Networking monitoring method and system for critical patients

Abstract

The invention relates to the technical field of critical patient monitoring, in particular to a network monitoring method and system for critical patients. According to the method, the processed physiological data and the basic information of the patient are input into the Bayesian network to obtain the probability that the patient is abnormal, so that whether the patient is abnormal or not can be more accurately judged, whether early warning is sent out or not is determined according to the comparison result of the probability and the threshold value, a notification can be more effectively sent out for treatment, and the treatment efficiency is improved. Therefore, the monitoring and treatment efficiency is improved; whether the networking monitoring of the critical patient is qualified or not is judged through the average value of the ratios of the number of times of early warning in the multiple monitoring periods to the number of times of abnormity of the patient, so that whether the networking monitoring of the critical patient is qualified or not can be judged more quickly, and a corresponding processing mode is generated based on the reason of disqualification. Therefore, networking monitoring is more accurate, and the efficiency of networking monitoring of critical patients is further improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of critical patient monitoring, in particular to a critical patient networking monitoring method and system. BACKGROUND

[0002] The vital sign monitoring of critical patients is the core link of medical treatment. Traditional monitoring relies on bedside equipment, which has problems such as data island and response delay. With the development of Internet of Things (IoT), 5G communication, edge computing and artificial intelligence technology, networking monitoring system has gradually become an important technical support for intensive care.

[0003] Chinese patent publication No. CN117275648B discloses an intelligent nursing method for patients in an intensive care unit based on the Internet of Things, which comprises the following steps: obtaining the condition data of the patients in the ward, formulating and implementing the corresponding nursing plan; real-time monitoring of the physiological data of the patients through the sensor devices pre-deployed in the ward; collecting and analyzing the physiological data monitored by the sensor devices to determine the physiological state of the patients and analyze whether to issue a warning; judging the demand state of the patients based on their physiological state and the formulated nursing plan, and scheduling medical staff in real time according to the judgment result; collecting the nursing feedback of the medical staff and adjusting the nursing plan according to the physiological state of the patients.

[0004] As can be seen, the prior art has the following problems: it cannot accurately monitor the abnormalities of critical patients and issue warnings, and most of the time, the abnormalities of critical patients are discovered by humans, resulting in low monitoring efficiency. SUMMARY

[0005] Therefore, the present application provides a critical patient networking monitoring method and system to overcome the problem that the prior art cannot accurately monitor the abnormalities of critical patients and issue warnings, and most of the time, the abnormalities of critical patients are discovered by humans, resulting in low monitoring efficiency.

[0006] To achieve the above-mentioned purpose, the present application provides a critical patient networking monitoring method, comprising:

[0007] Real-time monitoring of the vital signs of the patient to obtain physiological data;

[0008] Obtaining the basic information of the patient;

[0009] Pretreating the physiological data of the patient based on Kalman filtering;

[0010] Inputting the pretreated physiological data and the basic information into a Bayesian network to obtain the probability of the patient's abnormality;

[0011] Determining whether to issue a warning based on the comparison result of the probability and the threshold value;

[0012] The network monitoring of critically ill patients is judged as qualified based on the average ratio of the number of warnings issued to the number of abnormalities in patients within multiple monitoring cycles. If the monitoring is not qualified, a corresponding processing method is generated based on the reason for the failure. The processing method includes adjusting the process noise covariance matrix in the Kalman filter, adjusting the optimization cycle of the Bayesian network, and adjusting the threshold of the night warning.

[0013] Adjustments are made based on the corresponding processing method.

[0014] Furthermore, the process of determining whether the online monitoring of critically ill patients is qualified based on the average ratio of the number of warnings issued to the number of abnormalities in patients within multiple monitoring periods includes: calculating the average ratio of the number of warnings issued to the number of abnormalities in patients within multiple monitoring periods; determining whether the online monitoring of critically ill patients is qualified based on the comparison result of the average value and a pre-stored preset average value; if the average value is greater than or equal to a first preset average value, the online monitoring of critically ill patients is qualified; if the average value is less than the first preset average value but greater than a second preset average value, the online monitoring of critically ill patients is qualified based on the variance of the ratio of the number of warnings issued to the number of abnormalities in patients within multiple monitoring periods; if the average value is less than or equal to the second preset average value, the online monitoring of critically ill patients is unqualified, and the reason for the unqualified online monitoring of critically ill patients is determined based on the difference between the preset average value and the average value.

[0015] Furthermore, the process of determining whether the online monitoring of critically ill patients is qualified based on the variance of the ratio of the number of warnings issued to the number of abnormalities in patients within multiple monitoring periods includes: calculating the variance of the ratio of the number of warnings issued to the number of abnormalities in patients within multiple monitoring periods; comparing the variance with a preset variance; if the variance is greater than the preset variance, adjusting the threshold based on the difference between the variance and the preset variance; if the variance is less than or equal to the preset variance, determining that the online monitoring of critically ill patients is unqualified, and determining the reason for the unqualified online monitoring of critically ill patients based on the difference between the preset average and the average.

[0016] Furthermore, the process of adjusting the threshold based on the difference between the variance and the preset variance includes: reducing the threshold based on the difference between the variance and the preset variance, wherein the difference is proportional to the reduction of the threshold.

[0017] Further, the process of determining the reasons for the failure of online monitoring of critically ill patients based on the difference between the preset average value and the average value includes: calculating the difference between the preset average value and the average value; determining the reasons for the failure of online monitoring of critically ill patients based on the comparison results of the difference with the pre-stored preset difference value; if the difference is greater than or equal to a first preset difference value, adjusting the process noise covariance matrix in the Kalman filter based on the ratio of the difference to the preset difference value; if the difference is less than the first preset difference value and greater than a second preset difference value, plotting the difference curve between the number of abnormal occurrences of patients and the number of warnings, and determining the reasons for the failure of online monitoring of critically ill patients based on the curve integral; if the difference is less than or equal to the second preset difference value, counting the number of abnormal ratios in multiple monitoring periods, and adjusting the optimization period of the Bayesian network based on the number; wherein, the abnormal ratio is the ratio of the number of warnings issued to the number of abnormal occurrences of patients in each monitoring period to a preset abnormal ratio value.

[0018] Furthermore, the process of adjusting the process noise covariance matrix in the Kalman filter based on the ratio of the difference to the preset difference includes: increasing the process noise covariance matrix in the Kalman filter based on the ratio of the difference to the preset difference, and the ratio is proportional to the increase of the process noise covariance matrix.

[0019] Furthermore, the process of determining the reasons for the failure of network monitoring of critically ill patients based on curve integrals includes: comparing the curve integral with a preset curve integral; if the curve integral is greater than the preset curve integral, adjusting the threshold of the nighttime warning based on the difference between the curve integral and the preset curve integral; if the curve integral is less than or equal to the preset curve integral, adjusting the optimization period of the Bayesian network based on the number of abnormal ratios in multiple monitoring periods.

[0020] Furthermore, the process of adjusting the nighttime warning threshold based on the difference between the curve integral and the preset curve integral includes: reducing the nighttime warning threshold based on the difference between the curve integral and the preset curve integral, and the difference is proportional to the reduction of the nighttime warning threshold.

[0021] Furthermore, the process of adjusting the optimization period of the Bayesian network based on the number of anomaly ratios within multiple monitoring periods includes: reducing the optimization period of the Bayesian network based on the number of anomalies, with the number being proportional to the reduction in the optimization period.

[0022] To achieve the above objectives, the present invention provides a networked monitoring system for critically ill patients, comprising:

[0023] The data monitoring unit is used to monitor the patient's vital signs in real time to obtain physiological data;

[0024] The information acquisition unit is used to acquire basic information about the patient.

[0025] A preprocessing unit, connected to the data monitoring unit, is used to preprocess the patient's physiological data based on Kalman filtering;

[0026] A probability output unit, which is connected to the information acquisition unit and the preprocessing unit respectively, is used to input the preprocessed physiological data and the basic information into a Bayesian network to obtain the probability of the patient having an abnormality.

[0027] An early warning unit, which is connected to the probability output unit, is used to determine whether to issue an early warning based on the comparison result between the probability and the threshold.

[0028] An analysis unit, connected to the early warning unit, is used to determine whether the network monitoring of critically ill patients is qualified based on the average ratio of the number of early warnings issued to the number of abnormalities in patients within multiple monitoring cycles. If the monitoring is unqualified, a corresponding processing method is generated based on the reason for the unqualification. The processing method includes adjusting the process noise covariance matrix in the Kalman filter, adjusting the optimization period of the Bayesian network, and adjusting the threshold of the nighttime early warning.

[0029] A control unit, connected to the analysis unit, is used to adjust based on the corresponding processing method.

[0030] Compared with existing technologies, the beneficial effects of this invention are as follows: By inputting processed physiological data and basic patient information into a Bayesian network, the probability of patient abnormalities is obtained, thereby enabling more accurate judgment of whether a patient is experiencing abnormalities. Based on the comparison between the probability and a threshold, an alert is issued, allowing for more effective notification and treatment, thus improving the efficiency of monitoring and treatment. Furthermore, by averaging the ratio of the number of alerts issued to the number of patient abnormalities across multiple monitoring periods, the network monitoring of critically ill patients is determined to be qualified, enabling faster assessment of the adequacy of network monitoring for critically ill patients. Based on the reasons for non-compliance, corresponding processing methods are generated, making network monitoring more accurate and further improving the efficiency of network monitoring for critically ill patients.

[0031] Furthermore, this invention determines whether the online monitoring of critically ill patients is qualified by using the average ratio of the number of warnings issued to the number of abnormalities in patients within multiple monitoring cycles. This allows for a faster determination of whether the online monitoring of critically ill patients is qualified, enabling more precise adjustments in the event of non-compliance, thereby further improving the accuracy of monitoring.

[0032] Furthermore, this invention determines whether the network monitoring of critically ill patients is qualified by using the variance of the ratio of the number of warnings issued to the number of abnormalities in patients within multiple monitoring cycles. This allows for a more accurate determination of whether the network monitoring of critically ill patients is qualified, enabling more precise adjustments in the event of non-compliance, thereby further improving the accuracy of monitoring.

[0033] Furthermore, by adjusting the warning threshold based on the difference between the variance and the preset variance, this invention can adjust the warning threshold when the probability fluctuates around the threshold and most probabilities are slightly less than the threshold, so that the warning can be issued more accurately based on the comparison results between the probability and the threshold, thereby enabling more accurate monitoring of critically ill patients.

[0034] Furthermore, by determining the reasons for the failure of network monitoring of critically ill patients based on the difference between preset average values ​​and average values, the present invention can more accurately determine the reasons for the failure of network monitoring of critically ill patients, thereby enabling more precise adjustments in the future and further improving the accuracy of network monitoring of critically ill patients.

[0035] Furthermore, by adjusting the process noise covariance matrix in the Kalman filter based on the ratio of the difference to a preset difference, this invention can reduce the impact of noise on the acquisition results during the model state transition, thereby making the acquired physiological data more accurate, and consequently making the output results of the Bayesian network more accurate.

[0036] Furthermore, this invention determines the reasons for the failure of online monitoring of critically ill patients by using curve integrals. It can determine whether the failure of online monitoring of critically ill patients is caused by inaccurate nighttime warnings based on the comparison results of curve integrals and preset curve integrals. This allows for more accurate and targeted adjustments to be made subsequently, resulting in more accurate online monitoring results.

[0037] Furthermore, by adjusting the threshold of nighttime warnings based on the difference between the curve integral and a preset curve integral, the present invention can make nighttime warnings more accurate, thereby improving the accuracy of network monitoring.

[0038] Furthermore, by adjusting the optimization period of the Bayesian network based on the number of anomaly ratios within multiple monitoring periods, this invention can more effectively adjust the optimization period of the Bayesian network, making the network more adaptable to the current stage of monitoring, thereby more accurately monitoring the physical condition of critically ill patients. Attached Figure Description

[0039] Figure 1 This is a schematic diagram of the structure of a networked monitoring system for critically ill patients according to an embodiment of the present invention;

[0040] Figure 2This is a flowchart illustrating the steps of a networked monitoring method for critically ill patients according to an embodiment of the present invention.

[0041] Figure 3 This is a flowchart illustrating the steps of determining the results of comparing the average of the ratio of the number of warnings issued to the number of abnormalities in patients within multiple monitoring periods with a pre-stored preset average value in an embodiment of the present invention.

[0042] Figure 4 This is a flowchart illustrating the steps of determining the comparison result between the preset average value and the difference between the preset average value and the pre-stored preset difference value in an embodiment of the present invention. Detailed Implementation

[0043] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0044] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0045] It should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0046] Please see Figure 1 As shown, it is a schematic diagram of the structure of the networked monitoring system for critically ill patients according to an embodiment of the present invention.

[0047] The system includes a data monitoring unit, an information acquisition unit, a preprocessing unit, a probability output unit, an early warning unit, an analysis unit, and a control unit.

[0048] The data monitoring unit is used to monitor the patient's vital signs in real time to obtain physiological data;

[0049] The information acquisition unit is used to acquire the patient's basic information;

[0050] The preprocessing unit is connected to the data monitoring unit and is used to preprocess the patient's physiological data based on Kalman filtering.

[0051] The probability output unit is connected to the information acquisition unit and the preprocessing unit respectively. It is used to input the preprocessed physiological data and the basic information into the Bayesian network to obtain the probability of the patient having an abnormality.

[0052] The warning unit is connected to the probability output unit, and is used to determine whether to issue a warning based on the comparison result of the probability and the threshold.

[0053] The analysis unit is connected to the early warning unit. It is used to determine whether the network monitoring of critically ill patients is qualified based on the average ratio of the number of early warnings issued to the number of abnormalities in patients within multiple monitoring cycles. If the monitoring is unqualified, a corresponding processing method is generated based on the reason for the unqualification. The processing method includes adjusting the process noise covariance matrix in the Kalman filter, adjusting the optimization period of the Bayesian network, and adjusting the threshold of the nighttime early warning.

[0054] The control unit is connected to the analysis unit and is used to make adjustments based on the corresponding processing method.

[0055] Specifically, in this embodiment, physiological data obtained by monitoring the patient's vital signs are input into a Bayesian network after being processed using Kalman filtering, along with the patient's basic personal information. The network outputs the probability of the patient exhibiting abnormalities, enabling a more accurate determination of whether the patient's physical condition is abnormal based on the patient's basic situation. Simultaneously, a warning is issued based on the comparison between the probability and a threshold; that is, a warning is issued when the probability is greater than or equal to the threshold, allowing the patient to receive treatment more promptly. Furthermore, the network monitoring of critically ill patients is judged based on the average ratio of the number of warnings issued to the number of abnormalities observed over multiple monitoring periods. If the monitoring is deemed unqualified, a corresponding processing method is generated based on the reason for the unqualified status, enabling a faster determination of the accuracy of the network monitoring results and allowing adjustments to be made based on the determined reasons when the detection results are deemed unqualified.

[0056] Please see Figure 2 The diagram shown is a flowchart illustrating the steps of a network monitoring method for critically ill patients according to an embodiment of the present invention.

[0057] The steps of the system described in this embodiment of the invention during actual operation include:

[0058] S1, through the data monitoring unit, monitors the patient's vital signs in real time to obtain physiological data;

[0059] S2, acquires the patient's basic information through the information acquisition unit;

[0060] S3, the patient's physiological data is preprocessed based on Kalman filtering by a preprocessing unit connected to the data monitoring unit;

[0061] S4, the preprocessed physiological data and basic information are input into a Bayesian network through a probability output unit that is connected to the information acquisition unit and the preprocessing unit respectively, to obtain the probability of the patient having an abnormality;

[0062] S5, the warning unit connected to the probability output unit determines whether to issue a warning based on the comparison result between the probability and the threshold;

[0063] S6, the analysis unit connected to the early warning unit determines whether the network monitoring of critically ill patients is qualified based on the average ratio of the number of early warnings issued to the number of abnormalities in patients within multiple monitoring cycles, and generates a corresponding processing method based on the reason for the failure when the monitoring is unqualified. The processing method includes adjusting the process noise covariance matrix in the Kalman filter, adjusting the optimization period of the Bayesian network, and adjusting the threshold of the nighttime early warning.

[0064] S7, the control unit connected to the analysis unit adjusts the process based on the corresponding processing method.

[0065] Please see Figure 3 The diagram illustrates the steps of determining whether the online monitoring of a critically ill patient is qualified based on the average of the ratio of the number of warnings issued to the number of abnormal patient occurrences within multiple monitoring periods, as shown in this embodiment of the invention, compared with a pre-stored preset average. The process of determining whether the online monitoring of a critically ill patient is qualified based on the average of the ratio of the number of warnings issued to the number of abnormal patient occurrences within multiple monitoring periods includes: calculating the average of the ratio of the number of warnings issued to the number of abnormal patient occurrences within multiple monitoring periods; determining whether the online monitoring of a critically ill patient is qualified based on the comparison result of the average value and the pre-stored preset average value; if the average value is greater than or equal to a first preset average value, the online monitoring of the critically ill patient is deemed qualified; if the average value is less than the first preset average value but greater than a second preset average value, the online monitoring of the critically ill patient is deemed qualified based on the variance of the ratio of the number of warnings issued to the number of abnormal patient occurrences within multiple monitoring periods; if the average value is less than or equal to the second preset average value, the online monitoring of the critically ill patient is deemed unqualified, and the reason for the unqualified online monitoring of the critically ill patient is determined based on the difference between the preset average value and the average value.

[0066] Specifically, in this embodiment, the average value L0 can be divided into a first preset average value L1 and a second preset average value L2. The first preset average value L1 = 0.87 and the second preset average value L2 = 0.60 are set in the average value standard. It should be noted that in other embodiments, the values ​​of L1 and L2 can also be determined based on relative monitoring requirements. The comparison process between the average value L and L1 and L2 is as follows:

[0067] If the average value L is greater than or equal to the first preset average value L1, the network monitoring of critically ill patients is deemed qualified.

[0068] If the average value L is less than the first preset average value L1 and greater than the second preset average value L2, it means that it is impossible to determine whether there are other factors that cause this result. Then, the network monitoring of critically ill patients is qualified based on the variance P of the ratio of the number of warnings issued to the number of abnormalities in patients within multiple monitoring cycles.

[0069] If the average value L is less than or equal to the second preset average value L2, the network monitoring of critically ill patients is deemed unqualified. Then, the reason for the unqualified network monitoring of critically ill patients is determined based on the difference Q between the preset average value and the average value.

[0070] Specifically, the process of determining whether the online monitoring of critically ill patients is qualified based on the variance of the ratio of the number of warnings issued to the number of abnormalities in patients within multiple monitoring periods in this embodiment of the invention includes: calculating the variance of the ratio of the number of warnings issued to the number of abnormalities in patients within multiple monitoring periods; comparing the variance with a preset variance; if the variance is greater than the preset variance, adjusting the threshold based on the difference between the variance and the preset variance; if the variance is less than or equal to the preset variance, determining that the online monitoring of critically ill patients is unqualified, and determining the reason for the unqualified online monitoring of critically ill patients based on the difference between the preset average value and the average value.

[0071] Specifically, in this embodiment, the preset variance P0 = 0.95, and the comparison process based on variance P and preset variance P0 is as follows:

[0072] If the variance P is greater than the preset variance P0, it indicates that the ratio dispersion is high, and that the probability of abnormality of patients output by the Bayesian network is distributed around the threshold, and most of the probabilities are slightly less than the threshold. Then, the threshold is adjusted based on the difference R between the variance and the preset variance.

[0073] If the variance P is less than or equal to the preset variance P0, it indicates that the ratio dispersion is low, and the network monitoring of critically ill patients is deemed unqualified. Then, the reason for the unqualified network monitoring of critically ill patients is determined based on the difference Q between the preset average and the average.

[0074] Specifically, the process of adjusting the threshold based on the difference between the variance and the preset variance in this embodiment of the invention includes: reducing the threshold based on the difference between the variance and the preset variance, and the difference is proportional to the reduction of the threshold.

[0075] Specifically, in this embodiment, the preset difference R0 = 0.15, and the comparison process based on the difference R and the preset difference R0 is as follows:

[0076] If the difference R is less than or equal to the preset difference R0, the threshold is adjusted to 0.95 times the original threshold.

[0077] If the difference R is greater than the preset difference R0, the threshold is adjusted to 0.8 times the original threshold.

[0078] Please see Figure 4 The diagram illustrates the steps of determining the reasons for the failure of online monitoring of critically ill patients based on the comparison between the difference between preset average values ​​and the average values ​​in an embodiment of the present invention. The process of determining the reasons for the failure of online monitoring of critically ill patients based on the difference between preset average values ​​and the average values ​​in an embodiment of the present invention includes: calculating the difference between preset average values ​​and the average values; determining the reasons for the failure of online monitoring of critically ill patients based on the comparison between the difference and the pre-stored preset differences; if the difference is greater than or equal to a first preset difference, adjusting the process noise covariance matrix in the Kalman filter based on the ratio of the difference to the preset difference; if the difference is less than the first preset difference and greater than a second preset difference, plotting the difference curve between the number of abnormal patient occurrences and the number of warnings, and determining the reasons for the failure of online monitoring of critically ill patients based on the curve integral; if the difference is less than or equal to the second preset difference, counting the number of abnormal ratios in multiple monitoring periods, and adjusting the optimization period of the Bayesian network based on the number; wherein, the abnormal ratio is the ratio of the number of warnings issued to the number of abnormal patient occurrences in each monitoring period to a preset abnormal ratio.

[0079] Specifically, in this embodiment, the difference Q0 can be divided into a first preset difference Q1 and a second preset difference Q2. The preset difference standard is set so that the first preset difference Q1 = 0.31 and the second preset difference Q2 = 0.14. It should be noted that in other embodiments, the values ​​of Q1 and Q2 can also be determined according to the corresponding monitoring requirements. The comparison process based on the difference Q with Q1 and Q2 is as follows:

[0080] If the difference Q is greater than or equal to the first preset difference Q1, it indicates that there is a problem in the process of preprocessing the data using Kalman filtering. Then, the process noise covariance matrix in Kalman filtering is adjusted based on the ratio T between the difference and the preset difference.

[0081] If the difference Q is less than the first preset difference Q1 and greater than the second preset difference Q2, it indicates that the monitoring may be unqualified due to the inaccuracy of the nighttime warning. Then, a difference curve between time-number of warnings and number of abnormalities in patients is plotted, and the reason for the unqualified network monitoring of critically ill patients is determined based on the curve integral U.

[0082] If the difference Q is less than or equal to the second preset difference Q2, it indicates that the optimization cycle of the Bayesian network is too long and the current network is not suitable for detecting the current data. In this case, the number of abnormal ratios in multiple monitoring cycles is counted, and the optimization cycle of the Bayesian network is adjusted based on the number V.

[0083] Specifically, the process of adjusting the process noise covariance matrix in the Kalman filter based on the ratio of the difference to a preset difference in this embodiment of the invention includes: increasing the process noise covariance matrix in the Kalman filter based on the ratio of the difference to a preset difference, and the ratio is proportional to the increase in the process noise covariance matrix.

[0084] Specifically, in this embodiment, the preset ratio T0 = 1.3, and the comparison process based on the ratio T and the preset ratio T0 is as follows:

[0085] If the ratio T is less than or equal to the preset ratio T0, the process noise covariance matrix in the Kalman filter will be adjusted to 1.2 times the original process noise covariance matrix.

[0086] If the ratio T is greater than the preset ratio T0, the process noise covariance matrix in the Kalman filter will be adjusted to 1.75 times the original process noise covariance matrix.

[0087] Specifically, the process of determining the reasons for the failure of network monitoring of critically ill patients based on curve integral in this embodiment of the invention includes: comparing the curve integral with a preset curve integral; if the curve integral is greater than the preset curve integral, adjusting the threshold of the nighttime warning based on the difference between the curve integral and the preset curve integral; if the curve integral is less than or equal to the preset curve integral, adjusting the optimization period of the Bayesian network based on the number of abnormal ratios in multiple monitoring periods.

[0088] Specifically, in this embodiment, the monitoring time is 10 cycles, with five days as one cycle. The preset curve integral U0 = 40. The comparison process based on the curve integral U and the preset curve integral U0 is as follows:

[0089] If the curve integral U is greater than the preset curve integral U0, it indicates that the night warning is inaccurate. Then, the threshold of the night warning is adjusted based on the difference W between the curve integral and the preset curve integral.

[0090] If the curve integral U is less than or equal to the preset curve integral U0, it indicates that the optimization period of the network is relatively long. In this case, the optimization period of the Bayesian network is adjusted based on the number V of the anomaly ratios within multiple monitoring periods.

[0091] Specifically, the process of adjusting the threshold of the night warning based on the difference between the curve integral and the preset curve integral in this embodiment of the invention includes: reducing the threshold of the night warning based on the difference between the curve integral and the preset curve integral, and the difference is proportional to the reduction of the threshold of the night warning.

[0092] Specifically, in this embodiment, the preset difference W0 between the curve integral and the preset curve integral is 5. The comparison process based on the difference W and the preset difference W0 is as follows:

[0093] If the difference W is less than or equal to the preset difference W0, the threshold for nighttime warning will be adjusted to 0.89 times the original threshold.

[0094] If the difference W is greater than the preset difference W0, the threshold for nighttime warning will be adjusted to 0.75 times the original threshold.

[0095] Specifically, the process of adjusting the optimization period of the Bayesian network based on the number of anomaly ratios within multiple monitoring periods in this embodiment of the invention includes: reducing the optimization period of the Bayesian network based on the number of anomalies, and the number of anomalies being proportional to the reduction in the optimization period.

[0096] Specifically, in this embodiment, the preset number V0 of the anomaly ratio within multiple monitoring periods is 4. The comparison process based on the number V and the preset number V0 is as follows:

[0097] If the quantity V is less than or equal to the preset quantity V0, the optimization period will be adjusted to 0.86 times the original optimization period.

[0098] If the quantity V is greater than the preset quantity V0, the optimization period will be adjusted to 0.65 times the original optimization period.

[0099] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

[0100] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for networked monitoring of critically ill patients, characterized in that, include: Real-time monitoring of the patient's vital signs to obtain physiological data; Obtain the patient's basic information; The patient's physiological data is preprocessed based on Kalman filtering; The preprocessed physiological data and basic information are input into a Bayesian network to obtain the probability of the patient exhibiting abnormalities. Whether to issue a warning is determined based on the comparison result between the probability and the threshold. The network monitoring of critically ill patients is judged as qualified based on the average ratio of the number of warnings issued to the number of abnormalities in patients within multiple monitoring cycles. If the monitoring is not qualified, a corresponding processing method is generated based on the reason for the failure. The processing method includes adjusting the process noise covariance matrix in the Kalman filter, adjusting the optimization cycle of the Bayesian network, and adjusting the threshold of the night warning. Adjustments are made based on the corresponding processing method.

2. The network monitoring method for critically ill patients according to claim 1, characterized in that, The process of determining whether the online monitoring of critically ill patients is qualified based on the average ratio of the number of warnings issued to the number of abnormalities in patients over multiple monitoring periods includes: Calculate the average of the ratio of the number of warnings issued to the number of times the patient exhibited abnormalities over multiple monitoring periods; The online monitoring of critically ill patients is deemed qualified based on the comparison between the average value and the pre-stored preset average value. If the average value is greater than or equal to the first preset average value, the online monitoring of critically ill patients is deemed qualified. If the average value is less than the first preset average value and greater than the second preset average value, the network monitoring of critically ill patients is determined to be qualified based on the variance of the ratio of the number of warnings issued to the number of abnormalities in patients within multiple monitoring cycles. If the average value is less than or equal to the second preset average value, the network monitoring of critically ill patients is deemed unqualified. The reason for the unqualified network monitoring of critically ill patients is then determined based on the difference between the preset average value and the average value.

3. The network monitoring method for critically ill patients according to claim 2, characterized in that, The process of determining whether the online monitoring of critically ill patients is qualified based on the variance of the ratio of the number of warnings issued to the number of abnormalities in patients over multiple monitoring periods includes: Calculate the variance of the ratio of the number of warnings issued to the number of times a patient exhibits abnormalities over multiple monitoring periods; The variance is compared with the preset variance; If the variance is greater than the preset variance, the threshold is adjusted based on the difference between the variance and the preset variance; If the variance is less than or equal to the preset variance, the online monitoring of critically ill patients is deemed unqualified. The reason for the unqualified online monitoring of critically ill patients is then determined based on the difference between the preset average value and the average value.

4. The network monitoring method for critically ill patients according to claim 3, characterized in that, The process of adjusting the threshold based on the difference between the variance and the preset variance includes: The threshold is reduced based on the difference between the variance and the preset variance, and the reduction in the difference is proportional to the reduction in the threshold.

5. The network monitoring method for critically ill patients according to claim 2, characterized in that, The process of determining the reasons for the failure of online monitoring of critically ill patients based on the difference between the preset average and the average includes: Calculate the difference between the preset average and the average; The reasons for the failure of online monitoring of critically ill patients are determined based on the comparison results between the difference and the pre-stored preset difference. If the difference is greater than or equal to the first preset difference, the process noise covariance matrix in the Kalman filter is adjusted based on the ratio of the difference to the preset difference. If the difference is less than the first preset difference and greater than the second preset difference, then a curve of the difference between time and the number of abnormalities and the number of warnings is plotted, and the reason for the failure of the network monitoring of critically ill patients is determined based on the curve integral. If the difference is less than or equal to the second preset difference, count the number of abnormal ratios in multiple monitoring periods, and adjust the optimization period of the Bayesian network based on the number. The abnormality ratio is the ratio of the number of warnings issued to the number of abnormalities in a patient during each monitoring cycle to the preset abnormality ratio.

6. The network monitoring method for critically ill patients according to claim 5, characterized in that, The process of adjusting the process noise covariance matrix in Kalman filtering based on the ratio of the difference to a preset difference includes: The process noise covariance matrix in the Kalman filter is increased based on the ratio of the difference to the preset difference, and the ratio is proportional to the increase in the process noise covariance matrix.

7. The network monitoring method for critically ill patients according to claim 5, characterized in that, The process of determining the reasons for unqualified online monitoring of critically ill patients based on curve integrals includes: Compare the line integral with the preset line integral; If the curve integral is greater than the preset curve integral, the threshold for nighttime warning is adjusted based on the difference between the curve integral and the preset curve integral. If the curve integral is less than or equal to the preset curve integral, the optimization period of the Bayesian network is adjusted based on the number of anomaly ratios within multiple monitoring periods.

8. The network monitoring method for critically ill patients according to claim 7, characterized in that, The process of adjusting the nighttime warning threshold based on the difference between the curve integral and the preset curve integral includes: The threshold for nighttime warnings is lowered based on the difference between the curve integral and the preset curve integral, and the difference is proportional to the reduction in the threshold for nighttime warnings.

9. The network monitoring method for critically ill patients according to claim 5, characterized in that, The process of adjusting the optimization period of a Bayesian network based on the number of anomaly ratios across multiple monitoring periods includes: The optimization cycle of Bayesian networks is reduced by increasing the number of networks, and the reduction in the number of networks is proportional to the reduction in the optimization cycle.

10. A networked monitoring system for critically ill patients, characterized in that, include: The data monitoring unit is used to monitor the patient's vital signs in real time to obtain physiological data; The information acquisition unit is used to acquire basic information about the patient. A preprocessing unit, connected to the data monitoring unit, is used to preprocess the patient's physiological data based on Kalman filtering; A probability output unit, which is connected to the information acquisition unit and the preprocessing unit respectively, is used to input the preprocessed physiological data and the basic information into a Bayesian network to obtain the probability of the patient having an abnormality. An early warning unit, which is connected to the probability output unit, is used to determine whether to issue an early warning based on the comparison result between the probability and the threshold. An analysis unit, connected to the early warning unit, is used to determine whether the network monitoring of critically ill patients is qualified based on the average ratio of the number of early warnings issued to the number of abnormalities in patients within multiple monitoring cycles. If the monitoring is unqualified, a corresponding processing method is generated based on the reason for the unqualification. The processing method includes adjusting the process noise covariance matrix in the Kalman filter, adjusting the optimization period of the Bayesian network, and adjusting the threshold of the nighttime early warning. A control unit, connected to the analysis unit, is used to adjust based on the corresponding processing method.

Images