Pulmonary embolism adverse event risk assessment method and device, equipment and storage medium

Abstract

The invention discloses a pulmonary embolism adverse event risk assessment method, device and equipment and a storage medium, a functional pulmonary embolism adverse event risk assessment model is constructed, three objective parameters of age, heart rate and blood oxygen partial pressure are directly incorporated into the model, weights of the age, the heart rate and the blood oxygen partial pressure are obtained through objective derivation (non-manual setting), and the risk assessment model is used for evaluating the risk of the pulmonary embolism adverse event. According to the method, the dependence on doctor experience is thoroughly eliminated, assessment deviation caused by subjective factors is avoided, the functional risk assessment model has continuity, and the problem that the assessment results of two patients with similar parameters are greatly different in the prior art is avoided.

Description

Technical Field

[0001] This invention belongs to the field of pulmonary embolism prognosis technology, and particularly relates to a method, device, equipment and storage medium for assessing the risk of adverse events of pulmonary embolism. Background Technology

[0002] Pulmonary embolism (PE) is a common and highly fatal / disabling condition in emergency departments, requiring timely and accurate treatment. Death from PE typically occurs within weeks of diagnosis. Short-term mortality rates for PE vary significantly: in many patients with non-severe PE, the mortality rate is less than 2%, while for those experiencing cardiopulmonary arrest, the mortality rate can exceed 95%. Therefore, the prognosis of acute PE (referring to the patient's recovery after treatment or intervention and the prediction of future health status) has high clinical value.

[0003] Several prognostic models for acute pulmonary embolism (PE) exist in clinical practice. Among them, the Pulmonary Embolism Severity Index (PESI) and its simplified version (sPESI), which integrate the severity of pulmonary embolism and comorbidities, are the most widely validated clinical scores to date. However, these models often include subjective variables, such as a history of malignant tumors and chronic cardiopulmonary diseases. When taking medical histories based on these variables, physicians may make incorrect assessments of the mortality risk for some patients. Computed tomography pulmonary angiography (CTPA) is considered the gold standard for diagnosing pulmonary embolism. However, the clinical practice of identifying pulmonary embolism patients solely based on the International Classification of Diseases (ICD) codes may include some patients with similar symptoms to those with pulmonary embolism but who do not actually have pulmonary embolism. Therefore, it is impossible to accurately predict the 30-day mortality rate of patients diagnosed with CTPA.

[0004] Given the limitations of current prognostic models, an objective, accurate, and simple clinical prognostic model is needed to help clinicians assess the risk of pulmonary embolism patients and support treatment decisions. For example, low-risk patients can be discharged earlier or receive comprehensive outpatient management, while high-risk patients require closer monitoring and aggressive treatment.

[0005] A 2021 study published in *EClinical Medicine* developed a mortality risk scoring model called PERFORM for patients diagnosed with pulmonary embolism via computed tomography angiography (CTPA). This model does not rely on the patient's medical history but assesses the risk of death within 30 days using objective parameters such as age, heart rate, and partial pressure of oxygen. Clinical data validated this model, showing better performance and more objective assessment parameters compared to traditional methods. However, the model has limitations. The segmentation of assessment parameters relies heavily on physician experience (e.g., age is divided into four segments: <65, ≥65 and <75, ≥75 and <85, ≥85), and the scoring indicators are discontinuous after segmentation. This can lead to significant differences in assessment results between two patients with similar parameters (e.g., 84-year-old and 85-year-old patients may have very different scores, while 75-year-old and 84-year-old patients may have very similar scores), limiting the accuracy of the PERFORM model. Summary of the Invention

[0006] Based on this, and in response to the aforementioned technical problems, a method, apparatus, device, and storage medium for assessing the risk of adverse events of pulmonary embolism are provided.

[0007] The technical solution adopted in this invention is as follows: As a first aspect of the present invention, a method for assessing the risk of adverse events of pulmonary embolism is provided, characterized in that it includes: Establish a risk assessment model for adverse events of pulmonary embolism in patients:

[0008] Where G represents the risk assessment index for adverse events of pulmonary embolism in patients, and k age k heartrate k PO2 The weights of patient age, heart rate, and blood oxygen partial pressure represent the impact of adverse event risk. The values ​​of these three factors are greater than 0 and less than 1, and satisfy the following conditions: ; Assume k heartrate and k PO2 The relative weighting coefficient k heartrate_PO2 The value of is greater than 0 and less than 1 and satisfies:

[0009] Then there is Therefore, the risk assessment model for adverse events of pulmonary embolism in patients is simplified as follows: ; Generate k within the range of values. age and k heartrate_PO2For each combination pair, the evaluation index G corresponding to each case sample is calculated according to the simplified model, and the AUC is calculated by plotting the ROC curve to obtain the AUC corresponding to each combination pair. The function Q is fitted by the one-to-one correspondence between each combination and AUC. ~ (k age , k heartrate_PO2 ), in k age and k heartrate_PO2 Find the k corresponding to the maximum function value within the range of values. age and k heartrate_PO2 The weight values ​​of each item in the evaluation model are obtained. The target patient's age, heart rate, and blood oxygen partial pressure are input into the assessment model to obtain the risk assessment index value of pulmonary embolism adverse events for the target patient. The higher the value, the higher the risk of pulmonary embolism adverse events.

[0010] As a second aspect of the present invention, a pulmonary embolism adverse event risk assessment device is provided, characterized in that it comprises: The first module is used to establish a risk assessment model for adverse events of pulmonary embolism in patients:

[0011] Where G represents the risk assessment index for adverse events of pulmonary embolism in patients, and k age k heartrate k PO2 The weights of patient age, heart rate, and blood oxygen partial pressure represent the impact of adverse event risk. The values ​​of these three factors are greater than 0 and less than 1, and satisfy the following conditions: ; The second module is used to assume k heartrate and k PO2 The relative weighting coefficient k heartrate_PO2 The value of is greater than 0 and less than 1 and satisfies:

[0012] Then there is Therefore, the risk assessment model for adverse events of pulmonary embolism in patients is simplified as follows: ; The third module is used to generate k within the range of values. age and k heartrate_PO2 For each combination pair, the evaluation index G corresponding to each case sample is calculated according to the simplified model, and the AUC is calculated by plotting the ROC curve to obtain the AUC corresponding to each combination pair. The fourth module is used to fit the function Q through the one-to-one correspondence between each combination and AUC. ~ (k age, k heartrate_PO2 ), in k age and k heartrate_PO2 Find the k corresponding to the maximum function value within the range of values. age and k heartrate_PO2 The weight values ​​of each item in the evaluation model are obtained. The fifth module is used to input the target patient's age, heart rate, and blood oxygen partial pressure into the assessment model to obtain the risk assessment index value of pulmonary embolism adverse events for the target patient. The higher the value, the higher the risk of pulmonary embolism adverse events.

[0013] As a third aspect of the present invention, an electronic device is provided, characterized in that it includes a storage module, the storage module including instructions loaded and executed by a processor, the instructions, when executed, causing the processor to perform the pulmonary embolism adverse event risk assessment method of the first aspect described above.

[0014] As a fourth aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing one or more programs, characterized in that, when the one or more programs are executed by a processor, they implement the pulmonary embolism adverse event risk assessment method of the first aspect described above.

[0015] This invention constructs a functionalized pulmonary embolism adverse event risk assessment model, which directly incorporates three objective parameters: age, heart rate, and partial pressure of oxygen. The weights of age, heart rate, and partial pressure of oxygen are derived objectively (not manually set), completely eliminating reliance on physician experience and avoiding assessment bias caused by subjective factors. Furthermore, the functionalized risk assessment model has continuity, avoiding the problem of large differences in assessment results for patients with similar parameters in existing technologies.

[0016] Furthermore, the prediction accuracy of this application is significantly better than that of existing models, reducing the risk of clinical misjudgment. It has excellent generalization stability, is suitable for long-term and diverse clinical scenarios, and is highly convenient to operate, adapting to actual clinical workflows. Attached Figure Description

[0017] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments: Figure 1 A flowchart of a method for assessing the risk of adverse events of pulmonary embolism provided in an embodiment of the present invention; Figure 2 A schematic diagram of a pulmonary embolism adverse event risk assessment device provided in an embodiment of the present invention; Figure 3 A schematic diagram of an electronic device provided in an embodiment of the present invention; Figure 4(a) and (b) show the ROC curves of the existing PERFORM model and the evaluation model (CON-PERFORM) of this embodiment relative to the training set, respectively; Figure 5 (a) and (b) show the ROC curves of the existing PERFORM model and the evaluation model (CON-PERFORM) of this embodiment relative to the test set, respectively. Detailed Implementation

[0018] The embodiments of the present invention will be described below with reference to the accompanying drawings. It should be noted that the embodiments described in this specification are not exhaustive and do not represent the only embodiments of the present invention. The corresponding embodiments below are only for clearly illustrating the inventive content of this patent and are not intended to limit its implementation. For those skilled in the art, different variations and modifications can be made based on the embodiments described. Any variations or modifications that fall within the technical concept and inventive content of this invention and are obvious are also within the protection scope of this invention.

[0019] like Figure 1 As shown in the figure, this application provides a method for assessing the risk of adverse events related to pulmonary embolism, the specific process of which is as follows: S101. Establish a risk assessment model for adverse events of pulmonary embolism in patients:

[0020] Where G represents the risk assessment index for adverse events of pulmonary embolism in patients, and k age k heartrate k PO2 The weights of the patient's age, heart rate, and partial pressure of oxygen (PO2) on the risk of adverse events.

[0021] For a patient, their age, heart rate, and blood oxygen partial pressure are known, while k age k heartrate k PO2 Since G is the variable to be determined, we have G = G(k) age , k heartrate , k PO2 In a specific set of k age k heartrate k PO2 The G value for each sample case can be calculated, and the performance of the above model can be evaluated by plotting an ROC curve. In this embodiment, AUC (Area Under Curve, which is the area under the ROC curve, and is often used as the evaluation criterion for the model) is defined as Q. Therefore: Q = Q(G) That is, Q=Q(G(k) age , k heartrate, k PO2 ))= Q'(k age , k heartrate , k PO2 ) This is a weighted k age , k heartrate , k PO2 Q is the independent variable, and it is a function of the dependent variable. Therefore, we need to find the weight k when Q is optimal. age , k heartrate , k PO2 .

[0022] analytic function It can be observed that when the independent variable is multiplied by the coefficient N (a positive number), the evaluation index is equivalent to being multiplied by N, which has no effect on AUC, that is, no effect on Q, i.e., Q'(N*k) age , N*k heartrate , N*k PO2 )= Q'(k age , k heartrate , k PO2 )=Q. Therefore, let k age k heartrate k PO2 The normalized weights, the values ​​of which are greater than 0 and less than 1, and satisfy the following conditions: Then there is .

[0023] S102. Model Simplification: Assume k heartrate and k PO2 The relative weighting coefficient k heartrate_PO2 The value of is greater than 0 and less than 1 and satisfies:

[0024] Then there is Based on this, the risk assessment model for adverse events of pulmonary embolism in patients is simplified as follows: .

[0025] That is, G=G'(k) age ,k heartrate_PO2 If ), then Q = Q(G) = Q ~ (k age ,k heartrate_PO2 ).

[0026] After the above transformation, the number of independent variables of the function is reduced by one, which simplifies the process and also facilitates three-dimensional display (intuitively showing how the evaluation effect of the model changes as the control parameters change).

[0027] S103. Generate k using a random number seed within a range of values ​​(greater than 0, less than 1). age and k heartrate_PO2 For each combination pair, the assessment index G corresponding to each case sample is calculated based on the simplified model, and the AUC is calculated by plotting the ROC curve, thus obtaining the AUC corresponding to each combination pair.

[0028] The case samples are historical cases. The number of combination pairs is greater than or equal to 1000; in this embodiment, the number of combination pairs is selected as 5000.

[0029] S104. Fit the function Q using the one-to-one correspondence between each combination and AUC. ~ (k age , k heartrate_PO2 ), in k age and k heartrate_PO2 Find the k corresponding to the maximum function value within the range of values. age and k heartrate_PO2 This allows us to obtain the optimal weight values ​​for each element of the evaluation model.

[0030] The one-to-one correspondence between each combination and AUC can form a polynomial, which can be used to fit the function Q. ~ (k age ,k heartrate_PO2 ), by calculating the function within the domain (k age ∈ (0, 1), k heartrate_PO2 Find the maximum point of ∈ (0, 1), compare the function values ​​at the maximum point with the boundary function values, and determine the largest function value in the domain and the corresponding value of the independent variable, i.e., the corresponding k. age With k heartrate_PO2 Substitute back into the formula and k can be obtained heartrate and k PO2 This means that the optimal weight values ​​for each element of the evaluation model were obtained.

[0031] The evaluation model that determines the optimal weight values ​​for each item can be deployed on a computer.

[0032] Table 1 shows the change in the maximum value of the fitted function as the number of combination pairs changes. It can be seen that as the number of combination pairs increases from 1000 to 6000, the fluctuation of the maximum value of the function is less than 0.02%, indicating that the evaluation model has high stability.

[0033] S105. Input the target patient's age, heart rate, and blood oxygen partial pressure into the assessment model to obtain the risk assessment index value of pulmonary embolism adverse events for the target patient. The larger the value, the higher the risk of pulmonary embolism adverse events.

[0034] Doctors can input the age, heart rate, and blood oxygen partial pressure of the target patient into the assessment model through human-computer interaction devices. Alternatively, the hospital information system can automatically input the age, heart rate, and blood oxygen partial pressure of the target patient into the assessment model after the doctor designates the patient. The operation is highly convenient and adaptable to the actual clinical workflow.

[0035] The target patient refers to a patient whose risk of pulmonary embolism adverse events needs to be assessed. The obtained risk assessment index values ​​for pulmonary embolism adverse events in the target patient can assist physicians in planning the patient's clinical treatment.

[0036] In this embodiment, an optimal threshold for the risk assessment index value of pulmonary embolism adverse events is set, with a value range between 54 and 56. Thus, for example, if the risk assessment index value of the target patient for pulmonary embolism adverse events is ≥55.6, the patient is determined to be a high-risk patient, and vice versa.

[0037] Figure 4 (a) and (b) show the ROC curves of the existing PERFORM model and the evaluation model (CON-PERFORM) of this embodiment relative to the training set, respectively. The performance comparison of the existing PERFORM model and the evaluation model (CON-PERFORM) of this embodiment relative to the training set is shown in Table 1.

[0038] Table 1 CON-PERFORM PERFORM AUC value 0.841 0.793 Optimal threshold 56.2 6 accuracy 0.783 0.716 Sensitivity 0.828 0.759 Specificity 0.779 0.712 As can be seen from Table 1: The AUC value of CON-PERFORM reached 0.841, which was significantly higher than that of the PERFORM model (0.793). This indicates that within the training set, CON-PERFORM has higher overall predictive accuracy for the 30-day mortality risk of pulmonary embolism patients and can more accurately distinguish between high-risk and low-risk patients.

[0039] CON-PERFORM achieved an accuracy of 0.783, a significant improvement over PERFORM's 0.716. This indicates that in the training samples, CON-PERFORM has a lower probability of misclassification and more reliable classification results when classifying risks based on objective parameters such as age, heart rate, and PO2.

[0040] CON-PERFORM has a sensitivity of 0.828, which is higher than PERFORM's 0.759, indicating a stronger ability to identify high-risk patients and effectively reducing the clinical risk of "missing high-risk patients". At the same time, its specificity is 0.779, which is higher than PERFORM's 0.712, meaning that it has a better ability to exclude low-risk patients and can reduce the waste of medical resources caused by "misjudging low-risk patients as high-risk".

[0041] The optimal threshold is more clinically relevant: the optimal threshold for CON-PERFORM is 56.2. This threshold is objectively derived from a large sample based on functional analysis, rather than relying on physician experience. This threshold can be used to conduct more standardized clinical risk stratification and provide a stable reference standard for treatment decisions.

[0042] Figure 5 (a) and (b) show the ROC curves of the existing PERFORM model and the evaluation model (CON-PERFORM) of this embodiment relative to the test set, respectively. The performance comparison of the existing PERFORM model and the evaluation model (CON-PERFORM) of this embodiment relative to the test set is shown in Table 2.

[0043] Table 2 CON-PERFORM PERFORM AUC value 0.769 0.74 Optimal threshold 54.1 5 accuracy 0.763 0.624 Sensitivity 0.786 0.786 Specificity 0.762 0.61 As shown in Table 2, the performance advantage of CON-PERFORM is further validated in the test samples independent of the training set, and it is more in line with real clinical application scenarios, as detailed below: CON-PERFORM's AUC value was 0.769, higher than PERFORM's 0.74, indicating that CON-PERFORM's risk prediction ability remained superior in new samples that were not involved in model training, without the problem of "training set overfitting", and its performance stability was more in line with actual clinical needs.

[0044] The accuracy of CON-PERFORM was 0.763, while that of PERFORM was only 0.624. The difference between the two (13.9 percentage points) further widened compared to the training set (6.7 percentage points), indicating that when faced with diverse clinical samples, CON-PERFORM, based on the objective evaluation logic of the functional model, has stronger anti-interference ability and more stable classification results.

[0045] Both had a sensitivity of 0.786 (both could effectively capture high-risk patients), but CON-PERFORM's specificity (0.762) was much higher than PERFORM's (0.61). This means that without reducing the identification rate of high-risk patients, CON-PERFORM can more accurately screen low-risk patients, which can directly support the clinical strategy of "discharging low-risk patients as early as possible and strengthening the monitoring of high-risk patients", reducing unnecessary hospitalization and improving the efficiency of medical resource utilization.

[0046] This application also uses two batches of test sets (batch 1: 2010-2017, batch 2: 2024-2025) to compare CON-PERFORM, as shown in Table 3.

[0047] Table 3 The first batch Second batch relative error AUC value 0.769 0.751 0.024 Optimal threshold 54.1 54.637 0.01 accuracy 0.763 0.773 0.014 Sensitivity 0.786 0.778 0.01 Specificity 0.762 0.773 0.015 As can be seen from Table 3: The relative errors of CON-PERFORM's AUC (0.769 vs 0.751), accuracy (0.763 vs 0.773), sensitivity (0.786 vs 0.778), and specificity (0.762 vs 0.773) were all less than 3% (the maximum relative error was 1.5% for specificity). This indicates that the model's risk prediction ability remains stable regardless of changes in sample collection time and sample size, and its performance will not degrade due to time differences in patient population characteristics (such as advancements in medical technology and changes in the composition of patients' underlying diseases).

[0048] The optimal thresholds for the two batches of samples were 54.1 and 54.637, respectively, with a relative error of only 1%. This indicates that there is no need to frequently adjust the risk stratification threshold in clinical applications. Risk assessment can be carried out quickly based on stable thresholds, reducing operational complexity and facilitating promotion and use in different medical institutions and at different time periods.

[0049] The second batch of samples (2024-2025) is relatively new clinical data. Its performance is consistent with that of the earlier samples (2010-2017), proving that the CON-PERFORM model has long-term effectiveness and can meet the clinical needs for pulmonary embolism risk assessment in the future, providing key data support for the large-scale application of the model.

[0050] As can be seen from the above, the pulmonary embolism adverse event risk assessment method provided in this application constructs a functionalized pulmonary embolism adverse event risk assessment model, which directly incorporates three objective parameters: age, heart rate, and partial pressure of oxygen. The weights of age, heart rate, and partial pressure of oxygen are obtained through objective derivation (not manually set), completely eliminating the reliance on doctors' experience and avoiding assessment bias caused by subjective factors. Furthermore, the functionalized risk assessment model has continuity, avoiding the problem of large differences in assessment results between patients with similar parameters in the prior art.

[0051] Furthermore, the prediction accuracy of this application is significantly better than that of existing models, reducing the risk of clinical misjudgment. It has excellent generalization stability, is suitable for long-term and diverse clinical scenarios, and is highly convenient to operate, adapting to actual clinical workflows.

[0052] The following describes in detail one or more embodiments of the pulmonary embolism adverse event risk assessment device of the present invention. Those skilled in the art will understand that these devices can be configured using commercially available hardware components through the steps taught in this solution. Figure 2 An embodiment of the present invention illustrates a pulmonary embolism adverse event risk assessment device, which includes a first module 11, a second module 12, a third module 13, a fourth module 14, and a fifth module 15.

[0053] Module 11, used for S101, establishes a risk assessment model for adverse events of pulmonary embolism in patients:

[0054] Where G represents the risk assessment index for adverse events of pulmonary embolism in patients, and k age k heartrate k PO2 The weights of the patient's age, heart rate, and partial pressure of oxygen (PO2) on the risk of adverse events.

[0055] For a patient, their age, heart rate, and blood oxygen partial pressure are known, while k age k heartrate k PO2 Since G is the variable to be determined, we have G = G(k) age , k heartrate , k PO2 In a specific set of k age k heartrate k PO2 The G value for each sample case can be calculated, and the performance of the above model can be evaluated by plotting an ROC curve. In this embodiment, AUC (Area Under Curve, which is the area under the ROC curve, and is often used as the evaluation criterion for the model) is defined as Q. Therefore: Q = Q(G) That is, Q=Q(G(k) age , k heartrate , k PO2 ))= Q'(k age , k heartrate , k PO2 ) This is a weighted k age , k heartrate , k PO2 Q is the independent variable, and it is a function of the dependent variable. Therefore, we need to find the weight k when Q is optimal. age , k heartrate , k PO2 .

[0056] analytic function It can be observed that when the independent variable is multiplied by the coefficient N (a positive number), the evaluation index is equivalent to being multiplied by N, which has no effect on AUC, that is, no effect on Q, i.e., Q'(N*k) age , N*k heartrate , N*k PO2 )= Q'(k age , k heartrate , k PO2 )=Q. Therefore, let k age kheartrate k PO2 The normalized weights, the values ​​of which are greater than 0 and less than 1, and satisfy the following conditions: Then there is .

[0057] The second module 12 is used for S102, assuming k heartrate and k PO2 The relative weighting coefficient k heartrate_PO2 The value of is greater than 0 and less than 1 and satisfies:

[0058] Then there is Based on this, the risk assessment model for adverse events of pulmonary embolism in patients is simplified as follows: .

[0059] That is, G=G'(k) age ,k heartrate_PO2 If ), then Q = Q(G) = Q ~ (k age ,k heartrate_PO2 ).

[0060] After the above transformation, the function's independent variable is reduced by one, achieving simplification and also facilitating 3D display.

[0061] The third module 13 is used by S103 to generate k using a random number seed within a range of values ​​(greater than 0, less than 1). age and k heartrate_PO2 For each combination pair, the assessment index G corresponding to each case sample is calculated based on the simplified model, and the AUC is calculated by plotting the ROC curve, thus obtaining the AUC corresponding to each combination pair.

[0062] The case samples are historical cases. The number of combination pairs is greater than or equal to 1000; in this embodiment, the number of combination pairs is selected as 5000.

[0063] Module 4, 14, is used in S104 to fit the function Q through the one-to-one correspondence between each combination and AUC. ~ (k age ,k heartrate_PO2 ), in k age and k heartrate_PO2 Find the k corresponding to the maximum function value within the range of values. age and k heartrate_PO2 This allows us to obtain the optimal weight values ​​for each element of the evaluation model.

[0064] The one-to-one correspondence between each combination and AUC can form a polynomial, which can be used to fit the function Q. ~ (kage ,k heartrate_PO2 ), by calculating the function within the domain (k age ∈ (0, 1), k heartrate_PO2 Find the maximum point of ∈ (0, 1), compare the function values ​​at the maximum point with the boundary function values, and determine the largest function value in the domain and the corresponding value of the independent variable, i.e., the corresponding k. age With k heartrate_PO2 Substitute back into the formula and k can be obtained heartrate and k PO2 This means that the optimal weight values ​​for each element of the evaluation model were obtained.

[0065] The evaluation model that determines the optimal weight values ​​for each item can be deployed on a computer.

[0066] Table 1 shows the change in the maximum value of the fitted function as the number of combination pairs changes. It can be seen that as the number of combination pairs increases from 1000 to 6000, the fluctuation of the maximum value of the function is less than 0.02%, indicating that the evaluation model has high stability.

[0067] The fifth module 15 is used in S105 to input the target patient's age, heart rate, and blood oxygen partial pressure into the assessment model to obtain the risk assessment index value of pulmonary embolism adverse events for the target patient. The larger the value, the higher the risk of pulmonary embolism adverse events.

[0068] Doctors can input the age, heart rate, and blood oxygen partial pressure of the target patient into the assessment model through human-computer interaction devices. Alternatively, the hospital information system can automatically input the age, heart rate, and blood oxygen partial pressure of the target patient into the assessment model after the doctor designates the patient. The operation is highly convenient and adaptable to the actual clinical workflow.

[0069] The target patient refers to a patient whose risk of pulmonary embolism adverse events needs to be assessed. The obtained risk assessment index values ​​for pulmonary embolism adverse events in the target patient can assist physicians in planning the patient's clinical treatment.

[0070] In this embodiment, an optimal threshold for the risk assessment index value of pulmonary embolism adverse events is set, with a value range between 54 and 56. Thus, for example, if the risk assessment index value of the target patient for pulmonary embolism adverse events is ≥55.6, the patient is determined to be a high-risk patient, and vice versa.

[0071] In summary, the pulmonary embolism adverse event risk assessment device provided in the above embodiments can execute the pulmonary embolism adverse event risk assessment methods provided in the foregoing embodiments.

[0072] Similar to the above concept, the above Figure 2The structure of the pulmonary embolism adverse event risk assessment device shown can be implemented as an electronic device. Figure 3 A schematic block diagram of the structure of an electronic device provided by an embodiment of the present invention is shown.

[0073] For example, the electronic device includes a storage module 21 and a processor 22. The storage module 21 includes instructions loaded and executed by the processor 22, which, when executed, cause the processor 22 to perform the steps described in the above-described section of this specification, "A Method for Assessing the Risk of Pulmonary Embolism Adverse Events," according to various exemplary embodiments of the present invention.

[0074] It should be understood that processor 22 can be a Central Processing Unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among these, the general-purpose processor can be a microprocessor or any conventional processor.

[0075] This invention also provides a computer-readable storage medium that stores one or more programs, which, when executed by a processor, implement the steps described in the above-described method for assessing adverse events of pulmonary embolism according to various exemplary embodiments of the invention.

[0076] Those skilled in the art will understand that all or some of the steps, systems, and apparatuses disclosed above, and their functional modules / units, can be implemented as software, firmware, hardware, or suitable combinations thereof. In hardware implementations, the division between functional modules / units mentioned above does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed collaboratively by several physical components. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit (ASIC). Such software can be distributed on a computer-readable storage medium, which may include computer-readable storage media (or non-transitory media) and communication media (or transient media).

[0077] As is known to those skilled in the art, the term computer-readable storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer-readable storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. Furthermore, it is known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0078] For example, the computer-readable storage medium may be an internal storage unit of the electronic device described in the foregoing embodiments, such as a hard disk or memory of the electronic device. The computer-readable storage medium may also be an external storage device of the electronic device, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, Flash Card, etc., provided on the electronic device.

[0079] The electronic devices and computer-readable storage media provided in the foregoing embodiments construct a functionalized pulmonary embolism adverse event risk assessment model, which directly incorporates three objective parameters: age, heart rate, and partial pressure of oxygen. The weights of age, heart rate, and partial pressure of oxygen are derived objectively (not manually set), completely eliminating reliance on doctors' experience and avoiding assessment bias caused by subjective factors. Furthermore, the functionalized risk assessment model has continuity, avoiding the problem of large differences in assessment results for patients with similar parameters in the prior art.

[0080] Furthermore, the prediction accuracy of this application is significantly better than that of existing models, reducing the risk of clinical misjudgment. It has excellent generalization stability, is suitable for long-term and diverse clinical scenarios, and is highly convenient to operate, adapting to actual clinical workflows.

[0081] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for assessing the risk of adverse events related to pulmonary embolism, characterized in that, include: Establish a risk assessment model for adverse events of pulmonary embolism in patients: Where G represents the risk assessment index for adverse events of pulmonary embolism in patients, and k age k heartrate k PO2 The weights of patient age, heart rate, and blood oxygen partial pressure represent the impact of adverse event risk. The values ​​of these three factors are greater than 0 and less than 1, and satisfy the following conditions: ; Assume k heartrate and k PO2 The relative weighting coefficient k heartrate_PO2 The value of is greater than 0 and less than 1 and satisfies: Then there is Therefore, the risk assessment model for adverse events of pulmonary embolism in patients is simplified as follows: ； Generate k within the range of values. age and k heartrate_PO2 For each combination pair, the evaluation index G corresponding to each case sample is calculated according to the simplified model, and the AUC is calculated by plotting the ROC curve to obtain the AUC corresponding to each combination pair. The function Q is fitted by the one-to-one correspondence between each combination and AUC. ~ (k age , k heartrate_PO2 ), in k age and k heartrate_PO2 Find the k corresponding to the maximum function value within the range of values. age and k heartrate_PO2 The weight values ​​of each item in the evaluation model are obtained. The target patient's age, heart rate, and blood oxygen partial pressure are input into the assessment model to obtain the risk assessment index value of pulmonary embolism adverse events for the target patient. The higher the value, the higher the risk of pulmonary embolism adverse events.

2. The method for assessing the risk of adverse events of pulmonary embolism according to claim 1, characterized in that, The number of the combination pairs is greater than or equal to 1000.

3. The method for assessing the risk of adverse events of pulmonary embolism according to claim 1, characterized in that, Also includes: If the target patient's pulmonary embolism adverse event risk assessment index value is greater than or equal to the optimal segmentation threshold, the patient is determined to be a high-risk patient; otherwise, the patient is a low-risk patient.

4. A device for assessing the risk of adverse events of pulmonary embolism, characterized in that, include: The first module is used to establish a risk assessment model for adverse events of pulmonary embolism in patients: Where G represents the risk assessment index for adverse events of pulmonary embolism in patients, and k age k heartrate k PO2 The weights of patient age, heart rate, and blood oxygen partial pressure represent the impact of adverse event risk. The values ​​of these three factors are greater than 0 and less than 1, and satisfy the following conditions: ; The second module is used to assume k heartrate and k PO2 The relative weighting coefficient k heartrate_PO2 The value of is greater than 0 and less than 1 and satisfies: Then there is Therefore, the risk assessment model for adverse events of pulmonary embolism in patients is simplified as follows: ； The third module is used to generate k within the range of values. age and k heartrate_PO2 For each combination pair, the evaluation index G corresponding to each case sample is calculated according to the simplified model, and the AUC is calculated by plotting the ROC curve to obtain the AUC corresponding to each combination pair. The fourth module is used to fit the function Q through the one-to-one correspondence between each combination and AUC. ~ (k age , k heartrate_PO2 ), in k age and k heartrate_PO2 Find the k corresponding to the maximum function value within the range of values. age and k heartrate_PO2 The weight values ​​of each item in the evaluation model are obtained. The fifth module is used to input the target patient's age, heart rate, and blood oxygen partial pressure into the assessment model to obtain the risk assessment index value of pulmonary embolism adverse events for the target patient. The higher the value, the higher the risk of pulmonary embolism adverse events.

5. An electronic device, characterized in that, The system includes a storage module comprising instructions loaded and executed by a processor, which, when executed, cause the processor to perform a method for assessing the risk of adverse events of pulmonary embolism according to any one of claims 1-3.

6. A computer-readable storage medium storing one or more programs, characterized in that, When the one or more programs are executed by the processor, they implement the method for assessing the risk of adverse events of pulmonary embolism as described in any one of claims 1-3.

Images