Doctor recommendation method and device based on data collected by electrocardiograph monitor

Abstract

The invention discloses a doctor recommendation method and device based on data collected by an electrocardiograph monitor, and particularly relates to the technical field of doctor recommendation, which comprises the following steps: determining data quality information and data delay information through waveform analysis and spectrum analysis of ECG signal data and the time delay condition of uploading the ECG signal data to a medical cloud platform, the method comprises the following steps: generating the incredibility of current ECG signal data, switching a response strategy and a cloud collaboration mechanism of an ECG monitor, generating a complexity score of the ECG signal data from the perspective of ECG feature complexity, obtaining a specific disease of a patient based on a historical medical record of the patient, and combining a historical healing record of a doctor to obtain a medical record of the ECG signal data. According to the method, the medical cloud platform determines the cure performance data of different doctors, comprehensively analyzes the cure performance data of the different doctors, determines the recommendation sequence of the doctors based on the current on-duty doctor, and helps to intelligently distribute the doctors and intelligently recommend the most suitable doctor by the medical cloud platform, thereby improving the diagnosis accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of doctor recommendation, and more specifically, to a doctor recommendation method and device based on data collected by an electrocardiogram monitor. Background Art

[0002] Electrocardiogram (ECG) monitoring technology plays a vital role in modern medicine and is widely used for early screening, real-time monitoring and remote diagnosis of cardiovascular diseases. With the development of telemedicine and medical cloud platforms, the transmission, storage and analysis of ECG data are gradually moving towards cloud-based and intelligent development. The current ECG data analysis mainly relies on a single data quality assessment method without fully considering the delay and overall accuracy of data transmission. At the same time, in terms of doctor recommendations, the existing system usually only matches doctors based on their professional direction, lacking a comprehensive assessment of doctors' actual healing ability and ECG data processing experience, resulting in patients not getting the best medical advice.

[0003] In order to solve the above defects, a technical solution is now provided. Summary of the invention

[0004] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a doctor recommendation method and device based on data collected by an electrocardiogram monitor to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A doctor recommendation method based on data collected by an electrocardiogram monitor comprises the following steps:

[0007] S1: Determine the data quality information of the current ECG signal data by waveform analysis and spectrum analysis of the ECG signal data, and determine the data delay information of the current ECG signal data by the time delay of uploading the ECG signal data to the medical cloud platform;

[0008] S2: Determine the unreliability of the current ECG signal data through data quality information and data delay information, and switch the response strategy of the ECG monitor and the cloud collaboration mechanism;

[0009] S3: Generate a complexity score for ECG signal data from the perspective of ECG feature complexity, and obtain the patient's specific disease based on the patient's historical medical history. Combined with the doctor's historical cure records, determine the cure performance data of different doctors.

[0010] S4: By comprehensively analyzing the healing performance data of different doctors, the recommended order of doctors is determined based on the current doctors on duty.

[0011] In a preferred embodiment, determining data quality information of current ECG signal data includes:

[0012] The data quality information of the current ECG signal data is determined by the waveform continuity abnormality coefficient and the spectrum energy proportion abnormality coefficient;

[0013] The acquisition logic of the waveform continuity abnormality coefficient is as follows: obtain the QRS wave in the ECG signal data within the monitoring interval, determine the time interval between adjacent R waves, set the time interval threshold, and when the time interval between two adjacent R waves exceeds the time interval threshold, it is regarded as a QRS wave missing;

[0014] The time intervals between adjacent R waves in the monitoring interval are obtained, and the time intervals between adjacent R waves in the monitoring interval are marked as: , where i=1, 2, 3, ..., I, I is a positive integer, i is the number of the adjacent R wave in the monitoring interval, and the time interval threshold is marked as: ;

[0015] The adjacent R waves whose time intervals between adjacent R waves in the monitoring interval exceed the time interval threshold are obtained, and the time intervals of adjacent R waves exceeding the time interval threshold are re-marked as: , where n=1, 2, 3, ..., N, and N is a positive integer;

[0016] Calculate the waveform continuity anomaly coefficient, the calculation formula is: ;in, is the waveform continuity anomaly coefficient.

[0017] In a preferred embodiment, the logic for obtaining the spectrum energy proportion abnormality coefficient is:

[0018] Perform fast Fourier transform on the ECG signal in the monitoring interval to obtain the energy distribution of each frequency component, divide the ECG signal into high-frequency band and low-frequency band, and determine the abnormal coefficient of spectrum energy ratio by energy ratio calculation;

[0019] The calculation formula of spectrum energy ratio abnormal coefficient is: ;in, is the spectrum energy ratio abnormal coefficient, , It is the lowest frequency standard in the low frequency band. It is the lowest frequency standard in the high frequency band. It is the highest frequency standard in the high frequency band of ECG signals. is the energy of the ECG signal at frequency f.

[0020] In a preferred embodiment, the data delay information of the current ECG signal data includes:

[0021] The data delay information of the current ECG signal data is represented by a delay volatility coefficient;

[0022] The acquisition logic of the delay volatility coefficient is: obtain the set of ECG signal data transmission delays in the monitoring interval, and mark the set of ECG signal data transmission delays in the monitoring interval as: { , , ,……, ,……, }; where m is the number of ECG signal data packets sampled continuously within the monitoring interval, and M is the number of ECG signal data transmission delays within the monitoring interval;

[0023] Calculate the average value of the ECG signal data packet delay using the following formula: ;in, is the average value of ECG signal data packet delay;

[0024] Calculate the standard deviation of the ECG signal data packet delay using the following formula: ;in, is the standard deviation of the ECG signal data packet delay;

[0025] The delay fluctuation coefficient of the ECG signal data packet delay is calculated using the following formula: ;in, is the delay volatility coefficient.

[0026] In a preferred embodiment, determining the unreliability of the current ECG signal data by using the data quality information and the data delay information includes:

[0027] The waveform continuity anomaly coefficient, spectrum energy ratio anomaly coefficient and delay volatility coefficient are weighted and summed to construct an untrustworthy assessment model and generate an untrustworthy assessment coefficient. The calculation formula of the untrustworthy assessment coefficient is: ;in, is the unreliable evaluation coefficient, is the proportional coefficient of the waveform continuous abnormal coefficient, is the proportional coefficient of the spectrum energy to the abnormal coefficient, is the proportionality coefficient of the delay fluctuation coefficient, , , Both are greater than 0.

[0028] In a preferred embodiment, the response strategy of the ECG monitor and the cloud collaboration mechanism include:

[0029] Setting a first threshold value of an untrustworthy evaluation coefficient and a second threshold value of an untrustworthy evaluation coefficient, wherein the first threshold value of the untrustworthy evaluation coefficient is greater than the second threshold value of the untrustworthy evaluation coefficient, and comparing the untrustworthy evaluation coefficient of the data collected by the electrocardiogram monitor of the current patient with the first threshold value of the untrustworthy evaluation coefficient and the second threshold value of the untrustworthy evaluation coefficient;

[0030] If the untrustworthy evaluation coefficient is greater than the first threshold of the untrustworthy evaluation coefficient, a first warning signal is generated, the ECG monitor starts the backup lead and switches to the original waveform storage, and in the cloud-edge collaborative mechanism, for the ECG monitor based on the patient's alarm, the remote diagnosis is suspended and the local alarm is prioritized;

[0031] If the untrustworthy evaluation coefficient is greater than the second threshold of the untrustworthy evaluation coefficient, and the untrustworthy evaluation coefficient is less than the first threshold, a second warning signal is generated, the ECG monitor starts the enhanced filtering algorithm and compresses the data for transmission, and the cloud-edge collaborative mechanism limits the upload of non-critical data;

[0032] If the untrustworthy assessment coefficient is less than the second threshold value of the untrustworthy assessment coefficient, no warning signal is generated, the ECG monitor runs in full-function mode, and is synchronized to the medical cloud platform in real time.

[0033] In a preferred embodiment, determining the healing performance data of different doctors includes:

[0034] The doctor's healing performance data is represented by the complexity healing coefficient and the specific disease healing coefficient;

[0035] The acquisition logic of the complexity cure coefficient is as follows: collecting ECG signal data within the monitoring interval, extracting key ECG features in the ECG signal data, standardizing the key ECG features and determining the deviation value of the key ECG features according to the standard value, collecting the patient's medical information, and scoring the risk impact of the key ECG features on the patient based on the patient's medical information to obtain the patient's impact score;

[0036] The deviation value of the key ECG features in the ECG signal data within the monitoring interval and the patient's impact score are taken as input, and the complexity evaluation coefficient is taken as output through the logistic regression model. The calculation formula of the complexity evaluation coefficient is:

[0037] ; Among them, FZD is the complexity evaluation coefficient, , , ,……, is the weight, , , ,……, Rate the patient's impact, , , ,……, is the deviation value of the key ECG feature.

[0038] Based on different historical ECG signal data, the complexity evaluation coefficients of different ECG signal data are divided into different intervals. The performance of different doctors in each complexity evaluation coefficient interval is statistically analyzed to obtain the cure rates of different doctors in different complexity evaluation coefficient intervals. The cure rates of different doctors in different complexity evaluation coefficient intervals are marked as: , where g=1, 2, 3, ..., G, G is a positive integer, and g is the number of different complexity evaluation coefficient intervals;

[0039] According to the cure rate of doctors in different complexity evaluation coefficient intervals and the overall average cure rate in different complexity evaluation coefficient intervals, the complexity cure coefficient is calculated. The calculation formula is: ;in, is the complexity cure coefficient, QJ is the cure rate of the doctor in the complexity evaluation coefficient interval where the current ECG signal data is located, is the average cure rate of all doctors in the complexity assessment coefficient interval where the previous ECG signal data is located.

[0040] In a preferred embodiment, the logic for obtaining the specific disease cure coefficient is:

[0041] According to the patient's medical history, the specific disease of the patient is initially obtained. According to the cure history records of different doctors for the specific disease, the cure rate of different doctors for the specific disease is obtained. The specific disease cure coefficient is calculated by the arithmetic mean. The calculation formula is: ;in, is the specific disease cure coefficient, h=1, 2, 3, ..., H, H is a positive integer, h is the specific disease of the patient, The doctor's cure rate for a specific disease.

[0042] In a preferred embodiment, determining the order of doctors' recommendations includes:

[0043] The weighted sum of the complexity cure coefficient and the specific disease cure coefficient is calculated to build a recommendation evaluation model and generate a recommendation evaluation coefficient. The calculation formula of the recommendation evaluation coefficient is: ;in, is the recommended evaluation coefficient, is the proportionality coefficient of the complexity healing coefficient, is the proportional coefficient of the cure coefficient for a specific disease, , They are both greater than 0;

[0044] When the ECG signal data uploaded by the patient to the medical cloud platform does not generate a warning signal, the recommended evaluation coefficients of all the doctors on duty are obtained according to the current doctor on duty, and the recommended evaluation coefficients of all the doctors on duty are sorted, and the doctors on duty with large recommended evaluation coefficients are recommended to the patient first.

[0045] In a preferred embodiment, a doctor recommendation device based on data collected by an electrocardiogram monitor includes a data quality module, a data delay module, an untrustworthy evaluation module, a healing performance data module, and a recommendation evaluation module;

[0046] A data quality module is used to collect ECG signal data, perform waveform analysis and spectrum analysis on the ECG signal data, and calculate data quality information;

[0047] Data delay module, used to monitor the delay of ECG signal data uploaded to the medical cloud platform and calculate data delay information;

[0048] An untrustworthy evaluation module is used to calculate the untrustworthy evaluation coefficient of ECG signal data according to data quality information and data delay information, and trigger the response strategy of the ECG monitor and the cloud collaboration mechanism based on the untrustworthy evaluation coefficient;

[0049] The cure performance data module is used to extract the key ECG features of the ECG signal and calculate the complex cure coefficient and the specific disease cure coefficient in combination with the doctor's cure rate records;

[0050] The recommendation evaluation module is used to calculate the recommendation evaluation coefficient of the current doctor on duty, sort the doctors on duty according to the recommendation evaluation coefficient, and give priority to recommending the doctors with the highest scores.

[0051] Technical effects and advantages of the present invention:

[0052] The present invention analyzes the real-time and accuracy of data collected by the ECG monitor, switches the response strategy of the ECG monitor and the cloud collaboration mechanism, and analyzes the doctor selection recommended by the medical cloud platform under remote treatment through the doctor's healing performance data. The present invention is suitable for patients to upload ECG data at home through portable ECG monitoring devices, and the medical cloud platform intelligently assigns doctors and intelligently recommends the most suitable doctors based on the doctor's historical diagnostic performance, thereby improving diagnostic accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to facilitate understanding by those skilled in the art, the present invention is further described below in conjunction with the accompanying drawings;

[0054] Figure 1 A schematic flow chart of a doctor recommendation method based on data collected by an electrocardiogram monitor according to the present invention;

[0055] Figure 2 The present invention is a schematic structural diagram of a doctor recommendation device based on data collected by an electrocardiogram monitor. DETAILED DESCRIPTION

[0056] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0057] Example 1

[0058] Figure 1 The present invention is a flowchart of a doctor recommendation method based on data collected by an electrocardiogram monitor, which specifically includes the following steps:

[0059] S1: Determine the data quality information of the current ECG signal data by waveform analysis and spectrum analysis of the ECG signal data, and determine the data delay information of the current ECG signal data by the time delay of uploading the ECG signal data to the medical cloud platform;

[0060] S2: Determine the unreliability of the current ECG signal data through data quality information and data delay information, and switch the response strategy of the ECG monitor and the cloud collaboration mechanism;

[0061] S3: Generate a complexity score for ECG signal data from the perspective of ECG feature complexity, and obtain the patient's specific disease based on the patient's historical medical history. Combined with the doctor's historical cure records, determine the cure performance data of different doctors.

[0062] S4: By comprehensively analyzing the healing performance data of different doctors, the recommended order of doctors is determined based on the current doctors on duty.

[0063] Patients carry portable ECG monitors to achieve remote monitoring of patients in different scenarios. By analyzing the data from the ECG monitor, the system automatically recommends suitable doctors to patients. During the recommendation process, the ECG signal data collected by the ECG monitor is usually time series, and in home monitoring, patients may move a lot, resulting in large ECG signal noise. The system needs to identify this situation. At the same time, when the network is unstable, the system may prioritize the transmission of key data fragments instead of continuous streaming to save bandwidth and ensure that important information is not lost. Therefore, the system first needs to determine the response strategy of the ECG monitor and the cloud collaboration mechanism to ensure that the ECG signal data uploaded to the cloud can be used by doctors.

[0064] In one embodiment, when data collected by an ECG monitor detects a serious problem, the system needs to respond immediately. However, due to network problems, remote diagnosis cannot be relied upon and local processing must be turned to. Therefore, the system needs to determine whether to continue remote monitoring or turn to local processing.

[0065] Through waveform analysis and spectrum analysis of the ECG signal data, the data quality information of the current ECG signal data is determined, and the data quality information of the current ECG signal data is determined by the waveform continuity abnormality coefficient and the spectrum energy ratio abnormality coefficient. Through the delay of uploading the ECG signal data to the medical cloud platform, the data delay information of the current ECG signal data is determined, and the data delay information of the current ECG signal data is determined by the delay volatility coefficient.

[0066] Among them, the waveform continuity abnormality coefficient is calculated by analyzing the QRS wave in the ECG signal. Under normal circumstances, the QRS wave should appear with a certain regularity. For example, when the heart rate is 60-100 times / minute, the interval between each QRS wave is about 0.6 to 1 second. If there should be about 5 to 8 QRS waves within 5 seconds, if less than this number is detected, it may be missing. In addition, the noise and interference of the ECG monitor itself, especially in the scenario of home monitoring, the missing QRS wave may be caused by motion artifacts or other noise, rather than the real missing physiological signal.

[0067] The acquisition logic of the waveform continuity abnormality coefficient is as follows: obtain the QRS wave in the ECG signal data within the monitoring interval, determine the time interval between adjacent R waves, set the time interval threshold, and when the time interval between two adjacent R waves exceeds the time interval threshold, it is regarded as a QRS wave missing;

[0068] It should be noted that the monitoring interval is a specific time period, usually 5 seconds. The R wave is a major peak in the QRS complex. Under normal circumstances, there is a certain regularity between adjacent R waves. The time interval threshold is used to evaluate the time length between adjacent R waves. When the time length between adjacent R waves is greater than the time interval threshold, it means that the R wave is missing, that is, the QRS wave fails to appear normally. The monitoring interval and time interval threshold are determined by professional staff.

[0069] The time intervals between adjacent R waves in the monitoring interval are obtained, and the time intervals between adjacent R waves in the monitoring interval are marked as: , where i=1, 2, 3, ..., I, I is a positive integer, i is the number of the adjacent R wave in the monitoring interval, and the time interval threshold is marked as: ;

[0070] The adjacent R waves whose time intervals between adjacent R waves in the monitoring interval exceed the time interval threshold are obtained, and the time intervals of adjacent R waves exceeding the time interval threshold are re-marked as: , where n=1, 2, 3, ..., N, and N is a positive integer;

[0071] Calculate the waveform continuity anomaly coefficient, the calculation formula is: ;in, is the waveform continuity anomaly coefficient.

[0072] It can be seen from the formula that the larger the waveform continuity abnormality coefficient is, the more times the QRS wave is missing in the monitoring interval, and the longer the time interval between adjacent R waves. In this case, possible reasons include the failure of the heart's sinoatrial node to generate impulses on time, the failure of electrical signals to be transmitted from the atrium to the ventricle, and the ventricle itself generating impulses but being misjudged as missing.

[0073] Among them, the spectrum energy ratio abnormal coefficient refers to the proportion of the total energy occupied by the high-frequency part in the ECG signal spectrum. For time domain signals, the signal can be converted from the time domain to the frequency domain through Fourier transform, so as to observe the energy distribution of different frequency components. The high-frequency energy ratio is the ratio of the energy in a specific high-frequency band to the energy of the entire signal. The main useful signal components of normal physiological signals (whether ECG or ECG) are usually concentrated in a lower frequency range. For example, the key waveforms of ECG signals (such as P waves, QRS complex waves, and T waves) are usually in a relatively low frequency band. When the high-frequency energy ratio is low, it means that the main components of the signal are still concentrated in the expected frequency range, and the signal quality is high. If the high-frequency energy ratio increases abnormally, it may mean that more high-frequency noise or interference (such as electromyographic interference, electromagnetic interference, or distortion during transmission) is mixed in the signal, thereby affecting the correct interpretation of the signal.

[0074] The acquisition logic of the spectrum energy ratio abnormal coefficient is as follows: perform fast Fourier transform on the ECG signal in the monitoring interval, obtain the energy distribution of each frequency component, divide the ECG signal into high frequency band and low frequency band, and determine the spectrum energy ratio abnormal coefficient by energy ratio calculation;

[0075] The calculation formula of spectrum energy ratio abnormal coefficient is: ;in, is the spectrum energy ratio abnormal coefficient, , It is the lowest frequency standard in the low frequency band. It is the lowest frequency standard in the high frequency band. It is the highest frequency standard in the high frequency band of ECG signals. is the energy of the ECG signal at frequency f.

[0076] It should be noted that the lowest frequency standard of the low-frequency band is the starting frequency when the overall energy of the signal is counted (for example, 0.5Hz can be taken or a value can be set based on actual conditions), the lowest frequency standard of the high-frequency band is divided, and the components below this frequency are classified as low-frequency bands, and the highest frequency standard of the high-frequency band of the ECG signal is the upper limit when counting energy (for example, it can be set to 40Hz, 50Hz or higher, depending on the specific application scenario and device characteristics).

[0077] It can be seen from the formula that the larger the abnormal coefficient of spectrum energy ratio, the higher the proportion of high-frequency energy in the ECG signal in the total signal energy, and the higher the proportion of high-frequency components in the signal. In this case, possible reasons include noise interference at the acquisition end or during transmission, such as enhanced high-frequency noise (such as electromyographic interference, electromagnetic interference); improper signal compression or filtering, such as failure to effectively suppress high-frequency noise during preprocessing; equipment failure or transmission distortion, such as causing abnormal changes in signal energy distribution.

[0078] Among them, the delay fluctuation coefficient refers to the time interval from the ECG signal acquisition device uploading the data to the medical cloud platform to the data being successfully received on the cloud platform. The delay fluctuation coefficient reflects the impact of comprehensive factors such as transmission links, network congestion, and intermediate processing links.

[0079] The acquisition logic of the delay volatility coefficient is: obtain the set of ECG signal data transmission delays in the monitoring interval, and mark the set of ECG signal data transmission delays in the monitoring interval as: { , , ,……, ,……, }; where m is the number of ECG signal data packets sampled continuously within the monitoring interval, and M is the number of ECG signal data transmission delays within the monitoring interval;

[0080] Calculate the average value of the ECG signal data packet delay using the following formula: ;in, is the average value of ECG signal data packet delay;

[0081] Calculate the standard deviation of the ECG signal data packet delay using the following formula: ;in, is the standard deviation of the ECG signal data packet delay;

[0082] It should be noted that the standard deviation of the ECG signal data packet delay indicates the fluctuation or change of the delay time. The system may have an average transmission delay, but the delay between each data packet may not be stable. The greater the delay volatility, the worse the real-time and stability of the system transmission.

[0083] The delay fluctuation coefficient of the ECG signal data packet delay is calculated using the following formula: ;in, is the delay volatility coefficient.

[0084] It can be seen from the formula that the delay volatility coefficient represents the relative volatility coefficient. The larger the delay volatility coefficient is, the more likely it is that the delay between data packets is not stable, indicating that the data collected by the ECG monitor is unstable and may be discontinuous or unreliable after being transmitted to the cloud platform.

[0085] The waveform continuity anomaly coefficient, spectrum energy ratio anomaly coefficient and delay volatility coefficient are weighted and summed to construct an untrustworthy assessment model and generate an untrustworthy assessment coefficient. The calculation formula of the untrustworthy assessment coefficient is: ;in, is the unreliable evaluation coefficient, is the proportional coefficient of the waveform continuous abnormal coefficient, is the proportional coefficient of the spectrum energy to the abnormal coefficient, is the proportionality coefficient of the delay fluctuation coefficient, , , Both are greater than 0.

[0086] It can be seen from the formula that the larger the waveform continuity abnormality coefficient, the spectrum energy ratio abnormality coefficient and the delay fluctuation coefficient are, the larger the unreliable assessment coefficient is, indicating that the reliability of the data collected by the current patient's ECG monitor is poor and may not be used as real-time patient medical data. On the contrary, the smaller the waveform continuity abnormality coefficient, the spectrum energy ratio abnormality coefficient and the delay fluctuation coefficient are, the smaller the unreliable assessment coefficient is, indicating that the reliability of the data collected by the current patient's ECG monitor is good and may be used as real-time patient medical data.

[0087] Setting a first threshold value of an untrustworthy evaluation coefficient and a second threshold value of an untrustworthy evaluation coefficient, wherein the first threshold value of the untrustworthy evaluation coefficient is greater than the second threshold value of the untrustworthy evaluation coefficient, and comparing the untrustworthy evaluation coefficient of the data collected by the electrocardiogram monitor of the current patient with the first threshold value of the untrustworthy evaluation coefficient and the second threshold value of the untrustworthy evaluation coefficient;

[0088] If the untrustworthy evaluation coefficient is greater than the first threshold of the untrustworthy evaluation coefficient, a first warning signal is generated, the ECG monitor starts the backup lead and switches to the original waveform storage, and in the cloud-edge collaborative mechanism, for the ECG monitor based on the patient's alarm, the remote diagnosis is suspended and the local alarm is prioritized;

[0089] If the untrustworthy evaluation coefficient is greater than the second threshold of the untrustworthy evaluation coefficient, and the untrustworthy evaluation coefficient is less than the first threshold, a second warning signal is generated, the ECG monitor starts the enhanced filtering algorithm and compresses the data for transmission, and the cloud-edge collaborative mechanism limits the upload of non-critical data;

[0090] If the untrustworthy assessment coefficient is less than the second threshold value of the untrustworthy assessment coefficient, no warning signal is generated, the ECG monitor runs in full-function mode, and is synchronized to the medical cloud platform in real time.

[0091] Example 2

[0092] The above embodiment helps to ensure that patients receive timely treatment under early warning signals by focusing on the real-time and accuracy of data collected by the ECG monitor. This embodiment analyzes the doctors recommended by the medical cloud platform under remote treatment through the doctors' healing performance data.

[0093] In one embodiment, the doctor's healing performance data is represented by a complexity healing coefficient and a specific disease healing coefficient, wherein the complexity healing coefficient extracts key ECG features from the collected ECG data, compares them with standard or ideal values, calculates deviation values, reflects the severity or abnormality of the disease, combines the impact score of key ECG features on the patient, trains through a logistic regression model, quantifies the complexity of collecting ECG data, and obtains the doctor's preference for processing ECG data of different types of complexity based on the doctor's records of processing ECG data of different complexities.

[0094] It should be noted that it is generally believed that the complexity of ECG data is inversely proportional to the cure rate. Therefore, when using the linear regression model for training, the deviation value of the key ECG features extracted from the current ECG data and the impact score of the key ECG features on the patient are used as input, and the cure rate is used as output, which can quantify the complexity of the current ECG signal data.

[0095] Secondly, doctors have different preferences at different levels of data complexity. For example, doctors who perform well in high-complexity ECG signal data may be more suitable for dealing with difficult cases. By comparing the actual cure rate and the average cure rate of each doctor at different complexity levels, the advantages and disadvantages of each doctor in the treatment of patients with specific complexity can be revealed. This method can quantify the complexity of ECG data collection and recommend the most suitable diagnosis and treatment expert with the help of the doctor's historical treatment records, which will help improve the rationality of medical resource allocation and clinical treatment effectiveness.

[0096] The acquisition logic of the complexity cure coefficient is as follows: collecting ECG signal data within the monitoring interval, extracting key ECG features in the ECG signal data, standardizing the key ECG features and determining the deviation value of the key ECG features according to the standard value, collecting the patient's medical information, and scoring the risk impact of the key ECG features on the patient based on the patient's medical information to obtain the patient's impact score;

[0097] It should be noted that key ECG characteristics include QRS wave width, ST segment deviation, heart rate variability, and QT interval prolongation, and the patient's medical information includes age, medical history, complications, etc.

[0098] The deviation value of the key ECG features in the ECG signal data within the monitoring interval and the patient's impact score are taken as input, and the complexity evaluation coefficient is taken as output through the logistic regression model. The calculation formula of the complexity evaluation coefficient is:

[0099] ; Among them, FZD is the complexity evaluation coefficient, , , ,……, is the weight, , , ,……, Rate the patient's impact, , , ,……, is the deviation value of the key ECG feature.

[0100] It should be noted that when training the logistic regression model, the complexity assessment coefficient can be trained by the inverse representation of the cure rate. The final cure rate is trained based on the actual different ECG signal data. That is, when the cure rate is high, it means that the patient's ECG data is close to the ideal standard, the deviation value of the key ECG characteristics is small, and the patient's impact score is small, that is, the data complexity is relatively low.

[0101] Based on different historical ECG signal data, the complexity evaluation coefficients of different ECG signal data are divided into different intervals. The performance of different doctors in each complexity evaluation coefficient interval is statistically analyzed to obtain the cure rates of different doctors in different complexity evaluation coefficient intervals. The cure rates of different doctors in different complexity evaluation coefficient intervals are marked as: , where g=1, 2, 3, ..., G, G is a positive integer, and g is the number of different complexity evaluation coefficient intervals;

[0102] According to the cure rate of doctors in different complexity evaluation coefficient intervals and the overall average cure rate in different complexity evaluation coefficient intervals, the complexity cure coefficient is calculated. The calculation formula is: ;in, is the complexity cure coefficient, QJ is the cure rate of the doctor in the complexity evaluation coefficient interval where the current ECG signal data is located, is the average cure rate of all doctors in the complexity assessment coefficient interval where the previous ECG signal data is located.

[0103] It can be seen from the formula that the larger the complexity cure coefficient is, the more suitable the doctor may be for processing ECG signal data of the current complexity, which means that the doctor's success rate in processing the current ECG signal data may be higher. Therefore, it may be more recommended that the doctor analyze the current ECG signal data and treat the patient.

[0104] Among them, the specific disease cure coefficient is based on the patient's previous medical history, including previous disease records, treatment plans, surgical history, drug history and other information, and uses the ECG monitor to extract key ECG features from the ECG data. Combined with the built-in automatic diagnosis algorithm, it preliminarily determines the possible diseases of the patient and preliminarily obtains the specific diseases that the patient may have. Based on the cure of specific diseases by different doctors, it reflects the relative treatment level of doctors.

[0105] The logic for obtaining the specific disease cure coefficient is as follows: based on the patient's past medical records, the specific disease that the patient has is initially obtained; based on the cure history records of different doctors for the specific disease, the cure rates of different doctors for the specific disease are obtained; and the specific disease cure coefficient is calculated by the arithmetic mean. The calculation formula is: ;in, is the specific disease cure coefficient, h=1, 2, 3, ..., H, H is a positive integer, h is the specific disease of the patient, The doctor's cure rate for a specific disease.

[0106] From the formula, we can see that the larger the specific disease cure coefficient is, the stronger the doctor's overall ability to cure the patient's specific disease is; conversely, a lower coefficient indicates a relatively poor treatment effect. The specific disease cure coefficient can serve as an important quantitative basis for doctor performance evaluation, doctor recommendation and clinical decision-making.

[0107] The weighted sum of the complexity cure coefficient and the specific disease cure coefficient is calculated to build a recommendation evaluation model and generate a recommendation evaluation coefficient. The calculation formula of the recommendation evaluation coefficient is: ;in, is the recommended evaluation coefficient, is the proportionality coefficient of the complexity healing coefficient, is the proportional coefficient of the cure coefficient for a specific disease, , Both are greater than 0.

[0108] It can be seen from the formula that the larger the complexity cure coefficient and the specific disease cure coefficient, the larger the recommended evaluation coefficient, which means that the doctor may be able to provide more effective help for the treatment of the current patient and may be more suitable for analyzing the patient's ECG signal data. Conversely, the smaller the complexity cure coefficient and the specific disease cure coefficient, the smaller the recommended evaluation coefficient, which means that the doctor may not be able to provide more effective help for the treatment of the current patient and may not be suitable for analyzing the patient's ECG signal data.

[0109] When the ECG signal data uploaded by the patient to the medical cloud platform does not generate a warning signal, the recommended evaluation coefficients of all the doctors on duty are obtained according to the current doctor on duty, and the recommended evaluation coefficients of all the doctors on duty are sorted. The doctor on duty with a larger recommended evaluation coefficient is recommended to the patient for further analysis of the patient's ECG signal data.

[0110] The present invention analyzes the real-time and accuracy of data collected by the ECG monitor, switches the response strategy of the ECG monitor and the cloud collaboration mechanism, and analyzes the doctor selection recommended by the medical cloud platform under remote treatment through the doctor's healing performance data. The present invention is suitable for patients to upload ECG data at home through portable ECG monitoring devices, and the medical cloud platform intelligently assigns doctors and intelligently recommends the most suitable doctors based on the doctor's historical diagnostic performance, thereby improving diagnostic accuracy.

[0111] Example 3

[0112] A doctor recommendation device based on data collected by an electrocardiogram monitor, specifically comprising a data quality module, a data delay module, an untrustworthy evaluation module, a cure performance data module and a recommendation evaluation module;

[0113] A data quality module is used to collect ECG signal data, perform waveform analysis and spectrum analysis on the ECG signal data, and calculate data quality information;

[0114] Data delay module, used to monitor the delay of ECG signal data uploaded to the medical cloud platform and calculate data delay information;

[0115] An untrustworthy evaluation module is used to calculate the untrustworthy evaluation coefficient of ECG signal data according to data quality information and data delay information, and trigger the response strategy of the ECG monitor and the cloud collaboration mechanism based on the untrustworthy evaluation coefficient;

[0116] The cure performance data module is used to extract the key ECG features of the ECG signal and calculate the complex cure coefficient and the specific disease cure coefficient in combination with the doctor's cure rate records;

[0117] The recommendation evaluation module is used to calculate the recommendation evaluation coefficient of the current doctor on duty, sort the doctors on duty according to the recommendation evaluation coefficient, and give priority to recommending the doctors with the highest scores.

[0118] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0119] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0120] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0121] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0122] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0123] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0124] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks or optical disks.

[0125] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A doctor recommendation method based on data collected by an electrocardiogram monitor, characterized in that: The following steps are involved: S1: Determine the data quality information of the current ECG signal data by waveform analysis and spectrum analysis of the ECG signal data, and determine the data delay information of the current ECG signal data by the time delay of uploading the ECG signal data to the medical cloud platform; S2: Determine the unreliability of the current ECG signal data through data quality information and data delay information, and switch the response strategy of the ECG monitor and the cloud collaboration mechanism; S3: Generate a complexity score for ECG signal data from the perspective of ECG feature complexity, and obtain the patient's specific disease based on the patient's historical medical history. Combined with the doctor's historical cure records, determine the cure performance data of different doctors. S4: By comprehensively analyzing the healing performance data of different doctors, the recommended order of doctors is determined based on the current doctors on duty.

2. A doctor recommendation method based on data collected by an electrocardiogram monitor according to claim 1, characterized in that: Determine the data quality information of the current ECG signal data, including: The data quality information of the current ECG signal data is determined by the waveform continuity abnormality coefficient and the spectrum energy proportion abnormality coefficient; The acquisition logic of the waveform continuity abnormality coefficient is as follows: obtain the QRS wave in the ECG signal data within the monitoring interval, determine the time interval between adjacent R waves, set the time interval threshold, and when the time interval between two adjacent R waves exceeds the time interval threshold, it is regarded as a QRS wave missing; The time intervals between adjacent R waves in the monitoring interval are obtained, and the time intervals between adjacent R waves in the monitoring interval are marked as: , where i=1, 2, 3, ..., I, I is a positive integer, i is the number of the adjacent R wave in the monitoring interval, and the time interval threshold is marked as: ; The adjacent R waves whose time intervals between adjacent R waves in the monitoring interval exceed the time interval threshold are obtained, and the time intervals of adjacent R waves exceeding the time interval threshold are re-marked as: , where n=1, 2, 3, ..., N, and N is a positive integer; Calculate the waveform continuity anomaly coefficient, the calculation formula is: ;in, is the waveform continuity anomaly coefficient.

3. The method for recommending doctors based on data collected by an electrocardiogram monitor according to claim 2, characterized in that: The logic for obtaining the spectrum energy ratio abnormality coefficient is as follows: Perform fast Fourier transform on the ECG signal in the monitoring interval to obtain the energy distribution of each frequency component, divide the ECG signal into high-frequency band and low-frequency band, and determine the abnormal coefficient of spectrum energy ratio by energy ratio calculation; The calculation formula of spectrum energy ratio abnormal coefficient is: ;in, is the spectrum energy ratio abnormal coefficient, , It is the lowest frequency standard in the low frequency band. It is the lowest frequency standard in the high frequency band. It is the highest frequency standard in the high frequency band of ECG signals. is the energy of the ECG signal at frequency f.

4. The method for recommending doctors based on data collected by an electrocardiogram monitor according to claim 3, characterized in that: Data delay information of the current ECG signal data, including: The data delay information of the current ECG signal data is represented by a delay volatility coefficient; The acquisition logic of the delay volatility coefficient is: obtain the set of ECG signal data transmission delays in the monitoring interval, and mark the set of ECG signal data transmission delays in the monitoring interval as: { , , ,……, ,……, }; where m is the number of ECG signal data packets sampled continuously within the monitoring interval, and M is the number of ECG signal data transmission delays within the monitoring interval; Calculate the average value of the ECG signal data packet delay using the following formula: ;in, is the average value of ECG signal data packet delay; Calculate the standard deviation of the ECG signal data packet delay using the following formula: ;in, is the standard deviation of the ECG signal data packet delay; The delay fluctuation coefficient of the ECG signal data packet delay is calculated using the following formula: ;in, is the delay volatility coefficient.

5. The method for recommending doctors based on data collected by an electrocardiogram monitor according to claim 4, characterized in that: The unreliability of the current ECG signal data is determined through data quality information and data delay information, including: The waveform continuity anomaly coefficient, spectrum energy ratio anomaly coefficient and delay volatility coefficient are weighted and summed to construct an untrustworthy assessment model and generate an untrustworthy assessment coefficient. The calculation formula of the untrustworthy assessment coefficient is: ;in, is the unreliable assessment coefficient, is the proportional coefficient of the waveform continuous abnormal coefficient, is the proportional coefficient of the spectrum energy to the abnormal coefficient, is the proportionality coefficient of the delay fluctuation coefficient, , , Both are greater than 0.

6. The method for recommending doctors based on data collected by an electrocardiogram monitor according to claim 5, characterized in that: The response strategy and cloud collaboration mechanism for replacing ECG monitors include: Setting a first threshold value of an untrustworthy evaluation coefficient and a second threshold value of an untrustworthy evaluation coefficient, wherein the first threshold value of the untrustworthy evaluation coefficient is greater than the second threshold value of the untrustworthy evaluation coefficient, and comparing the untrustworthy evaluation coefficient of the data collected by the electrocardiogram monitor of the current patient with the first threshold value of the untrustworthy evaluation coefficient and the second threshold value of the untrustworthy evaluation coefficient; If the untrustworthy evaluation coefficient is greater than the first threshold of the untrustworthy evaluation coefficient, a first warning signal is generated, the ECG monitor starts the backup lead and switches to the original waveform storage, and in the cloud-edge collaborative mechanism, for the ECG monitor based on the patient's alarm, the remote diagnosis is suspended and the local alarm is prioritized; If the untrustworthy evaluation coefficient is greater than the second threshold of the untrustworthy evaluation coefficient, and the untrustworthy evaluation coefficient is less than the first threshold, a second warning signal is generated, the ECG monitor starts the enhanced filtering algorithm and compresses the data for transmission, and the cloud-edge collaborative mechanism limits the upload of non-critical data; If the untrustworthy assessment coefficient is less than the second threshold value of the untrustworthy assessment coefficient, no warning signal is generated, the ECG monitor runs in full-function mode, and is synchronized to the medical cloud platform in real time.

7. The method for recommending doctors based on data collected by an electrocardiogram monitor according to claim 6, characterized in that: Determine the healing performance data of different doctors, including: The doctor's healing performance data is represented by the complexity healing coefficient and the specific disease healing coefficient; The acquisition logic of the complexity cure coefficient is as follows: collecting ECG signal data within the monitoring interval, extracting key ECG features in the ECG signal data, standardizing the key ECG features and determining the deviation value of the key ECG features according to the standard value, collecting the patient's medical information, and scoring the risk impact of the key ECG features on the patient based on the patient's medical information to obtain the patient's impact score; The deviation value of the key ECG features in the ECG signal data within the monitoring interval and the patient's impact score are taken as input, and the complexity evaluation coefficient is taken as output through the logistic regression model. The calculation formula of the complexity evaluation coefficient is: ; Among them, FZD is the complexity evaluation coefficient, , , ,……, is the weight, , , ,……, Rate the patient's impact, , , ,……, is the deviation value of the key ECG feature, Based on different historical ECG signal data, the complexity evaluation coefficients of different ECG signal data are divided into different intervals. The performance of different doctors in each complexity evaluation coefficient interval is statistically analyzed to obtain the cure rates of different doctors in different complexity evaluation coefficient intervals. The cure rates of different doctors in different complexity evaluation coefficient intervals are marked as: , where g=1, 2, 3, ..., G, G is a positive integer, and g is the number of different complexity evaluation coefficient intervals; According to the cure rate of doctors in different complexity evaluation coefficient intervals and the overall average cure rate in different complexity evaluation coefficient intervals, the complexity cure coefficient is calculated. The calculation formula is: ;in, is the complexity cure coefficient, QJ is the cure rate of the doctor in the complexity evaluation coefficient interval where the current ECG signal data is located, is the average cure rate of all doctors in the complexity assessment coefficient interval where the previous ECG signal data is located.

8. The method for recommending doctors based on data collected by an electrocardiogram monitor according to claim 7, characterized in that: The logic for obtaining the specific disease cure coefficient is: According to the patient's medical history, the specific disease of the patient is initially obtained. According to the cure history records of different doctors for the specific disease, the cure rate of different doctors for the specific disease is obtained. The specific disease cure coefficient is calculated by the arithmetic mean. The calculation formula is: ;in, is the specific disease cure coefficient, h=1, 2, 3, ..., H, H is a positive integer, h is the specific disease of the patient, The doctor's cure rate for a specific disease.

9. The method for recommending doctors based on data collected by an electrocardiogram monitor according to claim 8, characterized in that: Determine your doctor's referral sequence, including: The weighted sum of the complexity cure coefficient and the specific disease cure coefficient is calculated to build a recommendation evaluation model and generate a recommendation evaluation coefficient. The calculation formula of the recommendation evaluation coefficient is: ;in, is the recommended evaluation coefficient, is the proportionality coefficient of the complexity healing coefficient, is the proportional coefficient of the cure coefficient for a specific disease, , They are both greater than 0; When the ECG signal data uploaded by the patient to the medical cloud platform does not generate a warning signal, the recommended evaluation coefficients of all the doctors on duty are obtained according to the current doctor on duty, and the recommended evaluation coefficients of all the doctors on duty are sorted, and the doctors on duty with large recommended evaluation coefficients are recommended to the patient first.

10. A doctor recommendation device based on data collected by an electrocardiogram monitor, characterized in that: It includes data quality module, data delay module, untrustworthy assessment module, cure performance data module and recommendation assessment module; A data quality module is used to collect ECG signal data, perform waveform analysis and spectrum analysis on the ECG signal data, and calculate data quality information; Data delay module, used to monitor the delay of ECG signal data uploaded to the medical cloud platform and calculate data delay information; An untrustworthy evaluation module is used to calculate the untrustworthy evaluation coefficient of ECG signal data according to data quality information and data delay information, and trigger the response strategy of the ECG monitor and the cloud collaboration mechanism based on the untrustworthy evaluation coefficient; The cure performance data module is used to extract the key ECG features of the ECG signal and calculate the complex cure coefficient and the specific disease cure coefficient in combination with the doctor's cure rate records; The recommendation evaluation module is used to calculate the recommendation evaluation coefficient of the current doctor on duty, sort the doctors on duty according to the recommendation evaluation coefficient, and give priority to recommending the doctors with the highest scores.

Images