An intelligent extraction method and system for key information of a patient recovery period

Abstract

The present application relates to the field of data processing, in particular to a kind of intelligent extraction method and system of patient recovery period key information, the motion data of acquisition through intelligent wearable equipment to identify the motion of user and motion intensity, and when motion intensity reaches certain value, the current motion duration of patient is accumulated.At the same time, the physical data of user is collected in real time, reference physical data matched with patient is determined from physical reference database in combination with current motion intensity, and the similarity of measured physical data and reference physical data is calculated.If similarity is too low, it can be inferred that patient may exist discomfort when exercising, at this time, the question template constructed in advance is used to execute question and answer process, and find the symptom entity that patient may appear when exercising.Finally, accumulated motion duration, symptom entity, measured physical data and motion intensity data are all extracted as key data to provide for doctor, so that doctor has relevant data to refer to and judge the recovery of patient.

Description

[0001] This application is a divisional application of the Chinese application with the application number 202510191037.5, the application date of March 21, 2025, and the invention name of "Intelligent extraction method and system for key information of patient recovery period". TECHNICAL FIELD

[0002] The present application relates to the field of data processing, and particularly relates to an intelligent extraction method and system for key information of patient recovery period. BACKGROUND

[0003] After receiving treatment, patients usually have an observation period. For inpatient patients, medical staff or professional instruments and equipment are used to observe the patients, so as to evaluate the recovery of the patients.

[0004] However, for patients recovering at home, the recovery of the patients can only be understood by relying on regular physical examination and self-description of the patients. During this period, doctors cannot understand the recovery of the patients by observing the physical sign information of the patients. If the condition of the patients deteriorates, doctors cannot intervene in time due to the lack of data collection means, which leads to the extension of the recovery period of the patients or even the difficult recovery of the patients. SUMMARY

[0005] Therefore, the purpose of the present application is to provide an intelligent extraction method and system for key information of patient recovery period to solve the problems in the background.

[0006] In order to achieve the above-mentioned purpose, the present application adopts the following technical scheme:

[0007] The intelligent extraction method for key information of patient recovery period of the present application comprises the following steps:

[0008] Obtaining the identity information of the patient, the measured physical sign data and the motion data of the patient in a target time period, wherein the measured physical sign data and the motion data are collected by the intelligent wearable device of the patient, and the target time period is a time period of a target time length before the current time point;

[0009] Calculating the current motion intensity of the patient based on the identity information of the patient and the motion data in the target time period, and adding the target time length to the cumulative motion time length of the current motion when the current motion intensity is greater than a set value;

[0010] Determining the target reference physical sign data of the patient in the target time period based on the current motion intensity of the patient, the identity information of the patient, and a pre-constructed physical sign reference database, wherein the physical sign reference database comprises reference physical sign data of multiple user groups when performing multiple intensity motions;

[0011] calculating a similarity between the measured vital sign data and the reference vital sign data; and performing a question and answer process with the patient when the similarity is less than a preset similarity threshold, wherein the question and answer process is used to extract a symptom entity;

[0012] constructing recovery period key information of the patient based on the accumulated exercise duration, the current exercise intensity, the measured vital sign data and the symptom entity of the patient in the target time period.

[0013] In an embodiment of the present application, the exercise data of the patient in the target time period is obtained, comprising:

[0014] obtaining the current exercise type and three-axis acceleration data collected by the smart wearable device, wherein the three-axis acceleration data comprises three-axis acceleration data at a plurality of time points ;

[0015] filtering the three-axis acceleration data respectively to obtain three-axis acceleration filtered data, wherein the three-axis acceleration filtered data comprises three-axis acceleration filtered data at a plurality of time points ;

[0016] merging the three-axis acceleration data at the plurality of time points to obtain a resultant acceleration ; ;

[0017] based on the resultant acceleration at the plurality of time points drawing an acceleration time domain variation waveform, extracting all peak points in the acceleration time domain variation waveform, and calculating a standard deviation of all peak points ;

[0018] performing a fast Fourier transform on the acceleration time domain variation waveform to obtain an acceleration frequency domain variation waveform, and extracting a main frequency component from the acceleration frequency domain variation waveform ;

[0019] constructing exercise data based on the standard deviation and the main frequency component .

[0020] In an embodiment of the present application, the current exercise intensity of the patient is calculated based on the identity information of the patient and the exercise data in the target time period, comprising:

[0021] determining a target weight sequence based on the current exercise type, the identity information of the patient and a pre-constructed exercise type-weight sequence relationship table, wherein the exercise type-weight sequence relationship table contains weight sequences of a plurality of user groups when performing a plurality of exercise types;

[0022] The motion data for the target time period are weighted based on the target weight sequence to obtain the current motion intensity. , wherein the current motion intensity The mathematical expression is:

[0023]

[0024] In the formula, As the first weight, It is the second weight.

[0025] In one embodiment of this application, the process of constructing the motion type-weight sequence relationship table includes:

[0026] Acquire multiple sample data, including user information of normal users, exercise type, measured vital sign sample data and exercise sample data of normal users during exercise within the target duration;

[0027] Based on the measured vital signs sample data, multiple sample data... The exercise intensity was labeled, and the measured vital sign sample data were normalized to obtain multiple sample data containing labels. The measured vital signs data include heart rate and blood pressure.

[0028] Based on the user information, multiple sample data containing tags are divided into multiple data units;

[0029] The data units are classified based on the type of motion to obtain multiple data sub-units;

[0030] Multiple intensity clusters are obtained by clustering multiple sample data within a data sub-unit based on motion intensity;

[0031] Calculate the mean motion intensity, mean peak acceleration standard deviation, and mean dominant frequency component of the acceleration frequency domain waveform for each intensity cluster; and construct a mathematical expression for motion intensity based on the mean motion intensity, mean peak acceleration standard deviation, and mean dominant frequency component of the acceleration frequency domain waveform.

[0032]

[0033] In the formula, For the first The average intensity of each intensity cluster's motion. For the first The standard deviation of the peak acceleration of each intensity cluster, For the first The mean value of the dominant frequency component of the acceleration frequency domain change waveform of each intensity cluster;

[0034] Form a system of equations from any two mathematical expressions of motion intensity, and apply the first weight... Second weight To solve this problem, when any data sub-unit has multiple distinct first weights... Or multiple different second weights At that time, for multiple different first weights Or multiple different second weights Calculate the mean to obtain the updated first weight. and the updated second weight ;

[0035] Based on the first weight of each data sub-unit Second weight Construct the motion type-weight sequence relationship table.

[0036] In one embodiment of this application, the process of constructing the vital signs reference database includes:

[0037] Based on heart rate and blood pressure, multiple sample data within each data sub-unit are clustered to obtain multiple vital sign clusters;

[0038] The sample data with abnormal exercise intensity are removed from the aforementioned vital sign clusters to obtain the filtered vital sign clusters;

[0039] Calculate the mean motion intensity of multiple sample data in the filtered vital sign cluster. Standard deviation of exercise intensity mean heart rate Heart rate standard deviation Mean blood pressure and blood pressure standard deviation ;

[0040] Based on the average exercise intensity Standard deviation of exercise intensity Constructing the range of motion intensity of the vital sign cluster Based on the aforementioned mean heart rate and heart rate standard deviation Heart rate range of constructing a cluster of vital signs ; and based on the mean blood pressure and the blood pressure standard deviation Blood pressure range for constructing a cluster of vital signs In the formula, This is a range adjustment parameter;

[0041] A vital signs reference database was constructed based on heart rate and blood pressure ranges across multiple exercise intensity ranges.

[0042] In an embodiment of the present application, the measured vital sign data includes measured heart rates and measured blood pressures at multiple time points, and the similarity between the measured vital sign data and the reference vital sign data is calculated, including:

[0043] The measured heart rates and the measured blood pressures at the multiple time points are normalized respectively, and the average value and the standard deviation of the normalized measured heart rates at the multiple time points are calculated , and the average value and the standard deviation of the normalized measured blood pressures at the multiple time points are calculated ; The heart rate distribution range and the blood pressure distribution range

[0044] of the target time period are constructed, wherein, is a range adjustment parameter; The similarity is calculated based on the heart rate reference range and the blood pressure reference range in the reference vital sign data, and the similarity includes a heart rate similarity and a blood pressure similarity

[0045] , wherein the mathematical expression of the heart rate similarity and the blood pressure similarity is:

[0046]

[0047]

[0048] In the formula, is the length of the heart rate reference range, is the length of the overlapping interval between the heart rate reference range and the heart rate distribution range, is the length of the blood pressure reference range, is the length of the overlapping interval between the blood pressure reference range and the blood pressure distribution range.

[0049] In an embodiment of the present application, a question and answer process is performed with the patient, including:

[0050] A pre-constructed question template is sent to the patient, wherein the question template is constructed based on possible symptom entities;

[0051] When receiving feedback information from the patient, the question and answer process is completed; when not receiving feedback information from the patient, the pre-constructed question template is sent to the patient until feedback information from the patient is received, or the question template is sent for more than a preset threshold number of times, and the question and answer process is completed.

[0052] ​​​​In one embodiment of this application, when the similarity is less than a preset similarity threshold, a question-and-answer process with the patient is executed, including:

[0053] The heart rate similarity Less than a preset similarity threshold or the blood pressure similarity If the similarity is less than a preset threshold, a question-and-answer process with the patient is executed.

[0054] In one embodiment of this application, it further includes:

[0055] Send the patient's key recovery information to the target audience.

[0056] This application also provides an intelligent system for extracting key information during the patient's recovery period, including:

[0057] The acquisition module is used to acquire the patient's measured vital signs data and movement data during a target time period, wherein the measured vital signs data and the movement data are collected by the patient's smart wearable device, and the target time period is the time period of the target duration before the current time point;

[0058] The intensity calculation module is used to calculate the patient's current exercise intensity based on the exercise data of the target time period; and when the current exercise intensity is greater than a set value, the target duration is added to the cumulative exercise duration of this exercise.

[0059] The reference vital signs determination module is used to determine the target reference vital signs data of the patient within the target time period based on the patient's current exercise intensity, the patient's identity information, and a pre-built vital signs reference database. The vital signs reference database includes reference vital signs data of multiple user groups when performing exercise of various intensities.

[0060] The question-and-answer module is used to calculate the similarity between the measured vital signs data and the reference vital signs data; if the similarity is less than a preset similarity threshold, a question-and-answer process with the patient is executed, wherein the question-and-answer process is used to extract symptom entities;

[0061] The information extraction module is used to construct key information about the patient's recovery period based on the patient's cumulative exercise duration, current exercise intensity, measured vital sign data of the target time period, and symptom entities.

[0062] The beneficial effects of the present application are: the intelligent extraction method and system for key information of a patient in a recovery period, which identifies the motion condition and motion intensity of a user through motion data collected by an intelligent wearable device, accumulates the current motion duration of the patient when the motion intensity reaches a certain value, simultaneously collects the physical data of the user in real time, determines the reference physical data matched with the patient from a physical reference database in combination with the current motion intensity, and calculates the similarity between the measured physical data and the reference physical data. If the similarity is too low, it can be inferred that the patient may have discomfort during the motion, at this time, the pre-constructed question template is used to perform the question and answer process to find the symptom entity that may appear during the motion of the patient. Finally, the accumulated motion duration, symptom entity, measured physical data and motion intensity data are all extracted as key data to provide to the doctor, so that the doctor has relevant data to refer to and judge the recovery condition of the patient. BRIEF DESCRIPTION OF DRAWINGS

[0063] The present application will be further described below in conjunction with the drawings and embodiments:

[0064] Figure 1 is a use scene diagram of an intelligent extraction method for key information of a patient in a recovery period in an embodiment of the present application;

[0065] Figure 2 is a flowchart of an intelligent extraction method for key information of a patient in a recovery period in an embodiment of the present application;

[0066] Figure 3 is a construction flowchart of a motion type-weight sequence relationship table and a physical reference database in the present application;

[0067] Figure 4 is a structure diagram of an intelligent extraction system for key information of a patient in a recovery period in an embodiment of the present application. DETAILED DESCRIPTION

[0068] The embodiments of the present application will be described below through specific concrete examples, and those skilled in the art can easily understand other advantages and effects of the present application from the disclosure of the present specification. The present application can also be implemented or applied through other different specific embodiments, and each detail in the present specification can be modified or changed based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.

[0069] It is to be noted that the drawings provided in the following embodiments only schematically illustrate the basic concept of the present application, and only the layers related to the present application are shown in the drawings, rather than being drawn according to the number, shape and size of the layers in actual implementation. The actual implementation of each layer can be a random change, and the layer layout pattern can be more complex.

[0070] In the following description, a large number of details are discussed to provide a more thorough explanation of the embodiments of the present application, however, it is obvious to those skilled in the art that the embodiments of the present application can be implemented without these specific details.

[0071] Figure 1 is a use scene diagram of an intelligent extraction method of key information of a patient recovery period in an embodiment of the present application, as shown in Figure 1 In the present application, the motion data (such as three-axis acceleration, body tilt angle, etc.) and the vital sign data (such as heart rate and blood pressure) of the user are collected by using the intelligent wearable device such as the smart bracelet. Limited by the data processing capability of the smart bracelet, the collected data is uploaded to the cloud for calculation and data processing. The motion intensity of the user, the cumulative motion duration, and the vital sign data of the user are extracted and provided to the target object as the basic data for judging the physical recovery condition of the user.

[0072] In addition, the information of the related personnel involved in the present application is obtained with full authorization, and the collection, use and processing of the related information need to comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0073] Figure 2 is a flowchart of an intelligent extraction method of key information of a patient recovery period in an embodiment of the present application, as shown in Figure 2 The intelligent extraction method of key information of a patient recovery period in the present embodiment can include the following steps:

[0074] S210, obtaining the identity information of the patient, the measured vital sign data and the motion data of the patient in a target time period, wherein the measured vital sign data and the motion data are collected by the intelligent wearable device of the patient, and the target time period is a time period of a target duration before the current time point;

[0075] Specifically, the identity information mainly includes age and gender. The vital sign data of people of different ages and genders is quite different when they are exercising, so it is necessary to distinguish them.

[0076] The measured vital sign data includes the measured heart rate and the measured blood pressure.

[0077] The motion data mainly includes motion type data collected by the smart wearable device and three-axis acceleration data. The acquisition process is as follows:

[0078] S211, acquiring the current motion type and three-axis acceleration data collected by the smart wearable device, wherein the three-axis acceleration data includes three-axis acceleration at multiple time points ;

[0079] The three-axis acceleration data generated by the user under different motion types is different in corresponding motion intensity. For example, when riding a bicycle and running, similar three-axis acceleration data generated often corresponds to different motion intensity, so the motion type recognition function of the smart bracelet is needed to identify the current motion type of the user.

[0080] The smart bracelet is usually equipped with various sensors, such as accelerometers, gyroscopes, heart rate sensors, etc. These sensors can monitor various physiological parameters and motion states of the user in real time. For example, the accelerometer can detect the moving direction and speed change of the user; the gyroscope can sense the direction and rotation of the wrist. In the prior art, the motion type of the user can be identified by recognizing the step frequency, step length, body tilt angle, arm swing mode, etc.

[0081] The three-axis acceleration is an important basic data of the user's motion intensity.

[0082] S212, filtering the three-axis acceleration data respectively to obtain three-axis acceleration filtering data, wherein the three-axis acceleration filtering data includes three-axis acceleration at multiple time points after filtering ;

[0083] The application adopts low-pass filtering to smooth the signal, the purpose is to remove noise, smooth the signal to facilitate the extraction of useful motion information.

[0084] S213, merging the three-axis acceleration at multiple time points to obtain the combined acceleration ;

[0085]

[0086] The combined acceleration after merging is one-dimensional data, converting three-dimensional data into one-dimensional data can effectively simplify the data dimension, which is convenient for subsequent data processing.

[0087] S214, based on the combined acceleration at multiple time points draw the acceleration time domain variation waveform, extract all peak points in the acceleration time domain variation waveform, and calculate the standard deviation of all peak points ;

[0088] The peak point reflects the characteristics of the waveform, including the amplitude characteristics and the frequency characteristics. The standard deviation of the amplitude of the peak point can reflect the stability of the acceleration change.

[0089] S215, performing a fast Fourier transform on the acceleration time-domain change waveform to obtain an acceleration frequency-domain change waveform, and extracting a main frequency component from the acceleration frequency-domain change waveform ;

[0090] By extracting the main frequency component of the acceleration frequency-domain change waveform, the main frequency of the acceleration change can be obtained, reflecting the speed information of the acceleration change.

[0091] S216, constructing motion data based on the standard deviation and the main frequency component .

[0092] The present application mainly uses the stability of the acceleration change and the speed of the acceleration change to reflect the motion intensity, so the standard deviation and the main frequency component are mainly extracted.

[0093] The present application adopts a real-time collection method, that is, once the data extraction process is triggered, the measured physical sign data and motion data of multiple time points in the target time period before the current time point are used as basic data for analysis, processing and extraction. The trigger condition can be a timing trigger or a proactive trigger.

[0094] S220, calculating the current motion intensity of the patient based on the identity information of the patient and the motion data of the target time period; and when the current motion intensity is greater than a set value, adding the target time length to the cumulative motion time length of the current motion;

[0095] In the present application, the motion intensity of the patient is identified by the motion type data and the three-axis acceleration data. In the prior art, heart rate and blood pressure data may be introduced to determine the motion intensity of the patient, but the present application is aimed at patients in the recovery period, so it is most likely that abnormal heart rate and blood pressure data will occur during exercise. Introducing heart rate and blood pressure will cause problems in determining the motion intensity. Therefore, only the motion type data and the three-axis acceleration data are used to identify the motion intensity of the patient, and the specific process is as follows:

[0096] S221, determining a target weight sequence based on the current motion type, the identity information of the patient and a pre-constructed motion type-weight sequence relationship table, wherein the motion type-weight sequence relationship table contains the weight sequence of a plurality of user groups when performing a plurality of motion types;

[0097] S222, weight the motion data of the target time period based on the target weight sequence, to obtain a current motion intensity wherein the current motion intensity is mathematically expressed as:

[0098]

[0099] wherein, is a first weight, is a second weight.

[0100] In order to more accurately describe the motion intensity of the patient, different weight sequences are constructed in advance for different motion types and user groups to weight the standard deviation and the dominant frequency component and obtain the current motion intensity.

[0101] Figure 3 is a construction flowchart of the motion type-weight sequence relationship table and the physical sign reference database in the present application, as shown in Figure 3 Specifically, the construction process of the motion type-weight sequence relationship table in the present application includes:

[0102] (1) obtaining a plurality of sample data, wherein the sample data includes user information of a normal user, a motion type, measured physical sign sample data and motion sample data sampled by the normal user during motion in a target time period;

[0103] The sample data is generated by the normal user wearing a smart bracelet during motion, and is collected after being authorized by the normal user.

[0104] (2) labeling the motion intensity of the plurality of sample data based on the measured physical sign sample data, and normalizing the measured physical sign sample data to obtain a plurality of sample data containing labels, wherein the measured physical sign sample data includes heart rate and blood pressure;

[0105] Since the sample data is generated by the normal user, the heart rate and blood pressure thereof have reference significance. It is more reasonable to label the motion intensity by using the heart rate and blood pressure. However, considering the individual differences between different users, the normal user feedback can be used for labeling at the standard time. For example, the motion intensity is from 0-100, and the user can evaluate the intensity of the motion process by himself.

[0106] (3) dividing the plurality of sample data containing labels into a plurality of data units based on the user information;

[0107] (4) classifying the data units based on the motion type to obtain a plurality of data subunits;

[0108] Factors influencing the correlation between vital signs such as heart rate and blood pressure and exercise intensity mainly include age, gender, and exercise type. Therefore, this application first uses user information to divide multiple tagged sample data into multiple data units, ensuring that each data unit contains only sample data generated by user groups of a specific age and gender. Then, the data units are further divided so that each sub-unit contains only sample data generated when performing a single type of exercise.

[0109] (5) Clustering multiple sample data within a data sub-unit based on motion intensity to obtain multiple intensity clusters;

[0110] This application only clusters the intensity of the exercise, that is, it clusters the one-dimensional data. The K-means clustering algorithm can be used to divide it into a fixed number of intensity clusters, such as 2-5 clusters.

[0111] (6) Calculate the mean motion intensity, mean peak acceleration standard deviation, and mean dominant frequency component of the acceleration frequency domain waveform in each intensity cluster; and construct a mathematical expression for motion intensity based on the mean motion intensity, mean peak acceleration standard deviation, and mean dominant frequency component of the acceleration frequency domain waveform:

[0112]

[0113] In the formula, For the first The average intensity of each intensity cluster's motion. For the first The standard deviation of the peak acceleration of each intensity cluster, For the first The mean value of the dominant frequency component of the acceleration frequency domain change waveform of each intensity cluster;

[0114] For each intensity cluster, the distribution of motion intensity is similar. Since the sample data generated by a group of users performing the same type of motion within the cluster are similar, the motion data are also similar. This application calculates the mean motion intensity, the mean peak standard deviation of acceleration, and the mean of the main frequency component of the acceleration frequency domain change waveform within the cluster, thereby reflecting the relationship between motion intensity and motion data in each intensity cluster.

[0115] Since multiple clusters were defined in the preceding text, multiple mathematical expressions for motion intensity can be constructed. In this application, the mathematical expression for motion intensity is a linear equation in two variables. By solving the two equations simultaneously, the values ​​of the first weight and the second weight can be obtained. In this application, any two mathematical expressions for motion intensity are combined into a system of equations, thereby calculating one or more solutions for the first weight and the second weight.

[0116] (7) Form an equation group with the mathematical expressions of any two exercise intensities, and solve the first weight and the second weight When there are multiple different first weights or multiple different second weights for any data subunit, the average of the multiple different first weights or the multiple different second weights is calculated to obtain the updated first weight and the updated second weight ;

[0117] Finally, if there is only one set of solutions, the first weight and the second weight can be directly obtained, and if there are multiple sets of solutions, the average of each set of solutions can be calculated.

[0118] (8) Construct the exercise type-weight sequence relationship table based on the first weight and the second weight of each data subunit.

[0119] Through the above process, the weight sequence corresponding to the exercise data of different user groups during different exercises can be obtained, and the exercise intensity can be more accurate and consistent with the labeling rules when the exercise data of the current user is brought in.

[0120] When the exercise intensity of the patient in the current time period is calculated, if the exercise intensity is greater than a certain value, it is determined that the patient has high-intensity exercise, and the duration of the high-intensity exercise can reflect the physical recovery of the patient, so the duration of the target time period is added to the cumulative time of the current exercise. Key information: cumulative exercise duration.

[0121] S230, determining target reference sign data of the patient in the target time period based on the current exercise intensity of the patient, the identity information of the patient, and a pre-constructed sign reference database, wherein the sign reference database includes reference sign data of multiple user groups during multiple intensity exercises;

[0122] After obtaining the current exercise intensity calculated from the exercise data, the application also refers to the sign data of normal users and the sign data of the current patient to determine whether the patient has abnormal sign data. If the sign is abnormal, whether it is accompanied by abnormal symptoms, these are key information that needs to be extracted.

[0123] In the present application, the reference sign data of normal users during similar scene exercises is provided by the pre-constructed sign reference database. The construction method of the sign reference database is as follows:

[0124] For example Figure 3As shown, the construction of the vital signs reference database is based on the data sub-units described above.

[0125] (1) Cluster the multiple sample data in each data sub-unit based on heart rate and blood pressure to obtain multiple vital sign clusters;

[0126] This application uses heart rate and blood pressure for clustering, which can be viewed as clustering two-dimensional data; therefore, the DSCAN density clustering algorithm is employed. The clustering degree of data points and the number of clusters are controlled by setting the maximum distance within each cluster.

[0127] (2) Remove sample data with abnormal exercise intensity from the said physical sign cluster to obtain the filtered physical sign cluster;

[0128] This application employs a cyclic elimination method to remove abnormal sample data, specifically including:

[0129] Calculate the mean and variance of the movement intensity within each cluster of vital signs;

[0130] If the variance within a cluster of vital signs is greater than the preset variance threshold, it indicates that the clustering of motion intensity data within the cluster is insufficient. In this case, the deviation between each data point and the mean within the cluster is calculated, the sample data with the largest deviation is removed, and the calculation of the mean and variance of motion intensity within the cluster is repeated until the variance is less than or equal to the variance threshold.

[0131] (3) Calculate the mean motion intensity of multiple sample data in the filtered physical sign cluster. Standard deviation of exercise intensity mean heart rate Heart rate standard deviation Mean blood pressure and blood pressure standard deviation ;

[0132] After the previously filtered-out vital sign clusters, the heart rate and blood pressure distribution characteristics are similar, and the exercise intensity values ​​are also relatively clustered. We can consider a correlation between exercise intensity and heart rate / blood pressure within each cluster. That is, when a normal user group performs a certain type of exercise, if the exercise intensity reaches the range within the cluster, then the corresponding heart rate and blood pressure will most likely fall within the corresponding heart rate and blood pressure range of that cluster.

[0133] (4) Based on the average exercise intensity Standard deviation of exercise intensity Constructing the range of motion intensity of the vital sign cluster Based on the aforementioned mean heart rate and heart rate standard deviation Heart rate range of constructing a cluster of vital signs ; and based on the mean blood pressure and the blood pressure standard deviation Blood pressure range for constructing a cluster of vital signs In the formula, This is a range adjustment parameter;

[0134] (5) Construct a vital signs reference database based on heart rate and blood pressure ranges across multiple exercise intensity ranges.

[0135] Finally, a reference database of vital signs is constructed by building a range based on the mean and standard deviation.

[0136] In this application, the DSCAN clustering algorithm is used to obtain clusters, and then the motion intensity range is obtained. The constructed motion intensity range may have blind spots, that is, areas that cannot be covered. In this case, the range adjustment parameter can be adjusted. To cover the blind spots.

[0137] After constructing the vital signs reference database, the corresponding vital signs reference data (range) can be found by collecting the patient's user information, exercise type, and exercise intensity.

[0138] S240, calculate the similarity between the measured vital signs data and the reference vital signs data; if the similarity is less than a preset similarity threshold, execute a question-and-answer process with the patient, wherein the question-and-answer process is used to extract symptom entities;

[0139] In this application, similarity can reflect whether the patient's measured vital signs data match the corresponding reference vital signs data. If they match, it indicates that the recovery is good. If there is a large discrepancy, it indicates that the recovery is poor, or even that abnormal symptoms may occur.

[0140] The similarity calculation process is as follows:

[0141] S241, normalize the measured heart rate and measured blood pressure at multiple time points, and calculate the normalized measured heart rate at multiple time points. average Standard deviation And calculate the normalized measured blood pressure at multiple time points. average Standard deviation ;

[0142] This application employs min-max normalization to eliminate data discrepancies between heart rate and blood pressure, facilitating subsequent calculations and processing.

[0143] Then calculate the measured heart rate. average Standard deviation Actual blood pressure measurement average , standard deviation , the distribution of heart rate and blood pressure of the patient in the target time period.

[0144] S242, construct the heart rate distribution range of the target time period and the blood pressure distribution range , wherein, is a range adjustment parameter;

[0145] S243, calculate the similarity based on the heart rate reference range and the blood pressure reference range in the reference sign data, the similarity includes heart rate similarity and blood pressure similarity , wherein the heart rate similarity and the blood pressure similarity The mathematical expression is:

[0146]

[0147]

[0148] In the formula, is the length of the heart rate reference range, is the length of the overlapping interval of the heart rate reference range and the heart rate distribution range, is the length of the blood pressure reference range, is the length of the overlapping interval of the blood pressure reference range and the blood pressure distribution range.

[0149] The present application compares two ranges to obtain the similarity. Among them, indicates the length of the heart rate distribution range, indicates the ratio of the length of the heart rate distribution range to the length of the heart rate reference range, indicates the ratio of the length of the overlapping interval of the heart rate reference range and the heart rate distribution range to the length of the heart rate reference range. Both parameters can reflect the similarity of the range.

[0150] In the present application, , so that the reference range is larger, and the heart rate distribution range in the target time period is smaller. The advantage of this setting is that more reference data can be preserved as much as possible, and the main heart rate data in the target time period can be extracted. Increase the fault tolerance rate of abnormal judgment, avoid the emergence of too sensitive abnormal judgment mechanism.

[0151] After obtaining the similarity, if the similarity is too small, it means that the patient's signs during exercise do not conform to the big data rule, and some symptoms may occur, such as feeling strong or fast heartbeat, shortness of breath or difficulty breathing, dizziness or mild headache, chest discomfort or pain, blurred vision or temporary vision loss, facial flushing or tinnitus, etc.

[0152] Specifically, when the heart rate similarity is less than a preset similarity threshold or the blood pressure similarity is less than a preset similarity threshold, a question and answer process with the patient is performed.

[0153] In order to confirm whether the patient has the above symptoms, the application constructs an inquiry template in advance, for example:

[0154] “Do you have the following symptoms: (1) feel strong or fast heartbeat; (2) shortness of breath or difficulty breathing; (3) dizziness or mild headache; (4) chest discomfort or pain; (5) blurred vision or temporary loss of vision; (6) facial flushing or tinnitus... If so, please reply to the corresponding serial number”

[0155] Then the pre-constructed question template is sent to the patient; when receiving feedback information from the patient, the question and answer process is completed; for example, returning (2), (5), the corresponding symptom entity “shortness of breath or difficulty breathing” and “blurred vision or temporary loss of vision” can be obtained.

[0156] When no feedback information is received from the patient, the pre-constructed question template is sent to the patient until feedback information is received from the patient, or the number of times of sending the question template exceeds a preset threshold, the question and answer process is completed. If the user does not reply all the time, after three times, no longer send, end communication, only extract the sign data.

[0157] S250, based on the cumulative exercise duration of the patient, the current exercise intensity, the measured sign data of the target time period and the symptom entity, the recovery period key information of the patient is constructed.

[0158] Finally, the cumulative exercise duration, the current exercise intensity, the measured sign data of the target time period and the symptom entity related to the recovery situation are extracted as key information, and the recovery period key information of the patient is sent to the target object. The target object (such as medical staff or related monitoring personnel) can understand the recovery situation of the patient in real time according to the key information, make suggestions or intervention in time, and optimize the recovery process of the patient.

[0159] The application further provides an intelligent extraction method of key information of a patient in a recovery period, which recognizes the motion condition and motion intensity of a user through motion data collected by an intelligent wearable device, accumulates the current motion duration of the patient when the motion intensity reaches a certain value, collects the physical sign data of the user in real time, determines the reference physical sign data matched with the patient from a reference physical sign database in combination with the current motion intensity, and calculates the similarity between the measured physical sign data and the reference physical sign data. If the similarity is too low, it can be inferred that the patient may have discomfort during the motion, at which time a question and answer process is performed by using a pre-constructed question template to find the symptom entity that may appear during the motion of the patient. Finally, the accumulated motion duration, the symptom entity, the measured physical sign data and the motion intensity data are all extracted as key data to provide the doctor, so that the doctor has relevant data to refer to and judge the recovery condition of the patient.

[0160] As shown in Figure 4 , the application further provides an intelligent extraction system of key information of a patient in a recovery period, comprising:

[0161] An acquisition module is configured to acquire measured physical sign data and motion data of a patient in a target time period, wherein the measured physical sign data and the motion data are collected by an intelligent wearable device of the patient, and the target time period is a time period of a target duration before a current time point;

[0162] An intensity calculation module is configured to calculate the current motion intensity of the patient based on the motion data of the target time period, and accumulate the target duration into the accumulated motion duration of the current motion when the current motion intensity is greater than a set value;

[0163] A reference physical sign determination module is configured to determine target reference physical sign data of the patient in the target time period based on the current motion intensity of the patient, the identity information of the patient and a pre-constructed reference physical sign database, wherein the reference physical sign database comprises reference physical sign data of multiple user groups when performing multiple intensity motions;

[0164] A question and answer module is configured to calculate the similarity between the measured physical sign data and the reference physical sign data, and perform a question and answer process with the patient when the similarity is less than a preset similarity threshold, wherein the question and answer process is used to extract a symptom entity;

[0165] An information extraction module is configured to construct the key information of the patient in the recovery period based on the accumulated motion duration of the patient, the current motion intensity, the measured physical sign data of the target time period and the symptom entity.

[0166] The intelligent extraction system of key information in the recovery period of a patient provided in the present application identifies the motion condition and motion intensity of a user through the motion data collected by an intelligent wearable device, accumulates the current motion duration of the patient when the motion intensity reaches a certain value, simultaneously collects the physical data of the user in real time, determines the reference physical data matched with the patient from a reference physical database in combination with the current motion intensity, and calculates the similarity between the measured physical data and the reference physical data. If the similarity is too low, it can be inferred that the patient may have discomfort during the motion, at this time, the question and answer process is performed by using the pre-constructed question template to find the symptom entity that may appear during the motion of the patient. Finally, the accumulated motion duration, symptom entity, measured physical data and motion intensity data are all extracted as key data to provide the doctor, so that the doctor has relevant data to refer to and judge the recovery condition of the patient.

[0167] The embodiment also provides an electronic terminal, comprising a processor and a memory.

[0168] The memory is used for storing a computer program, and the processor is used for executing the computer program stored in the memory, so that the terminal executes any method in the embodiment.

[0169] The computer readable storage medium in the embodiment can be understood by those skilled in the art that all or part of the steps of the above-mentioned method embodiments can be completed by the hardware of the computer program. The foregoing computer program can be stored in a computer readable storage medium. The program executes the steps of the above-mentioned method embodiments when executed; and the foregoing storage medium includes ROM, RAM, magnetic disc or optical disc and various storage program codes.

[0170] The electronic terminal provided in the embodiment includes a processor, a memory, a transceiver and a communication interface, the memory and the communication interface are connected with the processor and the transceiver and complete the communication between each other, the memory is used for storing a computer program, the communication interface is used for communication, and the processor and the transceiver are used for running the computer program, so that the electronic terminal executes each step of the method.

[0171] In the embodiment, the memory can include random access memory (RAM) and can also include non-volatile memory, for example, at least one disk memory.

[0172] The processor described above can be a general processor, including a central processing unit (CPU), a network processor (NP), etc.; can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component.

[0173] In the above embodiments, although the present application has been described in conjunction with specific embodiments thereof, numerous alternatives, modifications and variations will be readily apparent to those of ordinary skill in the art in the light of the foregoing descriptions. The embodiments of the present application are intended to embrace all such alternatives, modifications and variations as falling within the scope of the appended claims.

[0174] The above embodiments only illustrate the principles and effects of the present application, but are not used to limit the present application. Any modification or change made by any person skilled in the art without departing from the spirit and scope of the present application shall be covered by the claims of the present application.

Claims

1. An intelligent extraction method of key information of a patient recovery period, characterized in that, The method comprises the steps of: Obtaining identity information of a patient, measured vital sign data and motion data of the patient in a target time period, wherein the measured vital sign data and the motion data are collected by a smart wearable device of the patient, and the target time period is a time period of a target time length before a current time point; obtaining motion data of the patient in the target time period includes: obtaining a current motion type and three-axis acceleration data collected by the smart wearable device, wherein the three-axis acceleration data includes three-axis acceleration at multiple time points ; filtering the three-axis acceleration data respectively to obtain three-axis acceleration filtering data, wherein the three-axis acceleration filtering data includes three-axis acceleration at the multiple time points after filtering ; merging the three-axis acceleration at the multiple time points to obtain combined acceleration ; drawing an acceleration time domain variation waveform based on the combined acceleration at the multiple time points , extracting all peak points in the acceleration time domain variation waveform, and calculating a standard deviation of all the peak points ; performing fast Fourier transform on the acceleration time domain variation waveform to obtain an acceleration frequency domain variation waveform, and extracting a main frequency component from the acceleration frequency domain variation waveform ; constructing motion data based on the standard deviation and the main frequency component ; calculating the current exercise intensity of the patient based on the patient's identity information and the exercise data of the target time period; and when the current exercise intensity is greater than a set value, adding the target duration to the cumulative exercise duration of the current exercise; determining the target reference sign data of the patient in the target time period based on the current exercise intensity of the patient, the patient's identity information, and the pre-constructed sign reference database, wherein the sign reference database comprises reference sign data of multiple user groups when performing exercises of multiple intensities; Calculate the similarity between the measured vital sign data and the target reference vital sign data; if the similarity is less than a preset similarity threshold, execute a question-and-answer process with the patient, wherein the question-and-answer process is used to extract symptom entities; the measured vital sign data includes measured heart rate and measured blood pressure at multiple time points, and calculating the similarity between the measured vital sign data and the target reference vital sign data includes: normalizing the measured heart rate and measured blood pressure at multiple time points respectively, and calculating the normalized measured heart rate at multiple time points. average Standard deviation And calculate the normalized measured blood pressure at multiple time points. average Standard deviation Construct the heart rate distribution range for the target time period. and blood pressure distribution range ,in, The range adjustment parameter is used; a similarity is calculated based on the heart rate reference range and blood pressure reference range in the target reference vital sign data, wherein the similarity includes heart rate similarity. Similarity to blood pressure The heart rate similarity Similarity to the blood pressure The mathematical expression is: wherein, is a length of the heart rate reference range, is a length of the overlapping interval of the heart rate reference range and the heart rate distribution range, is a length of the blood pressure reference range, is a length of the overlapping interval of the blood pressure reference range and the blood pressure distribution range; constructing the recovery period key information of the patient based on the cumulative exercise duration of the patient, the current exercise intensity, the measured sign data of the target time period, and the symptom entity; sending the recovery period key information of the patient to a target object.

2. The method as claimed in claim 1, wherein, The method comprises the steps of: calculating the current exercise intensity of the patient based on the patient's identity information and the exercise data of the target time period; and when the current exercise intensity is greater than a set value, adding the target duration to the cumulative exercise duration of the current exercise; weighting the motion data of the target time period based on the target weight sequence to obtain a current motion intensity wherein the current motion intensity is mathematically expressed as: wherein, is a first weight, is a second weight.

3. The method as claimed in claim 2, wherein the key information is extracted from the patient recovery period by using the intelligent extraction method. determining the target weight sequence based on the current exercise type, the patient's identity information, and the pre-constructed exercise type-weight sequence relationship table, wherein the exercise type-weight sequence relationship table contains the weight sequence of multiple user groups when performing multiple exercise types; The construction process of the exercise type-weight sequence relationship table comprises: annotating a plurality of sample data based on the measured physical sign sample data , and normalizing the measured physical sign sample data to obtain a plurality of sample data containing labels , wherein the measured physical sign sample data comprises heart rate and blood pressure; obtaining multiple sample data, wherein the sample data comprises user information of normal users, exercise types, measured sign sample data and exercise sample data sampled by normal users when exercising for a target duration; dividing multiple sample data containing labels into multiple data units based on the user information; classifying the data units based on exercise types to obtain multiple data subunits; clustering multiple sample data in the data subunits based on exercise intensity to obtain multiple intensity clusters; wherein, is the mean of the intensity cluster number, is the mean of the intensity cluster number, is the mean of the intensity cluster number, is the standard deviation of the acceleration peak value of the intensity cluster number, is the mean of the intensity cluster number, is the mean of the intensity cluster number; Solving the equation group composed of the mathematical expressions of any two exercise intensities, and solving the first weight and the second weight , when there are multiple different first weights or multiple different second weights for any data subunit, averaging the multiple different first weights or the multiple different second weights to obtain the updated first weight and the updated second weight ; a first weight for each data subunit and a second weight constructing the motion type-weight sequence relationship table.

4. The intelligent extraction method for key information during the patient's recovery period according to claim 3, characterized in that, calculating the mean value of exercise intensity, the mean value of acceleration peak standard deviation, and the mean value of acceleration frequency domain change waveform main frequency component in each intensity cluster; and constructing a mathematical expression of exercise intensity based on the mean value of exercise intensity, the mean value of acceleration peak standard deviation, and the mean value of acceleration frequency domain change waveform main frequency component: The construction process of the sign reference database comprises: clustering multiple sample data in each data subunit based on heart rate and blood pressure to obtain multiple sign clusters; calculating a mean of motion intensity of the plurality of sample data in the screened sign cluster , a standard deviation of motion intensity , a mean of heart rate , a standard deviation of heart rate , a mean of blood pressure , and a standard deviation of blood pressure ; based on the mean of the motion intensity , the standard deviation of the motion intensity , the motion intensity range of the sign cluster ; based on the mean of the heart rate and the standard deviation of the heart rate , the heart rate range of the sign cluster ; and based on the mean of the blood pressure and the standard deviation of the blood pressure , the blood pressure range of the sign cluster , where is a range adjustment parameter; eliminating sample data with abnormal exercise intensity from the sign clusters to obtain filtered sign clusters; 5. The method as claimed in claim 1, wherein the said method is used for the intelligent extraction of critical information during the recovery period of the patient. constructing a sign reference database based on the heart rate range and blood pressure range of multiple exercise intensity ranges. performing a question and answer process with the patient, comprising: sending a pre-constructed question template to the patient, wherein the question template is constructed based on possible symptom entities; 6. The method as claimed in claim 1, wherein the said method is used for the intelligent extraction of critical information during the recovery period of the patient. when receiving feedback information from the patient, completing the question and answer process; when not receiving feedback information from the patient, returning to send the pre-constructed question template to the patient until receiving feedback information from the patient, or when the number of question template sending exceeds a pre-set threshold, completing the question and answer process. when the heart rate similarity less than a preset similarity threshold, or the blood pressure similarity less than a preset similarity threshold, a question and answer process with the patient is performed.

7. An intelligent extraction system of critical information of a patient recovery period, characterized in that, when the similarity is less than a pre-set similarity threshold, performing a question and answer process with the patient, comprising: comprising: The acquisition module is used for acquiring measured vital sign data and motion data of a patient in a target time period, wherein the measured vital sign data and the motion data are collected by a smart wearable device of the patient, and the target time period is a time period of a target time length before a current time point; acquiring motion data of the patient in the target time period comprises: acquiring a current motion type and three-axis acceleration data collected by the smart wearable device, wherein the three-axis acceleration data comprises three-axis acceleration at multiple time points ; filtering the three-axis acceleration data respectively to obtain three-axis acceleration filtering data, wherein the three-axis acceleration filtering data comprises three-axis acceleration at the multiple time points after filtering ; merging the three-axis acceleration at the multiple time points to obtain combined acceleration ; drawing an acceleration time domain variation waveform based on the combined acceleration at the multiple time points , extracting all peak points in the acceleration time domain variation waveform, and calculating a standard deviation of all the peak points ; performing fast Fourier transform on the acceleration time domain variation waveform to obtain an acceleration frequency domain variation waveform, and extracting a main frequency component from the acceleration frequency domain variation waveform ; constructing motion data based on the standard deviation and the main frequency component ; an intensity calculation module, configured to calculate a current exercise intensity of the patient based on exercise data of the target time period, and add the target time period to a cumulative exercise time of the current exercise when the current exercise intensity is greater than a set value; a reference sign determination module, configured to determine target reference sign data of the patient in the target time period based on the current exercise intensity of the patient, identity information of the patient, and a pre-constructed sign reference database, wherein the sign reference database comprises reference sign data of multiple user groups when performing exercises of multiple intensities; The question and answer module is configured to calculate the similarity between the measured vital sign data and the target reference vital sign data, and perform a question and answer process with the patient when the similarity is less than a preset similarity threshold, wherein the question and answer process is configured to extract a symptom entity. The measured vital sign data includes measured heart rates and measured blood pressures at multiple time points. The similarity between the measured vital sign data and the target reference vital sign data is calculated by normalizing the measured heart rates and the measured blood pressures at the multiple time points, calculating the average value and the standard deviation of the normalized measured heart rates at the multiple time points, and calculating the average value and the standard deviation of the normalized measured blood pressures at the multiple time points. The heart rate distribution range and the blood pressure distribution range of a target time period are constructed, wherein the heart rate distribution range is calculated based on the average value and the standard deviation of the normalized measured heart rates at the multiple time points, and the blood pressure distribution range is calculated based on the average value and the standard deviation of the normalized measured blood pressures at the multiple time points. The similarity between the measured vital sign data and the target reference vital sign data is calculated based on the heart rate reference range and the blood pressure reference range in the target reference vital sign data, and the similarity includes a heart rate similarity and a blood pressure similarity. The mathematical expressions of the heart rate similarity and the blood pressure similarity are as follows. ​​​​​​​​​​​​​ wherein, is the length of the heart rate reference range, is the length of the overlapping interval of the heart rate reference range and the heart rate distribution range, is the length of the blood pressure reference range, is the length of the overlapping interval of the blood pressure reference range and the blood pressure distribution range; an information extraction module, configured to construct recovery period key information of the patient based on the cumulative exercise time of the patient, the current exercise intensity, measured sign data of the target time period, and a symptom entity; a sending module, configured to send the recovery period key information of the patient to a target object.

Images