Alarm precision improving method and device, equipment, storage medium and product

Abstract

The invention provides an alarm precision improvement method and device, equipment, a storage medium and a product, and belongs to the technical field of medical care information processing, and the method comprises the steps: obtaining a historical sample set, a real-time physical sign data set and a real-time prediction result of an identification early warning model; calculating the credibility of the real-time prediction result; when the credibility of the real-time prediction result is smaller than or equal to a preset credibility threshold value, generating a sign weight; correcting the physical sign data in the real-time physical sign data set according to the physical sign weight to generate a corrected physical sign data set and a corrected risk probability; and when the corrected risk probability is greater than a preset probability threshold, outputting an alarm signal, and when the corrected risk probability is less than or equal to the preset probability threshold, outputting a silence signal. According to the alarm precision improving method and device, the equipment, the storage medium and the product, the alarm precision of an EHR system can be improved, and the occurrence probability of false alarm and leakage alarm phenomena is reduced.

Description

Technical Field

[0001] The present invention belongs to the field of medical care information processing technology, and in particular relates to a method, device, equipment, storage medium and product for improving alarm accuracy. Background Art

[0002] An EHR (Electronic Health Record) system is a system specifically designed to record, store, and manage patient health data. Compared to traditional paper medical records, EHR systems can automatically collect and analyze patient vital signs through various devices, providing medical decision support and enabling more accurate and timely treatment plans.

[0003] In existing technologies, EHR systems typically incorporate an identification and early warning model. After collecting a patient's vital signs, the model predicts the patient's health risk probability based on that data and issues an alarm if the health risk probability exceeds a preset threshold, prompting medical staff to provide timely attention and make clinical decisions.

[0004] However, due to the limitations of the prediction accuracy of the identification and early warning model, the EHR system will inevitably have false alarms and missed alarms. In actual use, the false alarm phenomenon refers to the EHR system issuing an alarm even though the patient is not in critical condition. The frequent occurrence of this phenomenon will increase the workload of medical staff and reduce the medical staff's trust in the EHR system. The missed alarm phenomenon refers to the patient being in critical condition but the EHR system does not issue an alarm. This phenomenon will result in the patient being unable to receive timely treatment, causing serious adverse effects. To solve this problem, those skilled in the art will usually choose to establish a more complex deep learning model or increase the number of parameters to improve the prediction accuracy of the identification and early warning model. However, such a solution will greatly increase the computing cost and data acquisition cost, and therefore cannot meet the actual use needs. Summary of the Invention

[0005] In view of this, the present invention aims to propose a method, device, equipment, storage medium and product for improving alarm accuracy to solve the above technical problems.

[0006] To achieve the above object, the technical solution created by the present invention is implemented as follows: In a first aspect, an embodiment of the present invention provides a method for improving alarm accuracy, comprising: Obtaining a historical sample set, a real-time vital sign data set, and a real-time prediction result of an identification and warning model, wherein the historical sample set includes a plurality of historical prediction results and a historical vital sign data set corresponding to each historical prediction result; Calculate the credibility of real-time prediction results based on real-time vital sign data sets and historical sample sets; When the credibility of the real-time prediction result is less than or equal to the preset credibility threshold, determining the importance of each physical sign data in the historical physical sign data set according to the historical sample set, and generating a physical sign weight of the physical sign data according to the importance; Correcting the physical sign data in the real-time physical sign data set according to the physical sign weight to generate a corrected physical sign data set, and inputting the corrected physical sign data set into the identification and early warning model to generate a corrected risk probability; When the corrected risk probability is greater than the preset probability threshold, a high-risk corrected prediction result is generated and an alarm signal is output; when the corrected risk probability is less than or equal to the preset probability threshold, a low-risk corrected prediction result is generated and a silent signal is output.

[0007] Furthermore, the real-time prediction results include high-risk real-time prediction results and low-risk real-time prediction results; The historical sample set includes a high-risk historical sample subset and a low-risk historical sample subset; The high-risk historical sample subset includes high-risk historical prediction results and historical physical sign data sets corresponding to the high-risk historical prediction results; The low-risk historical sample subset includes low-risk historical prediction results and historical physical sign data sets corresponding to the low-risk historical prediction results; The calculation of the credibility of the real-time prediction result based on the real-time vital sign data set and the historical sample set includes: Calculate the consistency metric value of each historical vital sign data set in the high-risk historical sample subset, and sort the consistency metric values of each historical vital sign data set in ascending order to generate a high-risk historical consistency list; Calculate the consistency metric value of each historical vital sign data set in the low-risk historical sample subset, and sort the consistency metric values of each historical vital sign data set in ascending order to generate a low-risk historical consistency list; When the real-time prediction result is a high-risk real-time prediction result, the real-time vital sign dataset is used as a newly added vital sign dataset in the high-risk historical sample subset, the consistency metric value of the newly added vital sign dataset is calculated, and the credibility of the real-time prediction result is determined according to the sorting position of the consistency metric value of the newly added vital sign dataset in the high-risk historical consistency list; When the real-time prediction result is a low-risk real-time prediction result, the real-time vital sign data set is used as a new vital sign data set in the low-risk historical sample subset, the consistency measurement value of the new vital sign data set is calculated, and the credibility of the real-time prediction result is determined based on the sorting position of the consistency measurement value of the new vital sign data set in the low-risk historical consistency list.

[0008] Furthermore, before calculating the credibility of the real-time prediction result based on the real-time vital sign data set and the historical sample set, the alarm accuracy improvement method further includes: Determine the normal interval of the historical physical sign data set based on the historical sample set; The change direction and change rate of the real-time vital sign data set within a preset time period are obtained. If the real-time vital sign data set changes in a direction approaching the normal interval, the preset credibility threshold is adjusted downward, and the adjustment step size of the preset credibility threshold is inversely proportional to the change rate. If the real-time vital sign data set changes in a direction away from the normal interval, the preset credibility threshold is adjusted upward, and the adjustment step size of the preset credibility threshold is directly proportional to the change rate.

[0009] Furthermore, before calculating the credibility of the real-time prediction result based on the real-time vital sign data set and the historical sample set, the alarm accuracy improvement method further includes: The false alarm rate and missed alarm rate of the EHR system are obtained, and when the false alarm rate increases, the preset credibility threshold is increased, and when the missed alarm rate increases, the preset credibility threshold is decreased.

[0010] Furthermore, after calculating the credibility of the real-time prediction result based on the real-time vital sign data set and the historical sample set, the alarm accuracy improvement method further includes: When the credibility of the real-time prediction result is greater than the preset credibility threshold, the type of the real-time prediction result is judged. If the real-time prediction result is a high-risk real-time prediction result, an alarm signal is output; if the real-time prediction result is a low-risk real-time prediction result, a silent signal is output.

[0011] Furthermore, before correcting the vital sign data in the real-time vital sign data set according to the vital sign weight, the alarm accuracy improvement method further includes: Determine the normal interval of the historical physical sign data set based on the historical sample set; Obtain the object attribution of the real-time vital sign data set and the change direction of the real-time vital sign data set within a preset time period; if the real-time vital sign data set changes in a direction approaching the normal interval, the vital sign weight is adjusted downward, and the reduction adjustment step length when the real-time vital sign data set belongs to a high-risk object is smaller than the reduction adjustment step length when the real-time vital sign data set belongs to an ordinary object; if the real-time vital sign data set changes in a direction away from the normal interval, the vital sign weight is adjusted upward, and the increase adjustment step length when the real-time vital sign data set belongs to a high-risk object is larger than the increase adjustment step length when the real-time vital sign data set belongs to an ordinary object.

[0012] In a second aspect, an embodiment of the present invention further provides a device for improving alarm accuracy, comprising: An acquisition module is used to acquire a historical sample set, a real-time vital sign data set, and a real-time prediction result of an identification and warning model, wherein the historical sample set includes a plurality of historical prediction results and a historical vital sign data set corresponding to each historical prediction result; A calculation module, used to calculate the credibility of real-time prediction results based on the real-time vital sign data set and the historical sample set; a generating module, configured to determine the importance of each type of vital sign data in the historical vital sign data set based on the historical sample set when the credibility of the real-time prediction result is less than or equal to a preset credibility threshold, and generate a vital sign weight for the vital sign data based on the importance; A correction module is used to correct the physical sign data in the real-time physical sign data set according to the physical sign weight to generate a corrected physical sign data set, and input the corrected physical sign data set into the identification and early warning model to generate a corrected risk probability; The output module is used to generate a high-risk correction prediction result and output an alarm signal when the corrected risk probability is greater than the preset probability threshold; and to generate a low-risk correction prediction result and output a silent signal when the corrected risk probability is less than or equal to the preset probability threshold.

[0013] In a third aspect, an embodiment of the present invention further provides a device, including: one or more processors; a storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the alarm accuracy improvement method provided in any of the above embodiments.

[0014] In a fourth aspect, an embodiment of the present invention further provides a storage medium comprising computer-executable instructions, which, when executed by a computer processor, are used to execute the alarm accuracy improvement method provided in any of the above embodiments.

[0015] In a fifth aspect, an embodiment of the present invention further provides a computer program product, including a computer program, which, when executed by a processor, implements the alarm accuracy improvement method provided in any of the above embodiments.

[0016] Compared with the prior art, the alarm accuracy improvement method, device, equipment, storage medium and product created by the present invention have the following advantages: The present invention creates a method, device, equipment, storage medium and product for improving alarm accuracy, which can calculate the credibility of the real-time prediction result based on the real-time vital sign data set and the historical sample set. When the credibility of the real-time prediction result is less than or equal to the preset credibility threshold, it can generate a vital sign weight according to the importance of each vital sign data, and correct the real-time vital sign data set according to the vital sign weight, thereby generating a corrected vital sign data set. Subsequently, a corrected risk probability is generated by inputting the corrected vital sign data set into the identification warning model, and a corrected prediction result is generated by comparing the corrected risk probability with the preset probability threshold, and an alarm signal or a silent signal is output according to the corrected prediction result. Compared with the prior art, the present invention can avoid a substantial increase in computing cost and data acquisition cost, and can make the corrected risk probability have better sensitivity through the vital sign weight, so that the prediction result more accurately reflects the actual situation of the patient, thereby improving the alarm accuracy of the EHR system and reducing the probability of false alarms and missed alarms. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings, which constitute part of the present invention, are intended to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are intended to explain the present invention and do not constitute an undue limitation of the present invention. In the accompanying drawings: Figure 1 A flowchart of the method for improving alarm accuracy according to Example 1 of the present invention is created; Figure 2 A flowchart of the method for improving alarm accuracy according to the second embodiment of the present invention is created; Figure 3 A flowchart of the method for improving alarm accuracy as described in Example 3 of the present invention is created; Figure 4 This is a schematic diagram of the structure of the alarm accuracy improvement device according to the fourth embodiment of the present invention; Figure 5 This is a structural diagram of the device described in Example 5 of the present invention. DETAILED DESCRIPTION

[0018] The present invention will be further described in detail below with reference to the accompanying drawings and examples. It will be understood that the specific embodiments described herein are intended only to illustrate the present invention and are not intended to limit the present invention. It should also be noted that, for ease of description, the accompanying drawings only illustrate portions relevant to the present invention, not all structures.

[0019] Example 1 Figure 1 This is a flowchart of a method for improving alarm accuracy provided in Example 1 of the present invention, which specifically includes the following steps: Step 110: Acquire a historical sample set, a real-time vital sign data set, and a real-time prediction result of the identification and warning model, wherein the historical sample set includes a plurality of historical prediction results and a historical vital sign data set corresponding to each historical prediction result.

[0020] When the EHR system is operating, data collection equipment continuously collects patients' vital signs (such as heart rate, blood pressure, blood oxygen saturation, and blood sugar) and inputs this data into the identification and early warning model. The identification and early warning model then predicts the patient's health risks based on this data, providing timely alarms when health risks exist and reminding medical staff to make clinical decisions. Accordingly, the EHR system also stores the historical input data and output results of the identification and early warning model for recordkeeping and subsequent management.

[0021] To improve the alarm accuracy of the EHR system, this embodiment acquires a historical sample set, a real-time vital sign dataset, and real-time prediction results from the identification and early warning model. The historical sample set should include multiple historical prediction results (i.e., historical output results) and a historical vital sign dataset (i.e., historical input data) corresponding to each historical prediction result. The real-time vital sign dataset is the real-time input data for the identification and early warning model, while the real-time prediction results are the real-time output results generated by the identification and early warning model based on the real-time input data.

[0022] In the subsequent work process of the EHR system, the historical sample set will be used to calculate the credibility of subsequent prediction results, thereby avoiding false alarms and missed alarms caused by prediction results with low credibility.

[0023] It should be noted that because the identification and early warning model typically requires input of multiple types of vital sign data when predicting a patient's health risks, and because the acquisition devices used to collect different types of vital sign data have different sampling frequencies, those skilled in the art typically set the sampling period based on actual conditions so that multiple types of vital sign data can form a vital sign dataset within the same sampling period and be input into the identification and early warning model, thereby facilitating the identification and early warning model's output of prediction results. Accordingly, the historical vital sign dataset in this embodiment should include multiple types of vital sign data within the same historical sampling period, and the real-time vital sign dataset should include multiple types of vital sign data within the same real-time sampling period.

[0024] Step 120: Calculate the credibility of the real-time prediction result based on the real-time vital sign data set and the historical sample set.

[0025] After obtaining the historical sample set, the real-time vital sign data set, and the real-time prediction results, this embodiment can calculate the credibility of the real-time prediction results and compare the calculated results with the preset credibility threshold to determine the subsequent processing method. Specifically, when the credibility of the real-time prediction results is greater than the preset credibility threshold, it proves that the real-time prediction results can accurately reflect the actual situation of the patient. At this time, it can be determined whether to trigger an alarm based on the real-time prediction results. When the credibility of the real-time prediction results is less than or equal to the preset credibility threshold, it proves that there may be errors in the real-time prediction results. At this time, the real-time prediction results should be subsequently processed to improve the alarm accuracy of the EHR system.

[0026] Optionally, the real-time prediction results in this embodiment may include two types: high-risk real-time prediction results and low-risk real-time prediction results. The high-risk real-time prediction results indicate that the patient currently faces a greater health risk and an alarm needs to be triggered to remind medical staff to perform clinical treatment in a timely manner. The low-risk real-time prediction results indicate that the patient's current health risk is low and no alarm needs to be triggered.

[0027] The historical sample set can also be divided into a high-risk historical sample subset and a low-risk historical sample subset based on the type of historical prediction results. The high-risk historical sample subset includes high-risk historical prediction results (i.e., historical prediction results that triggered alarms in historical records) and the historical vital sign datasets corresponding to high-risk historical prediction results. The low-risk historical sample subset includes low-risk historical prediction results (i.e., historical prediction results that did not trigger alarms in historical records) and the historical vital sign datasets corresponding to low-risk historical prediction results.

[0028] Accordingly, the credibility of the real-time prediction results is calculated based on the real-time vital sign dataset and the historical sample set, which can be optimized as follows: Calculate the consistency metric value of each historical vital sign data set in the high-risk historical sample subset, and sort the consistency metric values of each historical vital sign data set in ascending order to generate a high-risk historical consistency list; Calculate the consistency metric value of each historical vital sign data set in the low-risk historical sample subset, and sort the consistency metric values of each historical vital sign data set in ascending order to generate a low-risk historical consistency list; When the real-time prediction result is a high-risk real-time prediction result, the real-time vital sign dataset is used as a newly added vital sign dataset in the high-risk historical sample subset, the consistency metric value of the newly added vital sign dataset is calculated, and the credibility of the real-time prediction result is determined according to the sorting position of the consistency metric value of the newly added vital sign dataset in the high-risk historical consistency list; When the real-time prediction result is a low-risk real-time prediction result, the real-time vital sign data set is used as a new vital sign data set in the low-risk historical sample subset, the consistency measurement value of the new vital sign data set is calculated, and the credibility of the real-time prediction result is determined based on the sorting position of the consistency measurement value of the new vital sign data set in the low-risk historical consistency list.

[0029] In existing technologies, the credibility of prediction results is typically determined using clustering algorithms. This involves determining the credibility of real-time data based on the similarity between the real-time data and historical data. The higher the similarity, the higher the credibility of the real-time data. However, due to the limitations of the prediction accuracy of the identification and early warning model, historical prediction results within the historical sample set may contain errors. Using traditional clustering algorithms to calculate the credibility of real-time prediction results can also result in errors in the credibility of the real-time prediction results, making it impossible to guarantee the accuracy of subsequent processing results.

[0030] Therefore, this embodiment will first calculate the consistency metric for each historical vital sign dataset in the high-risk sample subset and the low-risk sample subset, thereby using the consistency metric to determine the credibility of the historical sample dataset. It should be noted that the consistency metric is used to characterize the degree of similarity between different vital sign datasets when the identification and early warning model outputs the same type of prediction results. The higher the similarity, the more credible the historical prediction results generated based on the historical vital sign dataset.

[0031] The following describes the specific calculation method of the consistency metric value in this embodiment by taking the calculation of the consistency metric value of a historical vital sign dataset in the high-risk historical sample subset as an example: The first step is to calculate the minimum Euclidean distance A between the historical physical sign dataset and all historical physical sign datasets in the low-risk historical sample subset; The second step is to calculate the minimum Euclidean distance B between the historical vital sign dataset and other historical vital sign datasets in the high-risk historical sample subset; The third step is to calculate the ratio C of A to B and use C as the consistency measure of the historical vital sign dataset.

[0032] Since the Euclidean distance can represent the true distance between two points in space, the larger the ratio C, the closer the historical vital sign dataset is to the high-risk historical sample subset, that is, the higher the similarity between the historical vital sign dataset and other historical vital sign datasets in the high-risk historical sample subset. If the ratio C is smaller, the closer the historical vital sign dataset is to the low-risk historical sample subset, that is, the higher the similarity between the historical vital sign dataset and the historical vital sign datasets in the low-risk historical sample subset.

[0033] After completing the calculation of the consistency metric values of the historical vital sign data set, this embodiment will sort the consistency metric values in ascending order to generate a high-risk historical consistency list and a low-risk historical consistency list. Since the consistency metric value is larger the further back it is sorted in the consistency list, the later a certain consistency metric value is sorted in the consistency list, the higher the credibility of the prediction result generated by the vital sign data set corresponding to the consistency metric value. Accordingly, after the real-time vital sign data set is used as a new vital sign data set according to the type of real-time prediction result and the consistency metric value calculation of the new vital sign data set is completed, the credibility of the real-time prediction result can be determined by the sorting position of the consistency metric value of the new vital sign data set in the corresponding consistency list.

[0034] To facilitate the conversion of the ranking position of the consistency metric value of the newly added vital sign dataset in the corresponding consistency list into a specific credibility value, this embodiment will take the real-time prediction result of a high-risk real-time prediction result as an example to specifically illustrate the conversion method of the credibility value: Step 1: After calculating the consistency metric value of the newly added vital sign dataset, substitute the consistency metric value of the newly added vital sign dataset into the high-risk historical consistency list, and obtain the position number X1 of the consistency metric value of the newly added vital sign dataset; Step 2: Obtain the position number X2 of the consistency measurement value of the last historical vital sign dataset in the high-risk historical consistency list; Step 3: First calculate the ratio Y of X1 and X2, and use the ratio Y as the credibility of the prediction result (i.e., real-time prediction result) corresponding to the newly added vital sign dataset (i.e., real-time vital sign dataset).

[0035] For example, assume that the high-risk historical consistency list contains the consistency metric values of 100 historical vital sign datasets. In this case, the position number X2 of the consistency metric value of the last historical vital sign dataset is 100. If the consistency metric value of the newly added vital sign dataset is ranked 70th in the high-risk historical consistency list, its position number X1 is 70. In this case, the ratio Y is 70 / 100, so the reliability of the prediction result is 0.7.

[0036] As an optional implementation of this embodiment, when the credibility of the real-time prediction result is greater than a preset credibility threshold, this embodiment may add the following steps after calculating the credibility of the real-time prediction result based on the real-time vital sign dataset and the historical sample set: When the credibility of the real-time prediction result is greater than the preset credibility threshold, the type of the real-time prediction result is judged. If the real-time prediction result is a high-risk real-time prediction result, an alarm signal is output; if the real-time prediction result is a low-risk real-time prediction result, a silent signal is output.

[0037] When the credibility of the real-time prediction result exceeds the preset credibility threshold, it proves that the real-time prediction result can accurately reflect the patient's actual situation. If the real-time prediction result is a high-risk real-time prediction result, it proves that the patient needs timely treatment. At this time, an alarm signal should be output to enable the EHR system to promptly alarm. If the real-time prediction result is a low-risk real-time prediction result, it proves that the patient's current health risk is low. At this time, a silent signal can be output to reduce the alarm frequency of the EHR system.

[0038] Step 130: When the credibility of the real-time prediction result is less than or equal to the preset credibility threshold, the importance of each physical sign data in the historical physical sign data set is determined according to the historical sample set, and the physical sign weight of the physical sign data is generated according to the importance.

[0039] When the credibility of a real-time prediction result is less than or equal to a preset credibility threshold, it indicates that the real-time prediction result may contain errors. If the type of real-time prediction result is directly used to determine whether to trigger an alarm, the EHR system will frequently generate false alarms or missed alarms. To address this issue, this embodiment determines the importance of each type of vital sign data in the historical vital sign data set based on the historical sample set, and generates a sign weight for the vital sign data based on the importance. The sign weight is then used to improve the accuracy and sensitivity of the prediction results of the identification and early warning model, thereby more accurately reflecting the actual situation of the patient and improving the alarm accuracy of the EHR system.

[0040] It should be noted that the importance of physical sign data refers to the importance of the physical sign indicators corresponding to the physical sign data in predicting whether a patient has a health risk of a certain type of disease. For example, when predicting whether a patient has a health risk of cardiovascular disease, blood pressure and heart rate are two physical sign indicators commonly used in this field, and the importance of blood pressure is higher than that of heart rate. Therefore, when generating the physical sign weights of physical sign data based on the importance, the physical sign weight of the physical sign data reflecting blood pressure will be greater than the physical sign weight of the physical sign data reflecting heart rate, so that the subsequent prediction process has higher sensitivity and accuracy for the health risk of cardiovascular disease.

[0041] Step 140: Correct the physical sign data in the real-time physical sign data set according to the physical sign weight to generate a corrected physical sign data set, and input the corrected physical sign data set into the identification and early warning model to generate a corrected risk probability.

[0042] After generating the vital sign weights, this embodiment modifies the vital sign data in the real-time vital sign dataset based on the vital sign weights, thereby generating a modified vital sign dataset. This modified vital sign dataset is then input into the identification and early warning model to generate a modified risk probability. Because the vital sign weights reflect the importance of vital sign data in predicting disease and health risks, the modified risk probability generated by inputting the modified vital sign dataset into the identification and early warning model has greater accuracy and sensitivity for disease-related health risks, more accurately reflecting the patient's actual situation and thereby improving the alarm accuracy of the EHR system.

[0043] Step 150: When the revised risk probability is greater than the preset probability threshold, a high-risk revised prediction result is generated and an alarm signal is output; when the revised risk probability is less than or equal to the preset probability threshold, a low-risk revised prediction result is generated and a silent signal is output.

[0044] After obtaining the corrected risk probability, in order to facilitate the judgment of the prediction result type corresponding to the corrected risk probability, this embodiment will compare the corrected risk probability with the preset probability threshold. When the corrected risk probability is greater than the preset probability threshold, it proves that the patient has a higher disease-related health risk and needs to be promptly treated. At this time, a high-risk corrected prediction result should be generated and an alarm signal should be output to cause the EHR system to alarm. When the corrected risk probability is less than or equal to the preset probability threshold, it proves that the patient has a lower disease-related health risk. At this time, a low-risk corrected prediction result should be generated and a silent signal should be output to avoid an EHR system alarm.

[0045] This embodiment can calculate the credibility of the real-time prediction results based on the real-time vital sign data set and the historical sample set. When the credibility of the real-time prediction results is less than or equal to the preset credibility threshold, it can generate a vital sign weight according to the importance of each vital sign data, and correct the real-time vital sign data set according to the vital sign weight, thereby generating a corrected vital sign data set. Subsequently, a corrected risk probability is generated by inputting the corrected vital sign data set into the identification and early warning model, and a corrected prediction result is generated by comparing the corrected risk probability with the preset probability threshold, and an alarm signal or a silent signal is output according to the corrected prediction result. Therefore, a substantial increase in computing cost and data acquisition cost can be avoided, and the corrected risk probability can be made more sensitive through the vital sign weight, so that the prediction result more accurately reflects the actual situation of the patient, thereby improving the alarm accuracy of the EHR system and reducing the probability of false alarms and missed alarms.

[0046] As another optional implementation of this embodiment, before calculating the credibility of the real-time prediction result based on the real-time vital sign data set and the historical sample set, the alarm accuracy improvement method may further include the following steps: The false alarm rate and missed alarm rate of the EHR system are obtained, and when the false alarm rate increases, the preset credibility threshold is increased, and when the missed alarm rate increases, the preset credibility threshold is decreased.

[0047] During normal work, medical staff will respond to the alarm of the EHR system and go to the patient's area for clinical treatment, and record it in the EHR system according to the patient's actual situation after the treatment. Correspondingly, medical staff will also conduct ward rounds and inspections regularly when the EHR system does not alarm, so as to understand the patient's actual situation, and record it in the EHR system based on the results of the ward rounds and inspections. Therefore, the false alarm rate and missed alarm rate of the EHR system can be obtained based on the records of medical staff in the EHR system. When the false alarm rate increases, increasing the preset credibility threshold can make more use of the corrected prediction results to determine whether to trigger an alarm. When the missed alarm rate increases, lowering the preset credibility threshold can make more use of the real-time prediction results to determine whether to trigger an alarm, thereby improving the practical application effect of this method.

[0048] Example 2 Figure 2 This is a flow chart of the method for improving alarm accuracy provided by Example 2 of the present invention. This example is optimized based on the above example. In this example, before calculating the credibility of the real-time prediction result based on the real-time vital sign data set and the historical sample set, the following steps can be added: Determine the normal range of the historical vital sign data set based on the historical sample set; obtain the change direction and change rate of the real-time vital sign data set within a preset time period; if the real-time vital sign data set changes in a direction close to the normal range, the preset credibility threshold is lowered and adjusted, and the lowering adjustment step of the preset credibility threshold is inversely proportional to the change rate; if the real-time vital sign data set changes in a direction away from the normal range, the preset credibility threshold is increased and adjusted, and the higher adjustment step of the preset credibility threshold is directly proportional to the change rate.

[0049] Accordingly, the method for improving alarm accuracy provided in this embodiment specifically includes: Step 210: Acquire a historical sample set, a real-time vital sign data set, and a real-time prediction result of the identification and warning model, wherein the historical sample set includes a plurality of historical prediction results and a historical vital sign data set corresponding to each historical prediction result.

[0050] Step 220: Determine the normal interval of the historical vital sign dataset based on the historical sample set; obtain the change direction and change rate of the real-time vital sign dataset within a preset time period; if the real-time vital sign dataset changes in a direction close to the normal interval, the preset credibility threshold is lowered, and the adjustment step size of the preset credibility threshold is inversely proportional to the change rate; if the real-time vital sign dataset changes in a direction away from the normal interval, the preset credibility threshold is increased, and the adjustment step size of the preset credibility threshold is directly proportional to the change rate.

[0051] Typically, a patient's vital signs data changes with changes in their physical condition. When a patient is healthy, their vital signs data fluctuate within a certain range, which is considered the normal range. When a patient is ill, their vital signs data deviate from the normal range, and the greater the deviation, the more serious the patient's illness.

[0052] In view of this, this embodiment determines the normal interval of the historical vital sign dataset based on the historical sample set, so as to facilitate analysis of changes in the real-time vital sign dataset and thus determine the trend of changes in the patient's physical condition. It should be noted that since the patient's vital sign data will not trigger an alarm in the EHR system when it is within the normal range, when determining the normal interval of the historical vital sign dataset, the historical vital sign dataset in the low-risk historical sample subset should be selected.

[0053] Accordingly, to reflect changes in the real-time vital sign data set, this embodiment also acquires the direction and rate of change of the real-time vital sign data set within a preset time period. The direction of change in the real-time vital sign data can be used to determine whether the patient's physical condition is deteriorating or recovering, while the rate of change in the real-time vital sign data can be used to determine the rate of deterioration or recovery of the patient's physical condition, thereby facilitating flexible adjustment of the preset credibility threshold.

[0054] When the real-time vital sign dataset approaches the normal range, indicating that the patient's physical condition is progressing toward recovery, the preset confidence threshold can be lowered. When the preset confidence threshold is lowered, the EHR system will rely more on the type of real-time prediction results to determine whether to trigger an alarm, thereby reducing the EHR system's computing power requirements.

[0055] When the real-time vital sign dataset deviates from the normal range, indicating a deterioration in the patient's condition, the preset confidence threshold can be adjusted higher. With this increased confidence threshold, the EHR system will rely more heavily on the type of corrected prediction to determine whether to trigger an alarm, resulting in greater accuracy and sensitivity, allowing medical staff to promptly address deteriorating patients.

[0056] In addition, since under normal circumstances the patient's vital signs data will not change rapidly in a short period of time, when the change rate of the vital signs data is high, it can be proved that the patient's physical condition has changed abnormally.

[0057] When the real-time vital sign dataset changes toward the normal range, if the rate of change is high, it clearly does not conform to normal recovery patterns. In this case, the preset confidence threshold should be lowered in an inversely proportional manner to the rate of change (i.e., the higher the rate of change, the smaller the confidence threshold step). This allows the preset confidence threshold to be lowered slowly to prevent the EHR system from prematurely determining whether to trigger an alarm based on the real-time prediction results.

[0058] When the real-time vital sign dataset changes away from the normal range, if the rate of change is high, it indicates that the patient's physical condition is rapidly deteriorating. In this case, the adjustment step size for increasing the preset confidence threshold should be proportional to the rate of change (i.e., the higher the rate of change, the larger the adjustment step size for lowering the confidence threshold). This will quickly increase the preset confidence threshold, allowing the EHR system to more quickly and earlier determine whether to trigger an alarm by correcting the prediction results.

[0059] Step 230: Calculate the credibility of the real-time prediction result based on the real-time vital sign data set and the historical sample set.

[0060] Step 240: When the credibility of the real-time prediction result is less than or equal to the preset credibility threshold, the importance of each physical sign data in the historical physical sign data set is determined based on the historical sample set, and the physical sign weight of the physical sign data is generated based on the importance.

[0061] Step 250: Correct the physical sign data in the real-time physical sign data set according to the physical sign weight to generate a corrected physical sign data set, and input the corrected physical sign data set into the identification and early warning model to generate a corrected risk probability.

[0062] Step 260: When the revised risk probability is greater than the preset probability threshold, a high-risk revised prediction result is generated and an alarm signal is output; when the revised risk probability is less than or equal to the preset probability threshold, a low-risk revised prediction result is generated and a silent signal is output.

[0063] This embodiment adds the following steps before calculating the credibility of the real-time prediction result based on the real-time vital sign dataset and the historical sample set: determining the normal range of the historical vital sign dataset based on the historical sample set; obtaining the direction and rate of change of the real-time vital sign dataset within a preset time period; if the real-time vital sign dataset changes toward the normal range, adjusting the preset credibility threshold downward, with the adjustment step size of the preset credibility threshold being proportional to the rate of change; if the real-time vital sign dataset changes away from the normal range, adjusting the preset credibility threshold upward, with the adjustment step size of the preset credibility threshold being proportional to the rate of change. This allows the preset credibility threshold to be adjusted based on the direction and rate of change of the real-time vital sign dataset within a preset time period, thereby adapting to changes in the patient's physical condition.

[0064] Example 3 Figure 3 This is a flow chart of the method for improving alarm accuracy provided by Example 3 of the present invention. This embodiment is optimized based on the above embodiment. In this embodiment, before correcting the vital sign data in the real-time vital sign data set according to the vital sign weight, the following steps can be added: Determine the normal range of the historical vital sign data set based on the historical sample set; obtain the object affiliation of the real-time vital sign data set and the change direction of the real-time vital sign data set within a preset time period; if the real-time vital sign data set changes in a direction close to the normal range, the vital sign weight is adjusted downward, and the reduction adjustment step when the real-time vital sign data set belongs to a high-risk object is smaller than the reduction adjustment step when the real-time vital sign data set belongs to an ordinary object; if the real-time vital sign data set changes in a direction away from the normal range, the vital sign weight is adjusted upward, and the increase adjustment step when the real-time vital sign data set belongs to a high-risk object is larger than the increase adjustment step when the real-time vital sign data set belongs to an ordinary object.

[0065] Specifically, the method for improving alarm accuracy provided in this embodiment includes: Step 310: Obtain a historical sample set, a real-time vital sign data set, and a real-time prediction result of the identification and warning model. The historical sample set includes multiple historical prediction results and a historical vital sign data set corresponding to each historical prediction result.

[0066] Step 320: Calculate the credibility of the real-time prediction result based on the real-time vital sign data set and the historical sample set.

[0067] Step 330: When the credibility of the real-time prediction result is less than or equal to the preset credibility threshold, the importance of each physical sign data in the historical physical sign data set is determined based on the historical sample set, and the physical sign weight of the physical sign data is generated based on the importance.

[0068] Step 340: Determine the normal interval of the historical vital sign data set based on the historical sample set; obtain the object affiliation of the real-time vital sign data set and the change direction of the real-time vital sign data set within a preset time period; if the real-time vital sign data set changes in a direction close to the normal interval, the vital sign weight is adjusted downward, and the reduction adjustment step when the real-time vital sign data set belongs to a high-risk object is smaller than the reduction adjustment step when the real-time vital sign data set belongs to an ordinary object; if the real-time vital sign data set changes in a direction away from the normal interval, the vital sign weight is adjusted upward, and the increase adjustment step when the real-time vital sign data set belongs to a high-risk object is larger than the increase adjustment step when the real-time vital sign data set belongs to an ordinary object.

[0069] To ensure that the correction of vital sign data adapts to changes in the patient's physical condition, this embodiment adjusts the vital sign weights based on changes in the patient's physical condition. Similar to the above embodiment, this embodiment also determines the normal range of the historical vital sign dataset based on the historical sample set, so as to facilitate analysis of changes in the real-time vital sign dataset and determine the trend of changes in the patient's physical condition.

[0070] When the real-time vital sign data set changes toward the normal range, it proves that the patient's physical condition is tending toward recovery. At this time, the vital sign weight can be adjusted downward, thereby reducing the disease sensitivity of the identification and early warning model when generating the revised prediction results and reducing the probability of triggering an alarm.

[0071] When the real-time vital sign data set changes in the direction away from the normal range, it proves that the patient's physical condition tends to deteriorate. At this time, the vital sign weight can be adjusted to increase the disease sensitivity of the identification and early warning model when generating modified prediction results, so as to more accurately capture the patient's condition deterioration and issue an alarm in time.

[0072] In actual application, due to differences in patients' specific physical fitness (such as age and history of chronic diseases), they can be roughly divided into two types: high-risk patients and general patients. The adjustment step size of the physical sign weights varies according to the patient type. High-risk patients refer to patients with poor physical fitness (such as those aged over 60 or under 12, or those with long-term chronic diseases), while general patients refer to patients with good physical fitness (such as those aged between 12 and 60, with no history of chronic diseases).

[0073] Specifically, when adjusting the sign weight reduction, if the real-time sign data set belongs to a high-risk subject, the sign weight reduction adjustment step should be smaller than the sign weight reduction adjustment step when the real-time sign data set belongs to an ordinary subject, so that the sign weight of the high-risk subject can be reduced in a smoother manner, thereby enabling the high-risk subject to continue to obtain a higher disease prediction sensitivity during the recovery process.

[0074] When adjusting the weight of vital signs, if the real-time vital sign data set belongs to a high-risk subject, the adjustment step size of the weight of vital signs should be larger than the adjustment step size when the real-time vital sign data set belongs to an ordinary subject, so that the vital sign weight of the high-risk subject can be quickly increased when the condition worsens, thereby issuing an alarm in time.

[0075] Step 350: Correct the physical sign data in the real-time physical sign data set according to the physical sign weight to generate a corrected physical sign data set, and input the corrected physical sign data set into the identification and early warning model to generate a corrected risk probability.

[0076] Step 360: When the revised risk probability is greater than the preset probability threshold, a high-risk revised prediction result is generated and an alarm signal is output; when the revised risk probability is less than or equal to the preset probability threshold, a low-risk revised prediction result is generated and a silent signal is output.

[0077] This embodiment adds the following steps before correcting the vital sign data in the real-time vital sign data set according to the vital sign weights: determining the normal range of the historical vital sign data set based on the historical sample set; obtaining the subject affiliation of the real-time vital sign data set and the direction of change of the real-time vital sign data set within a preset time period; if the real-time vital sign data set changes in a direction closer to the normal range, the vital sign weight is adjusted downward, and the adjustment step length when the real-time vital sign data set is assigned to a high-risk subject is smaller than the adjustment step length when the real-time vital sign data set is assigned to an average subject; if the real-time vital sign data set changes in a direction away from the normal range, the vital sign weight is adjusted upward, and the adjustment step length when the real-time vital sign data set is assigned to a high-risk subject is larger than the adjustment step length when the real-time vital sign data set is assigned to an average subject. This allows the vital sign weights to be adjusted according to the direction of change of the real-time vital sign data within a preset time period and the subject affiliation, thereby adapting to changes in the patient's physical condition and physical fitness.

[0078] Example 4 Figure 4 The structural diagram of the alarm accuracy improvement device provided by the fourth embodiment of the present invention is as follows: Figure 4 As shown, the device includes: An acquisition module 410 is configured to acquire a historical sample set, a real-time vital sign data set, and a real-time prediction result of an identification and early warning model, wherein the historical sample set includes a plurality of historical prediction results and a historical vital sign data set corresponding to each historical prediction result; A calculation module 420 is used to calculate the credibility of the real-time prediction result based on the real-time vital sign data set and the historical sample set; A generating module 430 is configured to determine the importance of each type of vital sign data in the historical vital sign data set based on the historical sample set when the credibility of the real-time prediction result is less than or equal to a preset credibility threshold, and generate a vital sign weight for the vital sign data based on the importance; A correction module 440 is used to correct the physical sign data in the real-time physical sign data set according to the physical sign weight to generate a corrected physical sign data set, and input the corrected physical sign data set into the identification and early warning model to generate a corrected risk probability; Output module 450 is used to generate a high-risk revised prediction result and output an alarm signal when the revised risk probability is greater than a preset probability threshold, and to generate a low-risk revised prediction result and output a silent signal when the revised risk probability is less than or equal to the preset probability threshold.

[0079] The alarm accuracy improvement device provided in this embodiment can obtain the historical sample set, real-time vital sign data set and real-time prediction results of the identification warning model through the acquisition module, calculate the credibility of the real-time prediction results through the calculation module, and generate a vital sign weight through the generation module when the credibility of the real-time prediction results is less than or equal to the preset credibility threshold. The real-time vital sign data set is then corrected through the correction module to facilitate the generation of a corrected risk probability, and finally a corrected prediction result is generated through the output module, and whether to trigger an alarm is determined based on the type of the corrected prediction result. Compared with the existing technology, it can avoid a significant increase in computing cost and data acquisition cost, and can make the corrected risk probability have better sensitivity through the vital sign weight, so that the prediction result can more accurately reflect the actual situation of the patient, thereby improving the alarm accuracy of the EHR system and reducing the probability of false alarms and missed alarms.

[0080] Based on the above embodiment, the calculation module includes: The first calculation unit is used to calculate the consistency measurement value of each historical vital sign data set in the high-risk historical sample subset, and sort the consistency measurement values of each historical vital sign data set in order from small to large to generate a high-risk historical consistency list; The second calculation unit is used to calculate the consistency measurement value of each historical vital sign data set in the low-risk historical sample subset, and sort the consistency measurement values of each historical vital sign data set in order from small to large to generate a low-risk historical consistency list; a first credibility determination unit, configured to, when the real-time prediction result is a high-risk real-time prediction result, use the real-time vital sign dataset as a newly added vital sign dataset in the high-risk historical sample subset, calculate a consistency metric value of the newly added vital sign dataset, and determine the credibility of the real-time prediction result according to a sorting position of the consistency metric value of the newly added vital sign dataset in the high-risk historical consistency list; The second credibility determination unit is used to, when the real-time prediction result is a low-risk real-time prediction result, use the real-time vital sign data set as a newly added vital sign data set in the low-risk historical sample subset, calculate the consistency measurement value of the newly added vital sign data set, and determine the credibility of the real-time prediction result according to the sorting position of the consistency measurement value of the newly added vital sign data set in the low-risk historical consistency list.

[0081] Based on the above embodiment, the device includes: The first adjustment module is used to determine the normal range of the historical vital sign data set based on the historical sample set; obtain the change direction and change rate of the real-time vital sign data set within a preset time period; if the real-time vital sign data set changes in a direction close to the normal range, the preset credibility threshold is lowered and adjusted, and the adjustment step size of the preset credibility threshold is inversely proportional to the change rate; if the real-time vital sign data set changes in a direction away from the normal range, the preset credibility threshold is increased and adjusted, and the adjustment step size of the preset credibility threshold is directly proportional to the change rate.

[0082] Based on the above embodiment, the device includes: The second adjustment module is used to obtain the false alarm rate and missed alarm rate of the EHR system, increase the preset credibility threshold when the false alarm rate increases, and reduce the preset credibility threshold when the missed alarm rate increases.

[0083] Based on the above embodiment, the device includes: The third adjustment module is used to determine the normal range of the historical vital sign data set based on the historical sample set; obtain the object affiliation of the real-time vital sign data set and the change direction of the real-time vital sign data set within a preset time period; if the real-time vital sign data set changes in a direction close to the normal range, the vital sign weight is adjusted downward, and the reduction adjustment step when the real-time vital sign data set belongs to a high-risk object is smaller than the reduction adjustment step when the real-time vital sign data set belongs to an ordinary object; if the real-time vital sign data set changes in a direction away from the normal range, the vital sign weight is adjusted upward, and the increase adjustment step when the real-time vital sign data set belongs to a high-risk object is larger than the increase adjustment step when the real-time vital sign data set belongs to an ordinary object.

[0084] Based on the above embodiment, the device further includes: The real-time prediction result judgment module is used to judge the type of the real-time prediction result when the credibility of the real-time prediction result is greater than the preset credibility threshold. If the real-time prediction result is a high-risk real-time prediction result, an alarm signal is output; if the real-time prediction result is a low-risk real-time prediction result, a silent signal is output.

[0085] The alarm accuracy improvement device provided in the embodiment of the present invention can execute the alarm accuracy improvement method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0086] Example 5 Figure 5 This is a structural diagram of a device provided in Example 5 of the present invention. Figure 5 A block diagram of an exemplary device 12 suitable for use in implementing embodiments of the present invention is shown. Figure 5 The device 12 shown is only an example and should not bring any limitation to the functionality and scope of use of the embodiments of the present invention.

[0087] like Figure 5 As shown, device 12 is implemented as a general-purpose computing device. Components of device 12 may include, but are not limited to, one or more processors or processing units 16, a system memory 28, and a bus 18 that connects various system components (including system memory 28 and processing unit 16).

[0088] Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor, or a local bus using any of a variety of bus architectures. Examples of these architectures include, but are not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MAC) bus, an Enhanced ISA bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus.

[0089] Device 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by device 12, including volatile and non-volatile media, removable and non-removable media.

[0090] System memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache 32. Device 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 34 may be configured to read and write non-removable, non-volatile magnetic media ( Figure 5 Not shown, usually called a "hard drive"). Although Figure 5Although not shown, a magnetic disk drive for reading and writing to a removable non-volatile magnetic disk (e.g., a "floppy disk"), as well as an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to bus 18 via one or more data media interfaces. System memory 28 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of various embodiments of the present invention.

[0091] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in system memory 28. Such program modules 42 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data, each of which, or some combination thereof, may include an implementation of a network environment. Program modules 42 generally implement the functions and / or methodologies of the embodiments described herein.

[0092] Device 12 may also communicate with one or more external devices 14 (e.g., a keyboard, pointing device, display 24, etc.), one or more devices that enable a user to interact with device 12, and / or any device that enables device 12 to communicate with one or more other computing devices (e.g., a network card, a modem, etc.). This communication may occur via input / output (I / O) interface 22. Furthermore, device 12 may communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network such as the Internet) via network adapter 20. As shown, network adapter 20 communicates with other modules of device 12 via bus 18. It should be understood that, although not shown, other hardware and / or software modules may be used in conjunction with device 12, including but not limited to microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0093] The processing unit 16 executes various functional applications and data processing by running programs stored in the system memory 28, such as implementing the alarm accuracy improvement method provided in the embodiment of the present invention.

[0094] Example 6 Embodiment 6 of the present invention further provides a storage medium containing computer-executable instructions, which, when executed by a computer processor, are used to execute any of the alarm accuracy improvement methods provided in the above embodiments.

[0095] The computer storage medium of the embodiments of the present invention may adopt any combination of one or more computer-readable media. The computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0096] A computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0097] Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0098] Computer program code for performing the operations of the present invention may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone 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 through 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., through the Internet using an Internet service provider).

[0099] Example 7 An embodiment of the present invention provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the alarm accuracy improvement method provided in the above embodiment.

[0100] Note that the above are only preferred embodiments of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments, and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments and may include many other equivalent embodiments without departing from the concept of the present invention. The scope of the present invention is determined by the scope of the appended claims.

Claims

1. A method for improving alarm accuracy, characterized in that include: Obtaining a historical sample set, a real-time vital sign data set, and a real-time prediction result of an identification and warning model, wherein the historical sample set includes a plurality of historical prediction results and a historical vital sign data set corresponding to each historical prediction result; Calculate the credibility of real-time prediction results based on real-time vital sign data sets and historical sample sets; When the credibility of the real-time prediction result is less than or equal to the preset credibility threshold, determining the importance of each physical sign data in the historical physical sign data set according to the historical sample set, and generating a physical sign weight of the physical sign data according to the importance; Correcting the physical sign data in the real-time physical sign data set according to the physical sign weight to generate a corrected physical sign data set, and inputting the corrected physical sign data set into the identification and early warning model to generate a corrected risk probability; When the corrected risk probability is greater than the preset probability threshold, a high-risk corrected prediction result is generated and an alarm signal is output; when the corrected risk probability is less than or equal to the preset probability threshold, a low-risk corrected prediction result is generated and a silent signal is output.

2. The method for improving alarm accuracy according to claim 1, characterized in that: The real-time prediction results include high-risk real-time prediction results and low-risk real-time prediction results; The historical sample set includes a high-risk historical sample subset and a low-risk historical sample subset; The high-risk historical sample subset includes high-risk historical prediction results and historical physical sign data sets corresponding to the high-risk historical prediction results; The low-risk historical sample subset includes low-risk historical prediction results and historical physical sign data sets corresponding to the low-risk historical prediction results; The calculation of the credibility of the real-time prediction result based on the real-time vital sign data set and the historical sample set includes: Calculate the consistency metric value of each historical vital sign data set in the high-risk historical sample subset, and sort the consistency metric values of each historical vital sign data set in ascending order to generate a high-risk historical consistency list; Calculate the consistency metric value of each historical vital sign data set in the low-risk historical sample subset, and sort the consistency metric values of each historical vital sign data set in ascending order to generate a low-risk historical consistency list; When the real-time prediction result is a high-risk real-time prediction result, the real-time vital sign dataset is used as a newly added vital sign dataset in the high-risk historical sample subset, the consistency metric value of the newly added vital sign dataset is calculated, and the credibility of the real-time prediction result is determined according to the sorting position of the consistency metric value of the newly added vital sign dataset in the high-risk historical consistency list; When the real-time prediction result is a low-risk real-time prediction result, the real-time vital sign data set is used as a new vital sign data set in the low-risk historical sample subset, the consistency measurement value of the new vital sign data set is calculated, and the credibility of the real-time prediction result is determined based on the sorting position of the consistency measurement value of the new vital sign data set in the low-risk historical consistency list.

3. The method for improving alarm accuracy according to claim 1, characterized in that: Before calculating the credibility of the real-time prediction result based on the real-time vital sign data set and the historical sample set, the alarm accuracy improvement method further includes: Determine the normal interval of the historical physical sign data set based on the historical sample set; The change direction and change rate of the real-time vital sign data set within a preset time period are obtained. If the real-time vital sign data set changes in a direction approaching the normal interval, the preset credibility threshold is adjusted downward, and the adjustment step size of the preset credibility threshold is inversely proportional to the change rate. If the real-time vital sign data set changes in a direction away from the normal interval, the preset credibility threshold is adjusted upward, and the adjustment step size of the preset credibility threshold is directly proportional to the change rate.

4. The method for improving alarm accuracy according to claim 1, characterized in that: Before calculating the credibility of the real-time prediction result based on the real-time vital sign data set and the historical sample set, the alarm accuracy improvement method further includes: The false alarm rate and missed alarm rate of the EHR system are obtained, and when the false alarm rate increases, the preset credibility threshold is increased, and when the missed alarm rate increases, the preset credibility threshold is decreased.

5. The method for improving alarm accuracy according to claim 1, characterized in that: After calculating the credibility of the real-time prediction result based on the real-time vital sign data set and the historical sample set, the alarm accuracy improvement method further includes: When the credibility of the real-time prediction result is greater than the preset credibility threshold, the type of the real-time prediction result is judged. If the real-time prediction result is a high-risk real-time prediction result, an alarm signal is output; if the real-time prediction result is a low-risk real-time prediction result, a silent signal is output.

6. The method for improving alarm accuracy according to claim 1, characterized in that: Before modifying the vital sign data in the real-time vital sign data set according to the vital sign weight, the method for improving alarm accuracy further includes: Determine the normal interval of the historical physical sign data set based on the historical sample set; Obtain the object attribution of the real-time vital sign data set and the change direction of the real-time vital sign data set within a preset time period; if the real-time vital sign data set changes in a direction approaching the normal interval, the vital sign weight is adjusted downward, and the reduction adjustment step length when the real-time vital sign data set belongs to a high-risk object is smaller than the reduction adjustment step length when the real-time vital sign data set belongs to an ordinary object; if the real-time vital sign data set changes in a direction away from the normal interval, the vital sign weight is adjusted upward, and the increase adjustment step length when the real-time vital sign data set belongs to a high-risk object is larger than the increase adjustment step length when the real-time vital sign data set belongs to an ordinary object.

7. A device for improving alarm accuracy, characterized in that: include: An acquisition module is used to acquire a historical sample set, a real-time vital sign data set, and a real-time prediction result of an identification and warning model, wherein the historical sample set includes a plurality of historical prediction results and a historical vital sign data set corresponding to each historical prediction result; A calculation module, used to calculate the credibility of real-time prediction results based on the real-time vital sign data set and the historical sample set; a generating module, configured to determine the importance of each type of vital sign data in the historical vital sign data set based on the historical sample set when the credibility of the real-time prediction result is less than or equal to a preset credibility threshold, and generate a vital sign weight for the vital sign data based on the importance; A correction module is used to correct the physical sign data in the real-time physical sign data set according to the physical sign weight to generate a corrected physical sign data set, and input the corrected physical sign data set into the identification and early warning model to generate a corrected risk probability; The output module is used to generate a high-risk correction prediction result and output an alarm signal when the corrected risk probability is greater than the preset probability threshold; and to generate a low-risk correction prediction result and output a silent signal when the corrected risk probability is less than or equal to the preset probability threshold.

8. A device, characterized in that The device comprises: one or more processors; a storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the alarm accuracy improvement method as described in any one of claims 1-6.

9. A storage medium comprising computer executable instructions, wherein the computer executable instructions, when executed by a computer processor, are used to execute the alarm accuracy improvement method according to any one of claims 1 to 6.

10. A computer product, comprising a computer program, characterized in that: When the computer program is executed by a processor, the alarm accuracy improvement method according to any one of claims 1 to 6 is implemented.

Images