A method for generating electronic medical records for users of a health checkup hut

Abstract

The present invention relates to the field of data processing technology, and specifically to a method for generating electronic medical records for users of a health checkup hut, comprising: obtaining to-be-processed medical examination data of a target user and a reference medical examination data set of a reference user; obtaining a current contradiction index based on the deviation between the to-be-processed medical examination data of the target user and the normal data range in each data dimension, combined with the distribution of historical medical examination data, and screening the data dimensions to obtain contradiction dimensions; obtaining a reference priority based on the data change trend and time distribution of each reference user in each contradiction dimension, combined with data prediction results and the current contradiction index; obtaining information richness based on the data periodicity characteristics and data fluctuation characteristics of the reference user in each contradiction dimension within a preset time period; obtaining key data dimensions, and then generating an electronic medical record for the target user. The present invention improves the credibility of electronic medical records and the efficiency of doctors in discovering problems during review.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method for generating electronic medical records for users of a health examination room. Background Art

[0002] The Health Checkup Hut is a grassroots health service facility that integrates self-service physical examinations, health assessments, health interventions, and popular science education. By combining intelligent equipment, big data, and internet technology, it provides residents with convenient, free, or low-cost health management services. The Health Hut can upload valid detected data to the main control computer to generate a new electronic medical record. Doctors can use the electronic medical record to promptly guide any abnormalities discovered outside the hospital back to the hospital, forming a complete service system and closed-loop process system. Patients can log in to their personal health space to query their personal electronic medical records, such as physical examination information, outpatient records, hospitalization records, monitoring information, health assessments, etc., so they can find the cause of various physical discomforts in a timely manner and effectively prevent them.

[0003] The existing method of constructing electronic medical records is mainly reflected in the fact that when there is a contradiction between new test data and historical records, such as a user's blood sugar suddenly rises, the traditional system only performs simple overwriting and storage, lacks big data analysis and comparison, and makes the contradiction warning mechanism of long-term electronic medical records less effective. Summary of the Invention

[0004] In order to solve the technical problem that the traditional system only performs simple overwriting storage and lacks big data analysis and comparison, resulting in a poor conflict warning mechanism for long-term electronic medical record generation, the purpose of the present invention is to provide a method for generating user electronic medical records in a health check-up room. The technical solution adopted is as follows:

[0005] Obtain the target user's pending physical examination data and historical physical examination data in each data dimension, as well as reference users belonging to the same user category as the target user and their reference physical examination data sets;

[0006] Based on the deviation between the target user's to-be-processed physical examination data in each data dimension and the normal data range in the corresponding data dimension, combined with the data distribution in the corresponding data dimension in the historical physical examination data, the current contradiction index of the target user in each data dimension is obtained, and the data dimension is screened using the current contradiction index to obtain the contradiction dimension;

[0007] Based on the data change trend of each reference user in each conflict dimension in the reference physical examination dataset and the time distribution of the reference users' physical examinations, combined with the data prediction results under each conflict dimension and the current conflict index of the target user, the reference priority of each conflict dimension is obtained;

[0008] Within the preset time period of the physical examination data to be processed, obtaining the information richness of each contradiction dimension based on the data periodicity characteristics and data fluctuation characteristics of each reference user in the reference physical examination data set in each contradiction dimension;

[0009] Each conflict dimension of the target user is screened in combination with the reference priority and information richness to obtain a key data dimension; and an electronic medical record of the target user is generated based on the to-be-processed physical examination data of the key data dimension.

[0010] Preferably, the current conflict index of the target user in each data dimension is obtained based on the deviation between the target user's to-be-processed physical examination data in each data dimension and the normal data range in the corresponding data dimension, combined with the data distribution in the corresponding data dimension in the historical physical examination data, specifically including:

[0011] Determine the data difference factor of the target user in each data dimension based on the difference between the target user's to-be-processed physical examination data in each data dimension and the normal data range in the same data dimension;

[0012] Obtain the standard deviation and mean of the target user's historical physical examination data in each data dimension; obtain the target user's data deviation factor in each data dimension based on the difference between the target user's to-be-processed physical examination data in each data dimension and the mean and the standard deviation;

[0013] A current contradiction index of the target user in each data dimension is obtained according to the data difference factor and the data deviation factor; wherein both the data difference factor and the data deviation factor are positively correlated with the current contradiction index.

[0014] Preferably, the reference priority of each contradiction dimension is obtained based on the data change trend of each reference user in each contradiction dimension in the reference physical examination data set and the time distribution of the reference user's physical examination, combined with the data prediction result under each contradiction dimension and the current contradiction index of the target user, specifically including:

[0015] According to the data deviation of each reference user in each data dimension in the reference physical examination dataset, the reference users are divided into deviant users and unified users in each contradictory dimension;

[0016] Based on the data change trend of the unified users under each contradiction dimension and the time distribution of the unified users' physical examinations, combined with the data prediction results under each contradiction dimension, the future deviation factor of the unified users in each contradiction dimension is obtained;

[0017] Using a preset value as a future deviation factor of a deviating user in each conflict dimension; wherein the future deviation factor of a unified user in each conflict dimension is less than the preset value;

[0018] Under each contradiction dimension, the product of the cumulative sum of the future deviation factors corresponding to the unified user and the deviated user and the current contradiction index of the target user in the same contradiction dimension is normalized to obtain the reference priority of each contradiction dimension.

[0019] Preferably, dividing the reference users into deviated users and uniform users according to the data deviation of each reference user in each data dimension in the reference physical examination data set specifically includes:

[0020] Obtain the data deviation factor of each reference user in each contradictory dimension, record the reference users whose data deviation factor is greater than the preset deviation threshold as deviant users, and record the reference users whose data deviation factor is less than or equal to the preset deviation threshold as uniform users.

[0021] Preferably, the method of obtaining the future deviation factor of the unified user in each contradiction dimension based on the data change trend of the unified user in each contradiction dimension and the time distribution of the unified user undergoing physical examinations, combined with the data prediction results in each contradiction dimension, specifically includes:

[0022] Fitting the data of each unified user under each conflict dimension to obtain a data curve of each unified user under each conflict dimension;

[0023] Calculating the average change trend of the data curves of all unified users under the same contradiction dimension to obtain the average change trend of each contradiction dimension; calculating the average time interval of physical examinations for all unified users under the same contradiction dimension to obtain the average physical examination time length of each contradiction dimension; using the product of the average change trend and the average physical examination time length to determine the predicted data of the unified users of each contradiction dimension;

[0024] Based on the predicted data of the unified user of each contradiction dimension and the standard deviation and mean of the reference physical examination data of all unified users of the same contradiction dimension in the reference physical examination dataset, the future deviation factor of the unified user of each contradiction dimension is determined.

[0025] Preferably, the information richness of each contradiction dimension is obtained according to the data periodicity characteristics and data fluctuation characteristics of each reference user in the reference physical examination data set in each contradiction dimension within the preset time period of the physical examination data to be processed, specifically including:

[0026] The target user's pending medical examination data corresponds to a preset time period during which the medical examination took place;

[0027] Calculate the average usage frequency of the health check-up room by all reference users to obtain the first characteristic coefficient;

[0028] Obtaining a second characteristic coefficient for each contradiction dimension according to a fluctuation degree of the reference physical examination data of each reference user in each contradiction dimension in the reference physical examination data set within a preset time period;

[0029] Perform periodic feature analysis based on all reference physical examination data of each contradiction dimension at each time node within a preset time period to obtain the third characteristic coefficient of each contradiction dimension;

[0030] A normalized result of the product of the first characteristic coefficient, the second characteristic coefficient, and the third characteristic coefficient is used as the information richness of each contradiction dimension.

[0031] Preferably, obtaining the second characteristic coefficient of each contradiction dimension according to the fluctuation degree of the reference physical examination data of each reference user in each contradiction dimension in the reference physical examination data set within a preset time period specifically includes:

[0032] Obtain the mean of the reference physical examination data of all reference users at the same time node under each conflict dimension within a preset time period as the feature data of each conflict dimension at each time node;

[0033] The variance of the characteristic data of each contradiction dimension at all time nodes within a preset time period is used as the second characteristic coefficient of each contradiction dimension.

[0034] Preferably, the periodic characteristic analysis is performed based on all reference physical examination data of each contradiction dimension at each time node within a preset time period to obtain the third characteristic coefficient of each contradiction dimension, specifically including:

[0035] Perform Fourier transform on the characteristic data of all time nodes of each contradiction dimension within a preset time period, and obtain the inverse of the amplitude corresponding to the maximum frequency component in the Fourier transform result of each contradiction dimension as the third characteristic coefficient of each contradiction dimension.

[0036] Preferably, the step of screening each conflict dimension of the target user in combination with the reference priority and information richness to obtain key data dimensions specifically includes:

[0037] Calculate the normalized result of the product of the reference priority and information richness of each contradiction dimension to obtain the feature criticality of each contradiction dimension; take the contradiction dimension corresponding to the feature criticality greater than or equal to the preset critical threshold as the key data dimension.

[0038] Preferably, the method of using the current contradiction index to filter the data dimension to obtain the contradiction dimension specifically includes:

[0039] The data dimension corresponding to the current contradiction index being greater than or equal to the preset abnormality threshold is taken as the contradiction dimension.

[0040] The embodiments of the present invention have at least the following beneficial effects:

[0041] The present invention first collects the newly added physical examination data and historical physical examination data of the target user, and at the same time obtains reference users of the same type as the target user to provide a data reference basis. Then, the current contradiction index of the target user in each data dimension is obtained. The deviation of the current real-time collected data is analyzed by the difference between the physical examination data of the target user in real time and the normal range, combined with the distribution of the historical physical examination data of the target user, and then the possibility of contradiction in the physical examination data to be processed in each data dimension is evaluated to screen out the data dimensions with greater contradictions to participate in the subsequent group data feature analysis process. Secondly, the reference physical examination data set characterizes the group user characteristics. By analyzing the data change trend and time analysis of the group user characteristics, and taking into account the results of data prediction, the degree to which the data of each contradiction dimension needs to be given priority is quantified, that is, the reference priority characterizes the priority of the corresponding contradiction dimension. Further, by analyzing the periodic characteristics and fluctuations of the group data within the same preset time period, the richness of the data of each contradiction dimension is measured, thereby accurately identifying the key information in the data. Finally, by combining the results of the two feature analyses, we ensure that rich and reliable data dimensions are subsequently screened out, thereby improving the credibility of electronic medical records and the efficiency of doctors in discovering problems during review. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0043] Figure 1 This is a flowchart of the steps of a method for generating electronic medical records for users of a health checkup hut provided by the present invention;

[0044] Figure 2 is a flowchart of the steps of the method for obtaining the current contradiction index of each data dimension provided by the present invention;

[0045] Figure 3 is a flowchart of the steps of the method for obtaining the reference priority of each conflict dimension provided by the present invention;

[0046] Figure 4 is a flowchart of the steps of the method for obtaining the future deviation factor provided by the present invention;

[0047] Figure 5It is a flowchart of the steps of the method for obtaining the information richness of each contradiction dimension provided by the present invention. DETAILED DESCRIPTION

[0048] To further illustrate the technical means and effectiveness of the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail the specific implementation, structure, features, and effectiveness of a method for generating electronic medical records for users of a health checkup cabin proposed by the present invention. In the following description, different references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0049] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0050] The specific implementation scenario targeted by this invention is that when patients have chronic diseases or are recovering from illness, the health check-up cabin can quickly monitor their physical condition. This is extremely valuable for patients who need complication screening, regular condition monitoring, and recovery status assessment. After the patient completes the examination in the health check-up cabin, the system generates an electronic medical record based on the monitoring data. Doctors can remotely view the patient's electronic medical record via the internet to determine whether the patient needs to be recalled for examination or treatment.

[0051] The main purpose of this invention is to compare and analyze the real-time physical examination information of the current patient with the historical data set through big data analysis and comparison, so as to determine the data dimensions that are inconsistent with its own historical data and the historical data of the patient group, that is, key information, and then construct a new electronic medical record to provide doctors with more detailed data support for observing the patient's status.

[0052] The following describes in detail a method for generating electronic medical records for users of a health checkup hut provided by the present invention with reference to the accompanying drawings.

[0053] See also Figure 1 , which shows a flowchart of a method for generating electronic medical records for users of a health checkup room provided by one embodiment of the present invention, the method comprising the following steps:

[0054] Step S100 : obtaining the target user's to-be-processed physical examination data and historical physical examination data in each data dimension, as well as reference users belonging to the same user category as the target user and their reference physical examination data sets.

[0055] In the process of building electronic medical records, the patient's new physical examination data needs to be added to the original medical record to form a new electronic medical record. However, for patients with chronic diseases or patients in the recovery period, their physical condition will continue to change (disease progression or recovery). That is, after the examination in the health check-up hut, there will be information mismatch between the new physical examination data and the original medical record. For example, the indicators of chronic diseases may gradually deteriorate or improve over time. Therefore, the system needs to distinguish the reasons for the mismatch, so as to identify the real key information in the new physical examination data and eliminate the influence of other factors.

[0056] Based on this, we first construct a data set for the health check-up hut. Specifically, we collect the patient physical examination data in each health check-up hut based on the edge node algorithm. Since multi-source data may have differences in format, data units, etc., it is necessary to standardize the collected data, summarize the standardized data, form a patient group data set, and store the patient group data set in each health check-up hut.

[0057] It should be noted that the method for normalizing data is a well-known technology and is only briefly introduced here. The normalization method is shown in Table 1 below:

[0058] Table 1 Standardization method table

[0059] Normalized dimensions Technical Solution Terminology standardization SNOMED CT / LOINC code mapping, such as: "hypertension" → SNOMED:38341003 Unit unification UCUM (Uniform Unit of Measurement) converter: "mg / dL→mmol / L"=value×0.02586 Time alignment NTP time synchronization and local clock compensation algorithm Spatial standardization Geographic coordinate conversion (WGS84→GCJ02) + location code (such as community health station ID)

[0060] It should be understandable that the patient group dataset covers all hospitals that cooperate with the health check-up hut. The patient group dataset also covers the user data of all users who have undergone long-term testing in the health check-up hut. Therefore, after each historical user uses the health check-up hut to generate new test data, the user's past test data can be directly extracted from the database. At the same time, some of the physical examination data of other users who belong to the same user group as the user can be screened out to provide a data basis for comparative analysis of the newly added physical examination data.

[0061] Specifically, the target user's pending physical examination data and historical physical examination data in each data dimension are obtained, where the target user refers to the user who uses the health check-up hut to generate an electronic medical record this time; the pending physical examination data refers to the new physical examination data generated by the target user during this use, which is used to generate a new electronic medical record; one type of physical examination data corresponds to one data dimension, for example, blood sugar data is a data dimension; historical physical examination data refers to the physical examination data generated by the target user during past physical examinations.

[0062] Furthermore, based on the basic information of the target user and the address information of the health check-up hut, the user group with the same disease type, gender, age difference of less than 5 years and in the same region as the target user is screened in the patient group dataset as the reference users of the target user. The physical examination data of each reference user in each data dimension constitutes the reference physical examination dataset of the reference user.

[0063] Step S200: Based on the deviation between the target user's to-be-processed physical examination data in each data dimension and the normal data range in the corresponding data dimension, combined with the data distribution in the corresponding data dimension in the historical physical examination data, the current contradiction index of the target user in each data dimension is obtained, and the data dimension is screened using the current contradiction index to obtain the contradiction dimension.

[0064] Patients with different diseases can undergo corresponding tests in the health checkup cabin based on their individual testing needs and obtain the required physical data. By comparing the test data (new physical examination data) with the original medical records before the physical examination, the differences in each data dimension can be obtained, and the data dimensions with conflicts can be identified.

[0065] As a specific example, Figure 2 As shown, the specific steps of quantifying the current contradiction index of each data dimension can be implemented by steps S201 to S203, and the user screens the data dimensions in which contradictions exist.

[0066] Step S201 : determining a data difference factor of the target user in each data dimension based on the difference between the to-be-processed physical examination data of the target user in each data dimension and the normal data range in the same data dimension.

[0067] It should be noted that for each different data dimension, the normal data under normal conditions may be a standard value or a standard data range. Therefore, specific analysis is conducted for different situations. In essence, it is to obtain the difference between the target user's to-be-processed physical examination data in each data dimension and the normal data under the same data dimension.

[0068] When the normal data of a data dimension is a standard value, the absolute value of the difference between the target user's to-be-processed physical examination data in each data dimension and the standard value in the same data dimension is obtained, and normalized, to obtain the data difference factor of the target user in the data dimension.

[0069] When the normal data for a data dimension falls within a standard data range, and the target user's pending medical examination data exceeds the standard data range, the absolute value of the difference between the target user's pending medical examination data for each data dimension and the boundary value of the annotated data range for the same data dimension is obtained and normalized to obtain the target user's data difference factor for that data dimension. When the target user's pending medical examination data falls within the standard data range, the data difference factor for the target user's corresponding data dimension is set to 0, indicating that there is no abnormality for the target user in that data dimension.

[0070] The data difference factor represents the degree of deviation between the target user's newly added physical examination data and the normal situation in the corresponding data dimension. The larger the value, the more significant the deviation. It should be noted that the normalization method is a well-known technology and will not be described in detail here.

[0071] Step S202: Obtain the standard deviation and mean of the historical physical examination data of the target user in each data dimension; obtain the data deviation factor of the target user in each data dimension based on the difference between the target user's to-be-processed physical examination data in each data dimension and the mean and the standard deviation.

[0072] In this embodiment, in order to avoid the long time period between the collection time of historical data and the current newly added physical examination data, resulting in low data reference value, historical physical examination data of 10 physical examinations before the current physical examination data is generated are obtained for feature analysis.

[0073] Specifically, taking the i-th data dimension as an example, the calculation method of the data deviation factor of the target user in the i-th data dimension can be expressed as: ,in Indicates the data deviation factor of the target user in the i-th data dimension, represents the target user's pending physical examination data in the i-th data dimension, represents the personal data set consisting of the target user's historical physical examination data in the i-th data dimension, represents the mean of the target user’s personal data set in the i-th data dimension, Represents the standard deviation of the target user’s personal dataset in the i-th data dimension.

[0074] The calculation process for the data deviation factor is essentially a simple variation of the Z-Score normalization process. While the Z-Score normalization process measures the position of data points within a normal distribution, the data deviation factor eliminates directional differences by using absolute values, focusing on the degree of deviation rather than the direction. The data deviation factor represents the data deviation of the target user in the corresponding data dimension.

[0075] Step S203 , obtaining the current contradiction index of the target user in each data dimension according to the data difference factor and the data deviation factor; wherein both the data difference factor and the data deviation factor are positively correlated with the current contradiction index.

[0076] In a normal distribution, most data fall within the range of ±3σ of the mean, where σ represents the standard deviation. For patients with chronic diseases, if an indicator suddenly exceeds 3 times the historical standard deviation, it is very likely to indicate a worsening of the disease (such as a sudden rise in blood sugar in diabetes) or a detection error, and must be marked as contradictory data first.

[0077] Based on this feature, when the data deviates from the factor When the target user's data fluctuations fall within the normal range of the target user's historical band, the contradiction degree of the target user's newly added physical examination data in the i-th data dimension is low. When the target user's data fluctuates beyond the normal level, it may indicate that the condition has worsened or abnormal detection conditions have occurred, and the contradiction index needs to be strengthened.

[0078] Specifically, when the data deviate from the factor When the target user's contradiction weight in the corresponding data dimension is set to a preset parameter, the preset parameter value is a very small positive number, such as 0.01, which indicates that the possibility of contradiction in the target user's physical examination data in the data dimension is very small. However, the degree of contradiction can be further quantified by combining the value of the data difference factor. When the data deviates from the factor When As the contradiction weight of the target user in the i-th data dimension, the larger its value is, the greater the possibility that there is a contradiction in the physical examination data to be processed in the corresponding data dimension.

[0079] Finally, the normalized result of the product of the target user's contradiction weight in the i-th data dimension and the data difference factor in the i-th data dimension is used as the current contradiction index of the target user in the i-th data dimension.

[0080] The current contradiction index representation of the target user in each data dimension is analyzed by analyzing the difference between the target user's current real-time physical examination data and the normal range, combined with the distribution of the target user's historical physical examination data, to analyze the deviation of the current real-time collected data, and then evaluate the possibility of contradiction in the physical examination data to be processed in each data dimension.

[0081] The larger the value of the current contradiction index, the greater the possibility that the physical examination data to be processed in the corresponding data dimension contains contradictions and anomalies. The smaller the value of the current contradiction index, the smaller the possibility that the physical examination data to be processed in the corresponding data dimension contains contradictions and anomalies. Based on this, the current contradiction index is used to filter the data dimensions to obtain the contradiction dimension.

[0082] More specifically, the data dimension corresponding to the current contradiction index being greater than or equal to a preset anomaly threshold is considered the contradiction dimension. In this embodiment, the anomaly threshold is set to 0.4, and implementers can set this threshold based on the specific implementation scenario. The contradiction dimension represents the data dimension in which the target user's pending medical examination data contains contradictions and anomalies.

[0083] Step S300: According to the data change trend of each reference user in each contradiction dimension in the reference physical examination data set and the time distribution of the reference user's physical examination, combined with the data prediction result under each contradiction dimension and the current contradiction index of the target user, the reference priority of each contradiction dimension is obtained.

[0084] After screening out data dimensions with inconsistencies and anomalies in the newly added health checkup data, we conduct a conflict reference priority assessment on these data dimensions. By analyzing group data on these conflicting dimensions, we can not only screen out special data dimensions that differ from group data changes, but also predict the change trend of a certain data dimension of the current user based on group data changes. This allows us to provide different degrees of positive incentives for different conflicting dimensions, helping to identify key information in the newly added health checkup data.

[0085] As a specific example, Figure 3 As shown, the method for obtaining the reference priority of each contradiction dimension can be implemented by steps S301 to S304.

[0086] Step S301 : dividing the reference users according to the data deviation of each reference user in each data dimension in the reference physical examination data set to obtain deviated users and unified users in each conflicting dimension.

[0087] Obtain the data deviation factor of each reference user in each contradiction dimension, record the reference users whose data deviation factor is greater than the preset deviation threshold as the deviated users under this contradiction dimension, and record the reference users whose data deviation factor is less than or equal to the preset deviation threshold as the unified users under this contradiction dimension.

[0088] It should be noted that the data deviation factor for each conflicting dimension for the reference user is obtained in the same manner as the data deviation factor for each data dimension for the target user. Refer to step S202 and will not be further explained here. A deviation threshold is then used to categorize all reference users: users with more severe data deviation are designated as deviating users, while users with less severe data deviation are designated as uniform users. In this embodiment, the deviation threshold is set to 0.4; implementers can adjust this threshold based on their specific implementation scenario.

[0089] For all reference users, a deviating user indicates that the reference user possesses deviating characteristics. Deviating characteristics can indicate the opposite characteristics between the current data dimension and patients with the same disease. These characteristics are key information for the user, and the greater the degree of deviation, the more important the information. A consistent user indicates that the reference user possesses consistent characteristics. These consistent characteristics indicate that the current conflicting dimension is consistent with the normal variation trend within that data dimension. It should be noted that for each conflicting dimension, there are different consistent users and deviating users, meaning that the same reference user may have different characteristics in different conflicting dimensions.

[0090] Step S302 : According to the data change trend of the unified user in each contradiction dimension and the time distribution of the unified user's physical examination, combined with the data prediction result in each contradiction dimension, the future deviation factor of the unified user in each contradiction dimension is obtained.

[0091] The reference health examination dataset for each user in each conflict dimension represents the distribution of reference data that exhibits uniform and normal data changes within the corresponding conflict dimension. By analyzing the changing trends of the reference data for all users under each conflict, data prediction can be achieved. This means predicting the degree of deviation in a user's future health examination data based on group data trends and individual historical characteristics, providing a basis for proactive early warning of electronic medical records.

[0092] As a specific example, Figure 4 As shown, the method for obtaining the future deviation factor can be implemented by steps S3021 to S3023.

[0093] Step S3021 , fitting the data of each unified user in each conflict dimension to obtain a data curve of each unified user in each conflict dimension.

[0094] Specifically, the least squares method can be used to fit all reference data of each unified user under each contradiction, and then for each contradiction dimension, one unified user corresponds to one data curve. In other embodiments, the implementer can also select other data fitting methods for processing according to the specific implementation scenario.

[0095] Step S3022: Calculate the average value of the change trend of the data curves of all unified users under the same contradiction dimension to obtain the average change trend of each contradiction dimension; calculate the average value of the time intervals for physical examinations of all unified users under the same contradiction dimension to obtain the average physical examination time length of each contradiction dimension; and use the product of the average change trend and the average physical examination time length to determine the predicted data of the unified users of each contradiction dimension.

[0096] In the first step, we'll use any conflict dimension as an example. This conflict dimension is designated as the selected conflict dimension. Under this selected conflict dimension, the average slope of the data curves for all users is taken as the average change trend of the selected conflict dimension. It should be understood that the slope of each data curve can be obtained using well-known techniques, such as calculating the average slope of all data points on the data curve. This method will not be further elaborated here. The average change trend represents the average level of the change trend of the group data.

[0097] The second step is to obtain the average checkup interval for each user within the selected conflict dimension. The average of all the average checkup intervals is then calculated to obtain the average checkup duration for the selected conflict dimension. The average checkup duration represents the average level of checkup cycles for the group data.

[0098] In the third step, the product of the average change trend of the selected conflict dimension and the average length of the medical examination is used as the predicted data for the selected conflict dimension in terms of the unified characteristic corresponding to the unified user. The predicted data represents the results of data prediction based on the data change trend and medical examination period of reference users with unified characteristics under each conflict dimension. It reflects the expected performance of each conflict dimension in the next medical examination of this user group under normal circumstances.

[0099] This shows that the quantification process of predictive data, firstly, takes into account that individual user trends may be influenced by accidental factors, while the group average trend better reflects the general development of the disease. Secondly, considering that the frequency of physical examinations reflects the density of data updates, the average duration can be used to standardize the time scale. Assuming that the future changes in user indicators conform to the group average trend, and using the average duration as the prediction period, the average change trend and the prediction period can be used to determine the data prediction results based on unified characteristics for each conflicting dimension.

[0100] Step S3023 : determining the future deviation factor of the unified user for each contradiction dimension based on the predicted data of the unified user for each contradiction dimension and the standard deviation and mean of the reference physical examination data of all unified users for the same contradiction dimension in the reference physical examination dataset.

[0101] The future deviation factor for a unified user in each conflict dimension represents the data deviation factor corresponding to the predicted data for the unified user in each conflict dimension, reflecting the deviation of the predicted data for the user group under normal circumstances. The specific method for obtaining the data deviation factor for the target user in each data dimension is the same as that for step S202, and will not be further explained here.

[0102] The quantification process of the future deviation factor represents the process of analyzing the group trend of unified users (with less data deviation) and predicting the degree of deviation of their future physical examination data, thus providing a forward-looking risk assessment for electronic medical records.

[0103] Step S303 : Using a preset value as a future deviation factor of the deviating user in each conflict dimension; wherein the future deviation factor of the unified user in each conflict dimension is smaller than the preset value.

[0104] The current data of a deviant user exhibits characteristics that are opposite to those of similar patients (for example, while most diabetic patients have elevated blood sugar, one patient's blood sugar suddenly drops). This constitutes critical information, and its future trend is difficult to predict using a group average model. Therefore, a fixed high value is used to directly mark the user as high risk. In this embodiment, the preset value is 1, which is greater than or equal to the future deviation factor of the unified user, ensuring that the deviant user dominates subsequent priority calculations. Because the data characteristics of the deviant user conflict with those of the group, complex trend predictions for the deviant user are avoided, and a high preset value is used to quickly trigger an alert.

[0105] Step S304 , under each contradiction dimension, normalize the product of the cumulative sum of the future deviation factors corresponding to the unified user and the deviation user and the current contradiction index of the target user in the same contradiction dimension to obtain a reference priority of each contradiction dimension.

[0106] The reference priority for each conflict dimension represents the reference or priority level for the conflict dimension calculated by combining the future deviation factor and the current conflict index, providing a decision ranking for physician review. By integrating the current conflict and future risk corresponding to each conflict dimension, the data-driven predictive model is directly linked to clinical decision-making needs, enabling the upgrade of electronic medical records from recording anomalies to priority warnings.

[0107] In the first step, under the selected contradiction dimension, the cumulative sum of the future deviation factors of the unified users and the deviated users is calculated, and the user group represented by all reference users under the selected contradiction dimension is used to analyze the predicted risk, reflecting the future data deviation.

[0108] In the second step, the product of the accumulated sum and the target user's current conflict index for the selected conflict dimension is normalized and used as the reference priority for the selected conflict dimension. It should be noted that normalization methods are well-known and will not be detailed here. Implementers can choose a method based on their specific implementation scenario, such as maximum-minimum normalization.

[0109] Reference priorities indicate the priority level of consideration for corresponding conflicting dimensions. Specifically, dimensions with a high current conflict index and a large future deviation factor warrant prioritization. Furthermore, doctors can use reference priorities to quickly identify data dimensions with significant current anomalies and high future risk, improving the efficiency of electronic medical record review.

[0110] Step S400 , within a preset time period of the physical examination data to be processed, obtain the information richness of each contradiction dimension according to the data periodicity characteristics and data fluctuation characteristics of each reference user in the reference physical examination data set in each contradiction dimension.

[0111] Compared to comprehensive hospital exams, health checkup cabins are more convenient and targeted. Patients can frequently visit for testing to obtain short-term data changes. Once sufficient data is available, the data trend can reflect the changing characteristics of any data dimension. For example, if the temperature drops significantly on a given day, human cholesterol levels will generally increase. When patients undergo physical examinations at different times or seasons, the data obtained will be affected by physiological fluctuations, the environment, and other factors. For example, female patients may experience a 10%-15% drop in hemoglobin levels during their menstrual period, or cholesterol levels generally increase in winter.

[0112] Based on this, by periodically analyzing the data change trends of the reference users' reference physical examination datasets in different contradiction dimensions and comparing the characteristics of different contradiction dimensions changing over time, the data criticality of the target users in each contradiction dimension is determined.

[0113] As a specific example, Figure 5 As shown, the method for obtaining the information richness of each contradiction dimension can be implemented by steps S401 to S404.

[0114] Step S401: within a preset time period in which the target user's to-be-processed physical examination data corresponds to a physical examination, the average usage frequency of the health examination room by all reference users is calculated to obtain a first characteristic coefficient.

[0115] In this embodiment, the physical examination data of each data dimension in the patient group data set are divided according to months, that is, this embodiment performs feature analysis on the month in which the target user's pending physical examination data is subjected to this physical examination, wherein the preset time period is also the month in which the target user is subjected to this physical examination.

[0116] Then, within a preset time period, the average time intervals between each reference user's use of the health checkup cabin are calculated to obtain the average checkup duration for each reference user. The inverse of the mean of the average checkup durations for all reference users within the preset time period is taken as the average usage frequency, reflecting the data density distribution of the target user's use of the health checkup cabin in the month. The first characteristic coefficient indicates that the key information analysis process focuses on the data of the target user's checkup in the current month, ensuring that the analysis matches the timeliness of the current health status. For example, the blood pressure fluctuations of hypertensive patients in winter may be different from those in summer.

[0117] The greater the average frequency of use of the health check-up hut by all reference users within the preset time period, the richer the reference check-up information for each contradiction dimension in that month, and the more reliable the periodic analysis results.

[0118] Step S402 : obtaining a second characteristic coefficient of each contradiction dimension according to the fluctuation degree of the reference physical examination data of each reference user in each contradiction dimension in the reference physical examination data set within a preset time period.

[0119] Secondly, we analyze the fluctuations in the data for each conflict dimension for reference users over the same time period. Specifically, we obtain the mean of the reference health checkup data for all reference users at the same time point for each conflict dimension within a preset time period as the characteristic data for each conflict dimension at each time point. The variance of the characteristic data for each conflict dimension at all time points within the preset time period is used as the second characteristic coefficient for each conflict dimension. In this embodiment, each time point is considered to be one day.

[0120] Specifically, the preset time period can also be regarded as the current month. The first step is to calculate the mean of the reference physical examination data of all reference users under each contradiction dimension every day in the current month as the characteristic data of each contradiction dimension on each day, and use the variance of these characteristic data of each contradiction dimension in the current month as the second characteristic coefficient.

[0121] The larger the variance, the larger the value of the corresponding second characteristic coefficient, indicating that the reference group data fluctuates more dramatically within that month, which may reflect that the disease changes are complex or significantly affected by external factors (such as season and environment). The physical examination corresponding to such contradictory dimensions contains more key information.

[0122] Step S403 : performing periodic characteristic analysis on all reference physical examination data of each contradiction dimension at each time node within a preset time period to obtain a third characteristic coefficient of each contradiction dimension.

[0123] Specifically, a Fourier transform is performed on the characteristic data of each contradiction dimension at all time nodes within a preset time period. The inverse of the amplitude corresponding to the maximum frequency component in the Fourier transform result of each contradiction dimension is obtained as the third characteristic coefficient of each contradiction dimension. It should be noted that the Fourier transform method is a well-known technology and will not be further described here.

[0124] The amplitude corresponding to the maximum frequency component represents the strength of the frequency component and reflects the periodicity of the data distribution. A larger value indicates a stronger periodicity in the reference physical examination dataset corresponding to the contradictory dimension, while a smaller value indicates a weaker periodicity in the reference physical examination dataset corresponding to the contradictory dimension.

[0125] The larger the amplitude corresponding to the maximum frequency component, the more significant the periodicity (such as hemoglobin fluctuations during a fixed monthly menstrual period). Such fluctuations are predictable physiological phenomena, and the data criticality of the corresponding contradictory dimension is low. The smaller the amplitude, the weaker the periodicity, indicating that the data changes in the corresponding contradictory dimension are irregular (such as sudden deterioration of chronic diseases), which is more likely to reflect individual-specific abnormalities, and the corresponding contradictory dimension is highly critical.

[0126] Step S404: taking a normalized result of the product of the first characteristic coefficient, the second characteristic coefficient, and the third characteristic coefficient as the information richness of each contradiction dimension.

[0127] The first characteristic coefficient characterizes the extent to which the data of each conflicting dimension may contain key information based on the data density distribution within a preset time period. The second characteristic coefficient characterizes the extent to which the data of each conflicting dimension may contain key information based on the data fluctuations of the reference physical examination dataset within a preset time period. The third characteristic coefficient characterizes the extent to which the data of each conflicting dimension may contain key information based on its periodic characteristics.

[0128] Based on this, the information richness of each contradiction dimension was obtained by combining the characteristic analysis results of the three aspects, which reflects that the physical examination data of the contradiction dimensions with high information richness need to be paid priority attention.

[0129] It should be noted that the method for normalizing data in this embodiment is a well-known technology, and the implementer can choose it according to the specific implementation scenario, such as the maximum-minimum normalization method, which is not limited here.

[0130] Step S500 , screening each conflict dimension of the target user in combination with the reference priority and information richness to obtain key data dimensions; generating an electronic medical record of the target user based on the physical examination data to be processed in the key data dimensions.

[0131] Specifically, by integrating reference priority and information richness, we can identify the most valuable key data dimensions for clinical decision-making from conflicting dimensions and generate informative electronic medical records based on these conflicting dimensions. Essentially, this approach transforms data-driven analytical results into clinically applicable key information, enhancing the decision-support value of electronic medical records.

[0132] More specifically, the normalized result of the product of the reference priority and information richness of each contradiction dimension is calculated to obtain the feature criticality of each contradiction dimension. The larger the reference priority value of each contradiction dimension, the more doctors need to pay attention to the pending physical examination data in this contradiction dimension, as shown by the results of big data analysis. At the same time, the larger the information richness value of the corresponding contradiction dimension, the higher the richness of the physical examination data in this contradiction dimension, and further, the more critical the pathological information contained in this contradiction dimension, and the more attention it requires. By combining the results of the two aspects of feature analysis, it is ensured that the data dimensions subsequently screened out are both high-risk and information-reliable.

[0133] Furthermore, the greater the value of the characteristic criticality of each contradiction dimension, the higher the information criticality of the corresponding contradiction dimension, and the greater the degree of attention the doctor needs to pay. The smaller the value of the characteristic criticality of each contradiction dimension, the lower the information criticality of the corresponding contradiction dimension, and the doctor can reduce the degree of attention to this data dimension. Specifically, the contradiction dimension corresponding to the characteristic criticality greater than or equal to the preset critical threshold is regarded as the key data dimension. In this embodiment, the critical threshold is 0.5, and the implementer can set it according to the specific implementation scenario.

[0134] Finally, the key data dimensions are extracted for processing, and the key data is highlighted or structuredly annotated to improve the efficiency of doctor review. The basic data of non-key data dimensions is retained, but the display priority is lowered. It should be understood that data dimensions other than key data dimensions are non-key data dimensions.

[0135] Based on this, it is possible to filter out all key information in the newly added physical examination data set and overwrite the original electronic medical record content based on this key information to generate a new electronic medical record. At the same time, key information is highlighted during the overwriting process to increase doctors' attention when reviewing the electronic medical record.

[0136] In summary, by comparing the target user's newly added physical examination data with their original medical records and a group of patients with the same disease, we can screen out data dimensions that conflict with the original medical records, as well as the key information within these conflicting dimensions, allowing the constructed electronic medical records to more accurately reveal changing trends in the disease. At the same time, we can analyze the temporal characteristics of the patient population dataset. Based on the temporal changes in the characteristic values ​​of different data dimensions, we can identify those data dimensions that are more likely to express information that contradicts that of similar patients (non-periodic data dimensions). This allows us to accurately identify key information in the newly added data, improve the credibility of the electronic medical records, and improve the efficiency of doctors in identifying problems during review.

[0137] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. A method for generating electronic medical records for users of a health check-up room, characterized in that: The method comprises the following steps: Obtain the target user's pending physical examination data and historical physical examination data in each data dimension, as well as reference users belonging to the same user category as the target user and their reference physical examination data sets; Based on the deviation between the target user's to-be-processed physical examination data in each data dimension and the normal data range in the corresponding data dimension, combined with the data distribution in the corresponding data dimension in the historical physical examination data, the current contradiction index of the target user in each data dimension is obtained, and the data dimension is screened using the current contradiction index to obtain the contradiction dimension; The method for obtaining the current contradiction index is as follows: based on the difference between the target user's to-be-processed physical examination data in each data dimension and the normal data range in the same data dimension, determining the target user's data difference factor in each data dimension; obtaining the standard deviation and mean of the target user's historical physical examination data in each data dimension; based on the difference between the target user's to-be-processed physical examination data in each data dimension and the mean and the standard deviation, obtaining the target user's data deviation factor in each data dimension; obtaining the target user's current contradiction index in each data dimension according to the data difference factor and the data deviation factor; wherein both the data difference factor and the data deviation factor are positively correlated with the current contradiction index; Based on the data change trend of each reference user in each conflict dimension in the reference physical examination dataset and the time distribution of the reference users' physical examinations, combined with the data prediction results under each conflict dimension and the current conflict index of the target user, the reference priority of each conflict dimension is obtained; Among them, the method for obtaining the reference priority is: according to the data deviation of each reference user in each data dimension in the reference physical examination data set, the reference users are divided into deviated users and unified users under each contradiction dimension; according to the data change trend of the unified user under each contradiction dimension and the time distribution of the unified user undergoing physical examination, combined with the data prediction results under each contradiction dimension, the future deviation factor of the unified user in each contradiction dimension is obtained; a preset value is used as the future deviation factor of the deviated user in each contradiction dimension; wherein the future deviation factor of the unified user in each contradiction dimension is less than the preset value; under each contradiction dimension, the cumulative sum of the future deviation factors corresponding to the unified user and the deviated user and the product of the current contradiction index of the target user in the same contradiction dimension are normalized to obtain the reference priority of each contradiction dimension; Within the preset time period of the physical examination data to be processed, obtaining the information richness of each contradiction dimension based on the data periodicity characteristics and data fluctuation characteristics of each reference user in the reference physical examination data set in each contradiction dimension; Among them, the method for obtaining the information richness is: within the preset time period in which the target user's to-be-processed physical examination data corresponds to the physical examination; calculating the average value of the frequency of use of the health check-up cabin by all reference users to obtain a first characteristic coefficient; according to the degree of fluctuation of the reference physical examination data of each contradiction dimension in the reference physical examination data set of each reference user within the preset time period, obtaining the second characteristic coefficient of each contradiction dimension; performing periodic characteristic analysis on all reference physical examination data of each contradiction dimension at each time node within the preset time period to obtain the third characteristic coefficient of each contradiction dimension; and taking the normalized result of the product of the first characteristic coefficient, the second characteristic coefficient and the third characteristic coefficient as the information richness of each contradiction dimension; Each conflict dimension of the target user is screened in combination with the reference priority and information richness to obtain a key data dimension; and an electronic medical record of the target user is generated based on the to-be-processed physical examination data of the key data dimension.

2. The method for generating user electronic medical records in a health check-up room according to claim 1, characterized in that: The step of dividing the reference users into deviated users and uniform users according to the data deviation of each reference user in each data dimension in the reference physical examination data set specifically includes: Obtain the data deviation factor of each reference user in each contradictory dimension, record the reference users whose data deviation factor is greater than the preset deviation threshold as deviant users, and record the reference users whose data deviation factor is less than or equal to the preset deviation threshold as uniform users.

3. The method for generating electronic medical records of users in a health check-up room according to claim 2, characterized in that: The future deviation factor of the unified user in each contradiction dimension is obtained based on the data change trend of the unified user in each contradiction dimension and the time distribution of the unified user's physical examination, combined with the data prediction results in each contradiction dimension, specifically including: Fitting the data of each unified user under each conflict dimension to obtain a data curve of each unified user under each conflict dimension; Calculating the average change trend of the data curves of all unified users under the same contradiction dimension to obtain the average change trend of each contradiction dimension; calculating the average time interval of physical examinations for all unified users under the same contradiction dimension to obtain the average physical examination time length of each contradiction dimension; using the product of the average change trend and the average physical examination time length to determine the predicted data of the unified users of each contradiction dimension; Based on the predicted data of the unified user of each contradiction dimension and the standard deviation and mean of the reference physical examination data of all unified users of the same contradiction dimension in the reference physical examination dataset, the future deviation factor of the unified user of each contradiction dimension is determined.

4. The method for generating user electronic medical records in a health check-up room according to claim 1, characterized in that: The obtaining of the second characteristic coefficient of each contradiction dimension according to the fluctuation degree of the reference physical examination data of each reference user in each contradiction dimension in the reference physical examination data set within a preset time period specifically includes: Obtain the mean of the reference physical examination data of all reference users at the same time node under each conflict dimension within a preset time period as the feature data of each conflict dimension at each time node; The variance of the characteristic data of each contradiction dimension at all time nodes within a preset time period is used as the second characteristic coefficient of each contradiction dimension.

5. The method for generating user electronic medical records in a health check-up room according to claim 4, characterized in that: The periodic characteristic analysis is performed based on all reference physical examination data of each contradiction dimension at each time node within the preset time period to obtain the third characteristic coefficient of each contradiction dimension, specifically including: Perform Fourier transform on the characteristic data of all time nodes of each contradiction dimension within a preset time period, and obtain the inverse of the amplitude corresponding to the maximum frequency component in the Fourier transform result of each contradiction dimension as the third characteristic coefficient of each contradiction dimension.

6. The method for generating user electronic medical records in a health check-up room according to claim 1, characterized in that: The step of screening each conflict dimension of the target user in combination with the reference priority and information richness to obtain key data dimensions specifically includes: Calculate the normalized result of the product of the reference priority and information richness of each contradiction dimension to obtain the feature criticality of each contradiction dimension; take the contradiction dimension corresponding to the feature criticality greater than or equal to the preset critical threshold as the key data dimension.

7. The method for generating user electronic medical records in a health check-up room according to claim 1, characterized in that: The method of using the current contradiction index to filter the data dimension to obtain the contradiction dimension specifically includes: The data dimension corresponding to the current contradiction index being greater than or equal to the preset abnormality threshold is taken as the contradiction dimension.

Images