status data includes at least mean arterial pressure (MAP) and historical medication dosage data, which corresponds to vasoactive drugs and / or positive inotropic drugs. The calculation module is configured to convert historical medication dosage data into VIS values ​​based on a preset vasoactive-positive inotropic scoring formula, and to calculate the BPRI value based on MAP and VIS values, where BPRI = MAP / VIS. The evaluation module is configured to input the BPRI value, VIS value, and relevant feature vectors into a pre-trained machine learning model. The machine learning model outputs a status evaluation index of the target object within a preset time period. The feature vector consists of several core features selected from a candidate feature set using a feature selection algorithm. These core features include combinations of at least two types of data: laboratory test data, physiological monitoring data, clinical scale assessment data, and basic attribute and medical history data.

[0041] The data processing method and apparatus based on hemodynamic indicators and machine learning according to embodiments of the present invention, on the one hand, introduces BPRI, a composite of MAP and VIS values, and combines it with VIS values ​​to simultaneously quantify blood pressure levels and circulatory support strength, comprehensively reflecting the target subject's responsiveness to vasoactive drugs and / or positive inotropic drugs. This overcomes the shortcomings of existing methods that only focus on absolute blood pressure values, achieving effective integration of hemodynamic responsiveness dimensions. On the other hand, BPRI and VIS values ​​can be directly calculated, reducing the dependence of machine learning models on complex laboratory test data. This makes it suitable for low-latency, near-real-time assessment within the early time window after the target subject's final recovery of vital signs (or data access to the system), effectively overcoming the shortcomings of traditional data assessment models that suffer from warning lag due to waiting for and running electrophysiological data or alarm markers. The data processing method and apparatus based on hemodynamic indicators and machine learning according to embodiments of the present invention can be integrated into bedside monitors, clinical decision support systems (CDSS) or hospital information systems (HIS), achieving diverse forms and facilitating industrial application and promotion.

[0042] The present invention also provides a schematic diagram of the hardware structure of an electronic device.

[0043] like Figure 2 As shown, the electronic device 100 includes a processor 110 and a memory 120. The memory 120 is used to store an executable computer program, and the processor 110 is used to execute a data processing method based on hemodynamic indicators and machine learning, as described in any of the above method embodiments, under the control of the computer program.

[0044] The modules of the data processing device based on hemodynamic indicators and machine learning described above can be implemented by the processor 110 in this embodiment executing the computer program stored in the memory 120, or they can be implemented by other structures, which are not limited here.

[0045] This invention can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of the invention.

[0046] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example—but not limited to—electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination thereof. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0047] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0048] The computer program instructions used to perform the operations of this invention may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing state information from the computer-readable program instructions. This electronic circuitry can execute the computer-readable program instructions to implement various aspects of the invention.

[0049] Various aspects of the present invention are described herein with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-readable program instructions.

[0050] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that, when executed by the processor of the computer or other programmable data processing apparatus, they create means for implementing the functions / actions specified in one or more blocks of the flowchart and / or block diagram. These computer-readable program instructions can also be stored in a computer-readable storage medium that causes a computer, programmable data processing apparatus, and / or other device to operate in a particular manner; thus, the computer-readable medium storing the instructions comprises an article of manufacture that includes instructions for implementing aspects of the functions / actions specified in one or more blocks of the flowchart and / or block diagram.

[0051] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to perform the functions / actions specified in one or more boxes of a flowchart and / or block diagram.

[0052] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of an instruction containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions. It will be known to those skilled in the art that implementation in hardware, implementation in software, and implementation using a combination of software and hardware are equivalent.

[0053] The various embodiments of the present invention have been described above. These descriptions are exemplary and not exhaustive, and are not limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or technical improvements to the embodiments in the market, or to enable others skilled in the art to understand the embodiments disclosed herein. The scope of the invention is defined by the appended claims.

Claims

1. A data processing method based on hemodynamic parameters and machine learning, characterized in that, Includes the following steps: S1. Obtain status data: Obtain the status data of patients in the inpatient intensive care unit within a preset time window after emergency treatment and emergency shutdown from the medical database. The status data includes at least mean arterial pressure and historical drug dosage data, and the historical drug dosage data corresponds to vasoactive drugs and / or positive inotropic drugs. S2. Calculate VIS value: Based on the preset vasoactivity-positive inotropic score calculation formula, convert the historical medication dosage data into VIS value; S3. Calculate the BPRI value: The BPRI value is calculated based on the mean arterial pressure and the VIS value, where BPRI = MAP / VIS; S4. Calculate the evaluation results: Input the feature vector into the pre-trained machine learning model. The feature vector includes the BPRI value, the VIS value, and the state features. The machine learning model outputs the state evaluation index of the target object within a preset time period. The state features are composed of several core features selected from the candidate feature set by the feature selection algorithm. The core features include at least one of the following: laboratory test data, physiological monitoring data, clinical scale evaluation data, and basic attributes and medical history data.

2. The data processing method based on hemodynamic parameters and machine learning according to claim 1, characterized in that, The machine learning model is an ensemble learning model based on gradient boosting decision trees.

3. The data processing method based on hemodynamic parameters and machine learning according to claim 2, characterized in that, The machine learning model used is the XGBoost model, and the feature selection algorithm used is the Boruta algorithm.

4. The data processing method based on hemodynamic parameters and machine learning according to claim 1, characterized in that, The core characteristics include: anion gap, pH, sodium, SOFA score, creatinine, body temperature, age, and whether congestive heart failure is present.

5. The data processing method based on hemodynamic parameters and machine learning according to claim 1, characterized in that, It also includes the following steps: The interpretability of the state evaluation index output by the machine learning model is analyzed using the SHAP method.

6. The data processing method based on hemodynamic parameters and machine learning according to claim 1, characterized in that, It also includes the following steps: The BPRI value is compared with a first preset threshold, and / or the VIS value is compared with a second preset threshold; The target object is classified into different states based on the comparison results. When the BPRI value is lower than the first preset threshold and / or the VIS value is higher than the second preset threshold, a high-risk warning message is output.

7. The data processing method based on hemodynamic parameters and machine learning according to claim 1, characterized in that, The status data includes one or more of the following: demographic attribute data, physiological signs data, past medical history records, laboratory test indicators, and organ function score data.

8. A data processing device based on hemodynamic parameters and machine learning, characterized in that, include: The data acquisition module is configured to acquire status data of emergency-stopped patients in the inpatient intensive care unit within a preset time window from a medical database. The status data includes at least mean arterial pressure and historical drug dosage data, and the historical drug dosage data corresponds to vasoactive drugs and / or positive inotropic drugs. The calculation module is configured to convert the historical drug dosage data into VIS values ​​according to a preset vasoactivity-positive inotropic score calculation formula, and to calculate the BPRI value based on the mean arterial pressure and the VIS value, where BPRI = MAP / VIS; The evaluation module is configured to input the BPRI value, the VIS value, and the feature vector into a pre-trained machine learning model, and the machine learning model outputs the state evaluation index of the target object within a preset time period. The feature vector consists of several core features selected from the candidate feature set by a feature selection algorithm. The core features include a combination of at least two types of data: laboratory test data, physiological monitoring data, clinical scale evaluation data, and basic attribute and medical history data.

9. An electronic device, characterized in that, include: A processor and a memory, wherein computer program instructions are stored in the memory, wherein when the computer program instructions are executed by the processor, the processor causes the processor to perform the method of any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, causes the processor to perform the method of any one of claims 1-7.

Images