Diagnosis strategy analysis method and system based on waiting time estimation

Abstract

The invention relates to the field of data processing, in particular to a doctor-seeing strategy analysis method and system based on waiting time estimation, and aims to obtain a target department to be reserved by a user by analyzing symptom description of the user. Meanwhile, the time required by the user to go to each hospital is analyzed according to the required time period of the user, and the travel time and the waiting time in the hospital are considered at the same time. According to the method, the crowding degree and the average waiting time of each hospital in the specific time period are known, and the patient can be guided to select a relatively loose time period for appointment or registration, so that the waiting time before actual treatment is shortened, and the medical treatment efficiency is improved. The analysis can be used as an important decision-making reference to help to select a medical institution which is most suitable for the current situation for chronic disease patients who need to see a doctor frequently or the situation that the patients need to quickly arrive at the hospital in an emergency.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and in particular to a method and system for analyzing medical consultation strategies based on waiting time estimation. Background Art

[0002] With the popularity of online appointments, more and more people can make appointments online conveniently and quickly. For users who want to see a doctor quickly, they usually choose a hospital that is closer to them for appointments. However, the degree of congestion in different hospitals varies at different times. Therefore, unless it is an emergency, users may still have to wait for a long time before they can see a doctor after arriving at the hospital. Summary of the invention

[0003] In view of this, an object of the present invention is to provide a method and system for analyzing medical consultation strategies based on waiting time estimation to solve the problems in the background technology.

[0004] In order to achieve the above object, the present invention adopts the following technical solutions: A method for analyzing a medical consultation strategy based on waiting time estimation of the present invention comprises the following steps: Obtaining the current user's medical consultation demand information and the user's current location information, wherein the medical consultation demand information includes a medical consultation time period and a symptom description; Extracting symptom entities from the symptom description, and determining the target department with the highest probability for the current user based on the symptom entities and a symptom-department relationship data table, wherein the symptom-department relationship data table represents the corresponding probabilities between multiple symptom types and departments; Filter candidate hospitals based on the target department and the current location information; Sampling the consultation time period to obtain multiple candidate time points; using the multiple candidate time points as departure time points, estimating the estimated arrival time of the current user at the multiple candidate hospitals, and determining the estimated waiting time of the target departments of the multiple candidate hospitals based on the estimated arrival time and the pre-constructed waiting time relationship curves of the multiple hospitals; A medical consultation decision is made based on the multiple candidate time points, the estimated arrival times of the multiple candidate hospitals, and the estimated waiting times of the target departments of the multiple candidate hospitals.

[0005] In one embodiment of the present application, the method for constructing the symptom-department relationship data table includes: Obtain triage records; Extracting a plurality of triage sample data from the triage record, wherein the triage sample data includes a symptom description sample and a triage result; Extracting a symptom entity sample from the symptom description sample, wherein the symptom entity sample includes a body part entity and a disease entity; Vectorizing the symptom entity sample to obtain a symptom entity sample vector; and mixing the symptom entity sample vector into a sample vector with a label for clustering to obtain a plurality of clusters, wherein the label is used to indicate the symptom type; Using the labels of the sample vectors in the same cluster as the labels of the symptom entity sample vectors to obtain a plurality of labeled triage sample data; Divide the labeled multiple triage sample data based on the labels to obtain multiple data units; Calculate the number of triage results in each data unit Probability , ,in, For triage results The number of corresponding triage sample data, is the total number of triage sample data for the data unit, is the serial number of the triage result; A symptom-department relationship data table is constructed based on multiple labels and the probabilities of triage results corresponding to the multiple labels.

[0006] In one embodiment of the present application, determining the target department with the highest probability for the current user based on the symptom entity and the symptom-department relationship data table includes: Labeling the symptom entity to obtain a label of the symptom entity; Match the label of the symptom entity with the symptom-department relationship data table to obtain the probability of multiple triage results ; When there is only one symptom entity, the department in the triage result with the highest probability is extracted as the target department; when there are multiple symptom entities, the total probability of each triage result is calculated, and the department in the triage result with the highest total probability is taken as the target department.

[0007] In one embodiment of the present application, a method for constructing a waiting time relationship curve includes: Obtain multiple historical medical records from multiple departments of multiple hospitals, where the historical medical records include the registration time point and time of visit ; Based on the registration time point of each historical medical record and time of visit Calculate the waiting time for each visit ; Calculate the first average waiting time of multiple historical consultation records within each natural day ; Based on the first average waiting time Cluster multiple historical medical records based on the registration time point to obtain multiple sample clusters; The sample cluster whose number of records is greater than a preset number threshold is taken as the target sample cluster, and the regular time period corresponding to the target sample cluster is extracted; Divide each natural day into a plurality of sub-time periods, and divide the historical medical records in the target sample cluster into a plurality of sub-time periods based on the registration time points, to obtain a plurality of record units; Calculate the second average waiting time of multiple historical medical records in each record unit and standard deviation ; Based on the second average waiting time and standard deviation Construct a reference range of waiting time for the corresponding sub-time period; A waiting time relationship curve of a regular time period is constructed based on the waiting time reference ranges of multiple sub-time periods.

[0008] In one embodiment of the present application, the waiting time relationship curve includes a minimum reference curve, a typical reference curve and a maximum reference curve, and the waiting time relationship curve of the regular time period is constructed based on the waiting time reference ranges of multiple sub-time periods, including: Mapping the waiting time reference ranges of the multiple sub-time periods into a two-dimensional coordinate system to obtain a maximum waiting time line segment, an average waiting time line segment, and a minimum waiting time line segment of each sub-time period, wherein the horizontal axis of the two-dimensional coordinate system is the time axis, and the vertical axis of the two-dimensional coordinate system is the waiting time axis; Take the midpoint of the maximum waiting time segment, the midpoint of the average waiting time segment, and the midpoint of the minimum waiting time segment respectively to obtain the minimum value point, typical value point, and maximum value point of each sub-time period; The minimum value points, typical value points and maximum value points corresponding to multiple sub-time periods are fitted respectively to obtain a minimum reference curve, a typical reference curve and a maximum reference curve.

[0009] In one embodiment of the present application, the estimated waiting time of target departments of multiple candidate hospitals is determined based on the estimated arrival time and the pre-constructed waiting time relationship curves of multiple hospitals, including: Determine the waiting time relationship curve of the target department of each hospital in regular time periods based on the estimated arrival time; Map the estimated arrival time to the two-dimensional coordinate system to obtain the intersection of the estimated arrival time with the minimum reference curve, the typical reference curve and the maximum reference curve to obtain the minimum reference duration. , Typical reference duration and the maximum reference duration .

[0010] In one embodiment of the present application, a medical consultation decision is made based on the multiple candidate time points, the estimated arrival time of the multiple candidate hospitals, and the estimated waiting time of the target departments of the multiple candidate hospitals, including: Construct candidate plans by starting from any candidate time point and reaching the target department of any candidate hospital; Calculate the time cost of each candidate solution and risk costs ; Based on the time cost and the risk cost Calculate the evaluation value of each candidate solution ,in, , is the serial number of the candidate solution; The multiple candidate solutions with the highest evaluation values ​​are used as target solutions, and the target solutions are returned to the current user.

[0011] In one embodiment of the present application, the time cost of each candidate solution is calculated. and risk costs ,include: Calculate the total reference duration for each candidate solution, which is the estimated arrival time and the typical reference time. The estimated arrival time is the difference between the estimated arrival time and the candidate time point; Calculate the waiting time floating ratio for each candidate solution , ; Normalize the total reference duration of multiple candidate solutions to obtain the time cost of multiple candidate solutions ; Floating ratio of waiting time for multiple candidate solutions Perform normalization to obtain the risk costs of multiple candidate solutions .

[0012] In one embodiment of the present application, the multiple candidate time points are used as departure time points, and the estimated arrival time of the current user at the multiple candidate hospitals is estimated, including: Call the external navigation API interface, use multiple candidate time points as departure time points, the user's current location information as the starting point, and multiple candidate hospitals as the end point for navigation planning, and obtain the estimated arrival time of multiple travel modes for multiple options, where each option includes a candidate time point and a candidate hospital; When the travel mode selection information from the current user is received, the estimated arrival time of the current user at multiple candidate hospitals when the current user departs at multiple candidate time points is obtained.

[0013] The present application also provides a medical consultation strategy analysis system based on waiting time estimation, including: An acquisition module is used to acquire the current user's medical consultation demand information and the user's current location information, wherein the medical consultation demand information includes a medical consultation time period and a symptom description; A department determination module, used to extract symptom entities from the symptom description, and determine the target department with the highest probability for the current user based on the symptom entities and a symptom-department relationship data table, wherein the symptom-department relationship data table represents the corresponding probabilities between multiple symptom types and departments; A preliminary screening module, used to screen candidate hospitals based on the target department and the current location information; A duration estimation module is used to sample the consultation time period to obtain multiple candidate time points; using the multiple candidate time points as departure time points, estimating the estimated arrival time of the current user at multiple candidate hospitals, and determining the estimated waiting time of the target departments of the multiple candidate hospitals based on the estimated arrival time and the pre-constructed waiting time relationship curves of the multiple hospitals; The decision-making module is used to make a medical consultation decision based on the multiple candidate time points, the estimated arrival time of the multiple candidate hospitals, and the estimated waiting time of the target departments of the multiple candidate hospitals.

[0014] The beneficial effects of the present invention are as follows: a method and system for analyzing a medical strategy based on waiting time estimation of the present invention, by parsing the user's symptom description, obtains the target department that the user needs to make an appointment. At the same time, the time required for the user to go to each hospital is analyzed according to the user's required time period, while taking into account the travel time and the waiting time in the hospital. The present application understands the degree of congestion and the average waiting time of each hospital in a specific time period, and can guide patients to choose a more relaxed time period for appointment or registration, thereby shortening the waiting time before the actual medical treatment and improving the efficiency of medical treatment. For chronic patients who need frequent medical treatment or who need to reach the hospital quickly in an emergency, this type of analysis can serve as an important decision-making reference to help select the most suitable medical institution for the current situation. In addition, from a macro perspective, if a large number of users can make more reasonable medical choices based on such analysis, the resource utilization efficiency of the entire medical system will be improved, such as reducing the pressure on popular hospitals and balancing the patient flow between different medical institutions. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The present invention will be further described below in conjunction with the accompanying drawings and embodiments: Figure 1is a flow chart of a method for analyzing a medical consultation strategy based on waiting time estimation shown in an embodiment of the present application; Figure 2 This is an application scenario diagram of a method for analyzing a medical consultation strategy based on waiting time estimation shown in an embodiment of the present application; Figure 3 A schematic diagram of a waiting time relationship curve in an embodiment of the present application; Figure 4 A schematic diagram of the predicted waiting time in an embodiment of the present application; Figure 5 It is a structural diagram of a medical consultation strategy analysis system based on waiting time estimation shown in one embodiment of the present application. DETAILED DESCRIPTION

[0016] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.

[0017] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and thus the drawings only show the layers related to the present invention rather than being drawn according to the number, shape and size of the layers in actual implementation. In actual implementation, the type, quantity and proportion of each layer may be changed arbitrarily, and the layer layout may also be more complicated.

[0018] In the following description, numerous details are discussed to provide a more thorough explanation of embodiments of the present invention; however, it is apparent to one skilled in the art that embodiments of the present invention may be practiced without these specific details.

[0019] The relevant information involved in this application has been obtained with full consent and authorization, and the collection, use and processing of relevant information must comply with the relevant laws, regulations and standards of relevant countries and regions.

[0020] Figure 1 is a flow chart of a method for analyzing a medical consultation strategy based on waiting time estimation shown in an embodiment of the present application, such as Figure 1 As shown, a method for analyzing a medical consultation strategy based on waiting time estimation in this embodiment may include the following steps: S110, obtaining the current user's medical consultation demand information and the user's current location information, wherein the medical consultation demand information includes a medical consultation time period and a symptom description; Figure 2 This is an application scenario diagram of a medical treatment strategy analysis method based on waiting time estimation shown in an embodiment of the present application. As Figure 2 shown, in this embodiment, the user accesses the server through terminal devices 220 such as mobile terminals, computer PCs, and tablet computers, so as to upload the current user information, the current user's medical treatment demand information, and the user's current location information to the server 210. The server 210 automatically performs appointment registration recommendation and travel plan planning decision according to the user's medical treatment demand information and the user's current location information. Thus, it helps the user to quickly make a medical treatment decision.

[0021] S120, extract symptom entities from the symptom description, and determine the target department with the highest probability for the current user based on the symptom entities and the symptom - department relationship data table, where the symptom - department relationship data table represents the corresponding probabilities between multiple symptom types and departments; In the present application, the symptom entities are extracted by the matching method. First, a dictionary containing various possible symptom descriptions needs to be constructed. For example, symptoms of common diseases (such as headache, fever), medical terms (such as chest pain, dyspnea), and common expressions of patients (such as "headache", "out of breath"). Then, necessary pre - processing is performed on the input free text to improve the accuracy of matching. It includes: Word segmentation: splitting the text into words or phrases. Removing stop words: removing those common words that do not provide useful information in the matching process (such as "of", "is"). Lemmatization: converting the vocabulary to its basic form (for example, "headache" and "cephalalgia" are uniformly processed). Finally, the pre - processed text is matched with the constructed symptom dictionary, for example, exact matching or fuzzy matching, so as to extract the symptom entities.

[0022] In the present application, the symptom - department relationship data table is also obtained through big data analysis of triage records, so as to determine the triage results (such as the target department) corresponding to one or more extracted symptom entities.

[0023] The construction process of the symptom - department relationship data table includes: (1) Obtain triage records; triage records are usually generated by the emergency department of a hospital or a dedicated triage system and stored in the hospital's information management system, such as the hospital information system (HIS) or the electronic health record (EHR) system. After obtaining authorization, the triage records can be extracted from the hospital information system (HIS) or the electronic health record (EHR) system.

[0024] In this embodiment, when extracting triage records, it is also necessary to remove privacy data to avoid the leakage of patients' privacy data.

[0025] (2) extracting a plurality of triage sample data from the triage record, wherein the triage sample data includes a symptom description sample and a triage result; When extracting, the raw data needs to be cleaned of errors, missing values, and other inconsistencies. In addition, the format needs to be standardized, such as unifying the date format, symptom description, etc.

[0026] (3) extracting a symptom entity sample from the symptom description sample, wherein the symptom entity sample includes a body part entity and a disease entity; The extraction of symptom entities can be referred to in the previous article and will not be repeated here.

[0027] (4) vectorizing the symptom entity sample to obtain a symptom entity sample vector; and mixing the symptom entity sample vector into a sample vector with a label for clustering to obtain a plurality of clusters, wherein the label is used to represent the symptom type; In this application, word2vec word embedding technology is used to vectorize symptom entity samples. The vectorized symptom entity samples can be quickly labeled.

[0028] In this application, clustering is used to achieve fast labeling. By introducing labeled sample vectors, clustering algorithms can be guided to more accurately identify similarity patterns, especially when dealing with complex or fuzzy boundary data sets. At the same time, using existing label information to assist clustering can quickly classify a large number of unlabeled symptom entities without the need for complete manual labeling, thereby greatly improving labeling efficiency.

[0029] Specifically, the present application uses the type of symptoms as labels to classify a wide variety of symptoms, such as pain, fatigue, fever, digestive system symptoms, etc. Combined with body parts, the computer can quickly identify the symptom type corresponding to the current user's symptom description.

[0030] (5) Using the labels of the sample vectors in the same cluster as the labels of the symptom entity sample vectors to obtain a plurality of labeled triage sample data; Since similar symptom entity vectors are clustered together during clustering, the labels within the same cluster can be directly used as labels for the symptom entity sample vectors, thereby achieving fast clustering.

[0031] (6) Divide the labeled multiple triage sample data based on the labels to obtain multiple data units; The divided data units contain multiple triage results corresponding to each symptom type.

[0032] For example, the symptom type is "headache". If the headache is severe and comes on suddenly, you may need to consider neurology or the emergency department, especially if it is accompanied by symptoms such as changes in consciousness and blurred vision. Therefore, the triage result includes neurology or the emergency department. If the headache is accompanied by nasal congestion or runny nose, it may be caused by a cold or allergies, and it is recommended to go to the ENT department. Therefore, the triage result includes the ENT department. If the headache is frequent and related to stress or emotions, you may need to consult a psychologist or mental health department. Therefore, the triage result includes a psychologist or mental health department.

[0033] (7) Calculate the number of triage results in each data unit. Probability , ,in, For triage results The number of corresponding triage sample data, is the total number of triage sample data for the data unit, is the serial number of the triage result; In this embodiment, since the same symptom type may correspond to multiple triage results, this application uses the department as the triage result to calculate the probability of the recommended department corresponding to the symptom type, that is, the probability .

[0034] (8) Construct a symptom-department relationship data table based on multiple labels and the probabilities of triage results corresponding to the multiple labels.

[0035] The symptom-department relationship data table contains multiple symptom labels and the probabilities of recommended departments for multiple symptom labels. Based on this table, the process of making department appointment recommendations is as follows: S121, labeling the symptom entity to obtain a label of the symptom entity; Symptom entities can also be clustered by mixing them into labeled sample vectors, thereby achieving fast labeling.

[0036] S122, matching the label of the symptom entity with the symptom-department relationship data table to obtain the probability of multiple triage results ; S123, when there is only one symptom entity, extract the department in the triage result with the highest probability as the target department; when there are multiple symptom entities, calculate the total probability of each triage result, and take the department in the triage result with the highest total probability as the target department.

[0037] Since there may be one or more symptom entities in the symptom description, there are multiple probability values ​​for multiple triage results. For example, "headache" and "fatigue" appear in the symptom description at the same time; The triage results for “headache” included: neurology: 15%, emergency department: 16%, otolaryngology: 21%; The triage results corresponding to “fatigue” include: emergency department: 13%, internal medicine: 17%, cardiovascular department: 5%; The corresponding results are: Neurology: 15%, Emergency Department: 29%, Otolaryngology: 21%, Internal Medicine: 17%, and Cardiology: 5%.

[0038] The final recommendation was "Emergency Department".

[0039] S130, screening candidate hospitals based on the target department and the current location information; The candidate hospitals are hospitals within a certain distance from the current user. Since the purpose of this embodiment is to improve the speed of medical treatment, hospitals within 10-20 kilometers of the current location are selected as candidate hospitals for screening.

[0040] S140, sampling the consultation time period to obtain multiple candidate time points; using the multiple candidate time points as departure time points, estimating the estimated arrival time of the current user at multiple candidate hospitals, and determining the estimated waiting time of target departments of the multiple candidate hospitals based on the estimated arrival time and a pre-constructed waiting time relationship curve of the multiple hospitals; Since the corresponding travel time and waiting time may be different when using different departure times, in order to find the best travel time, the medical consultation time points are sampled first. Specifically, sampling is performed every 30 minutes to obtain multiple candidate time points.

[0041] Then this application estimates the arrival time by calling external third-party API interfaces, including: Call the external navigation API interface, use multiple candidate time points as departure time points, the user's current location information as the starting point, and multiple candidate hospitals as the end point for navigation planning, and obtain the estimated arrival time of multiple travel modes for multiple options, where each option includes a candidate time point and a candidate hospital; After obtaining the estimated arrival time of multiple travel modes for multiple options, in order to determine which travel mode the user will use (such as driving, taxi, public transportation, walking, etc.), multiple options need to be sent to the current user for confirmation. For example, "Please select your travel mode, Hospital A: driving, taxi, public transportation, walking; Hospital B: driving, taxi, public transportation, walking..." When the travel mode selection information from the current user is received, the estimated arrival time of the current user at multiple candidate hospitals when the current user departs at multiple candidate time points is obtained.

[0042] Finally, the mode of transportation to each hospital is determined based on the user's choice, and the estimated arrival time of the current user at multiple candidate hospitals is obtained when the user departs at multiple candidate time points.

[0043] Since the waiting time may be different when arriving at the target department of the hospital at different times, the application determines the possible waiting time by pre-constructing the waiting time relationship curves of multiple hospitals for different estimated arrival times.

[0044] The process of constructing the waiting time relationship curve includes: (1) Obtain multiple historical medical records from multiple departments of multiple hospitals, where the historical medical records include the registration time point and time of visit ; Among them, historical medical records can also be extracted from the hospital information system (HIS) or electronic health record (EHR) system. The historical medical records record the various time points when the patient registers online, registers at the hospital, and then enters the clinic for treatment. Since this application needs to extract the waiting time, it mainly extracts the registration time point and time of visit .

[0045] (2) Based on the registration time of each historical medical record and time of visit Calculate the waiting time for each visit ;

[0046] (3) Calculate the first average waiting time of multiple historical medical records within each natural day ; The waiting time for a doctor's appointment varies with different time periods of the day and different periods of the year. For example, the waiting time in the morning is shorter, while it is relatively longer at other times. Flu season and summer can cause the waiting time in some departments to be significantly longer. Therefore, in this application, we first calculate the first average waiting time of multiple historical doctor's appointment records in each natural day. , to find the changes in waiting time brought by different seasons.

[0047] (4) Based on the first average waiting time Cluster multiple historical medical records based on the registration time point to obtain multiple sample clusters; (5) taking a sample cluster whose number of records is greater than a preset number threshold as a target sample cluster, and extracting a regular time period corresponding to the target sample cluster; Through clustering, historical medical records with similar average waiting time and registration time points can be grouped together.

[0048] This application sets a smaller maximum distance within a cluster, so there may be some sample clusters with smaller data volumes. These sample clusters may be caused by unexpected collective injuries and other phenomena. In order to find the rules under normal circumstances, this application removes these clusters with smaller data volumes and only retains sample clusters whose number of records within the cluster is greater than the preset number threshold as target sample clusters.

[0049] Each target sample cluster represents a regular time period, and the average waiting time in the regular time period is similar.

[0050] (6) Dividing each natural day into a plurality of sub-time periods, and dividing the historical medical records in the target sample cluster into a plurality of sub-time periods based on the registration time point, thereby obtaining a plurality of record units; After finding the changes in the annual cycle, the changes in the corresponding natural daily cycle within each target sample cluster are analyzed.

[0051] First, the working time is divided into multiple sub-time periods. In this embodiment, since the sampling at the time point is performed once every 30 minutes in the previous text, 30 minutes is also used as the duration of the sub-time period here.

[0052] Then the historical medical records of the target sample cluster are divided into corresponding sub-time periods to complete the second division.

[0053] (7) Calculate the second average waiting time of multiple historical medical records in each record unit and standard deviation ; (8) Based on the second average waiting time and standard deviation Construct a reference range of waiting time for the corresponding sub-time period; Since the average daily waiting time and registration time in the same target sample cluster are similar during clustering, the daily fluctuations of the consultation records in the target sample cluster are also roughly similar. The waiting time gradually increases from the morning and gradually decreases before the end of the working day. Therefore, by calculating the second average waiting time of multiple historical consultation records in each record unit and standard deviation A reference range of three times the standard deviation was constructed.

[0054] (9) Construct a waiting time relationship curve for a regular time period based on the waiting time reference ranges of multiple sub-time periods.

[0055] After obtaining the reference range that reflects the regular changes, a fluctuation curve is constructed based on multiple reference ranges, including: (9-1) Mapping the waiting time reference ranges of the multiple sub-time periods into a two-dimensional coordinate system to obtain a maximum waiting time line segment, an average waiting time line segment, and a minimum waiting time line segment of each sub-time period, wherein the horizontal axis of the two-dimensional coordinate system is the time axis, and the vertical axis of the two-dimensional coordinate system is the waiting time axis; (9-2) Take the midpoint of the maximum waiting time segment, the midpoint of the average waiting time segment, and the midpoint of the minimum waiting time segment respectively to obtain the minimum value point, typical value point, and maximum value point of each sub-time period; (9-3) The minimum value points, typical value points and maximum value points corresponding to multiple sub-time periods are fitted respectively to obtain the minimum reference curve, typical reference curve and maximum reference curve.

[0056] Figure 3 is a schematic diagram of a waiting time relationship curve in an embodiment of the present application. The waiting time relationship curve finally constructed is as follows: Figure 3 shown.

[0057] Figure 4 Schematic diagram of the predicted waiting time in an embodiment of the present application. Figure 4 As shown in the figure, after obtaining the curve, the estimated arrival time of the current user can be used to predict the approximate waiting time, including: S141, determining a waiting time relationship curve of a target department in each hospital in a regular time period based on the estimated arrival time; S142, mapping the estimated arrival time to the two-dimensional coordinate system, obtaining the intersection of the estimated arrival time with the minimum reference curve, the typical reference curve and the maximum reference curve, and obtaining the minimum reference duration , Typical reference duration and the maximum reference duration .

[0058] When the estimated arrival time is mapped to the two-dimensional coordinate system, a vertical straight line can be generated, and the intersection of this straight line with the minimum reference curve, the typical reference curve and the maximum reference curve can be used as the prediction point. , Typical reference duration and the maximum reference duration .

[0059] S150, making a medical consultation decision based on the multiple candidate time points, the estimated arrival times of the multiple candidate hospitals, and the estimated waiting times of the target departments of the multiple candidate hospitals.

[0060] After obtaining the travel time and estimated waiting time to multiple hospitals at multiple candidate departure time points, this application makes decisions based on time cost and risk cost, including: S151, constructing a candidate plan by starting from any candidate time point and reaching a target department of any candidate hospital; For example, candidate plan 1 departs at 7:40 and is expected to arrive at the hospital in 1 hour. The estimated minimum reference waiting time is 10 minutes, the estimated typical reference waiting time is 13 minutes, and the estimated maximum reference waiting time is 18 minutes.

[0061] S152, calculate the time cost of each candidate solution and risk costs ; Risk cost and maximum reference time Related, if there is a large maximum reference duration , it means that the user may have to wait for a long time, so the risk cost is high.

[0062] Time cost in this application and risk costs The calculation is done as follows: S1521, calculating the total reference duration of each candidate solution, where the total reference duration is the estimated arrival duration and the typical reference duration The estimated arrival time is the difference between the estimated arrival time and the candidate time point; S1522, calculate the floating ratio of the waiting time of each candidate solution , ; From the above formula, we can see that the risk cost and typical reference time in this application are and the maximum reference duration Relevant and typical reference duration and the maximum reference duration The greater the deviation, the higher the risk cost. The time cost is related to the total duration predicted in the previous article.

[0063] S1523, normalizing the total reference durations of multiple candidate solutions to obtain the time costs of the multiple candidate solutions ; Floating ratio of waiting time for multiple candidate solutions Perform normalization to obtain the risk costs of multiple candidate solutions .

[0064] Since we only need to compare different solutions, we use normalization to compare the advantages and disadvantages of different solutions.

[0065] Normalization is performed using the min-max normalization algorithm.

[0066] S153, based on the time cost and the risk cost Calculate the evaluation value of each candidate solution ,in, , is the serial number of the candidate solution; In general, the importance of time cost is greater than risk cost. In this embodiment, , .

[0067] S154, taking the multiple candidate solutions with the highest evaluation values ​​as target solutions, and returning the target solutions to the current user.

[0068] Finally, the 3-5 solutions with the highest evaluation values ​​are selected and returned to the target user as a reference. If the target user selects one of the target solutions as the implementation plan, the user can directly refer to the implementation plan to make an online appointment, travel, and wait. This application can greatly facilitate users' medical travel and save users a lot of time by minimizing waiting time.

[0069] The present invention provides a medical strategy analysis method based on waiting time estimation, which parses the user's symptom description to obtain the target department that the user needs to make an appointment with. At the same time, the time required for the user to go to each hospital is analyzed according to the user's required time period, while taking into account the travel time and the waiting time in the hospital. The present application understands the degree of congestion and average waiting time of each hospital in a specific time period, and can guide patients to choose a more relaxed time period for appointment or registration, thereby shortening the waiting time before the actual medical treatment and improving the efficiency of medical treatment. For chronic patients who need frequent medical treatment or who need to reach the hospital quickly in an emergency, this type of analysis can serve as an important decision-making reference to help choose the most suitable medical institution for the current situation. In addition, from a macro perspective, if a large number of users can make more reasonable medical choices based on this type of analysis, the resource utilization efficiency of the entire medical system will be improved, such as reducing the pressure on popular hospitals and balancing the patient flow between different medical institutions.

[0070] like Figure 5 As shown, the present application also provides a medical consultation strategy analysis system based on waiting time estimation, including: An acquisition module is used to acquire the current user's medical consultation demand information and the user's current location information, wherein the medical consultation demand information includes a medical consultation time period and a symptom description; A department determination module, used to extract symptom entities from the symptom description, and determine the target department with the highest probability for the current user based on the symptom entities and a symptom-department relationship data table, wherein the symptom-department relationship data table represents the corresponding probabilities between multiple symptom types and departments; A preliminary screening module, used to screen candidate hospitals based on the target department and the current location information; A duration estimation module is used to sample the consultation time period to obtain multiple candidate time points; using the multiple candidate time points as departure time points, estimating the estimated arrival time of the current user at multiple candidate hospitals, and determining the estimated waiting time of the target departments of the multiple candidate hospitals based on the estimated arrival time and the pre-constructed waiting time relationship curves of the multiple hospitals; The decision-making module is used to make a medical consultation decision based on the multiple candidate time points, the estimated arrival time of the multiple candidate hospitals, and the estimated waiting time of the target departments of the multiple candidate hospitals.

[0071] The present invention provides a method and system for analyzing a medical strategy based on waiting time estimation, which parses the user's symptom description to obtain the target department that the user needs to make an appointment with. At the same time, the time required for the user to go to each hospital is analyzed according to the user's required time period, while taking into account the travel time and the waiting time in the hospital. The present application understands the degree of congestion and average waiting time of each hospital in a specific time period, and can guide patients to choose a more relaxed time period for appointment or registration, thereby shortening the waiting time before the actual medical treatment and improving the efficiency of medical treatment. For patients with chronic diseases who need to see a doctor frequently or who need to reach the hospital quickly in an emergency, this type of analysis can serve as an important decision-making reference to help choose the most suitable medical institution for the current situation. In addition, from a macro perspective, if a large number of users can make more reasonable medical choices based on this type of analysis, the resource utilization efficiency of the entire medical system will be improved, such as reducing the pressure on popular hospitals and balancing the patient flow between different medical institutions.

[0072] This embodiment also provides an electronic terminal, including: a processor and a memory; The memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory, so that the terminal executes any one of the methods in this embodiment.

[0073] The computer-readable storage medium in this embodiment can be understood by ordinary technicians in this field: all or part of the steps of implementing the above-mentioned method embodiments can be completed by hardware related to the computer program. The aforementioned computer program can be stored in a computer-readable storage medium. When the program is executed, the steps of the above-mentioned method embodiments are executed; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk and other media that can store program codes.

[0074] The electronic terminal provided in this embodiment includes a processor, a memory, a transceiver and a communication interface. The memory and the communication interface are connected to the processor and the transceiver and complete communication with each other. The memory is used to store computer programs, the communication interface is used to communicate, and the processor and the transceiver are used to run computer programs so that the electronic terminal executes each step of the above method.

[0075] In this embodiment, the memory may include a random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0076] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0077] In the above-mentioned embodiments, although the present invention has been described in conjunction with the specific embodiments of the present invention, many replacements, modifications and variations of these embodiments will be apparent to those of ordinary skill in the art based on the foregoing description. The embodiments of the present invention are intended to cover all such replacements, modifications and variations falling within the broad scope of the appended claims.

[0078] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. A method for analyzing medical consultation strategies based on waiting time estimation, characterized in that: Includes steps: Obtaining the current user's medical consultation demand information and the user's current location information, wherein the medical consultation demand information includes a medical consultation time period and a symptom description; Extracting symptom entities from the symptom description, and determining the target department with the highest probability for the current user based on the symptom entities and a symptom-department relationship data table, wherein the symptom-department relationship data table represents the corresponding probabilities between multiple symptom types and departments; Filter candidate hospitals based on the target department and the current location information; Sampling the consultation time period to obtain multiple candidate time points; using the multiple candidate time points as departure time points, estimating the estimated arrival time of the current user at the multiple candidate hospitals, and determining the estimated waiting time of the target departments of the multiple candidate hospitals based on the estimated arrival time and the pre-constructed waiting time relationship curves of the multiple hospitals; A medical consultation decision is made based on the multiple candidate time points, the estimated arrival times of the multiple candidate hospitals, and the estimated waiting times of the target departments of the multiple candidate hospitals.

2. A method for analyzing medical consultation strategies based on waiting time estimation according to claim 1, characterized in that: The method for constructing the symptom-department relationship data table includes: Obtain triage records; Extracting a plurality of triage sample data from the triage record, wherein the triage sample data includes a symptom description sample and a triage result; Extracting a symptom entity sample from the symptom description sample, wherein the symptom entity sample includes a body part entity and a disease entity; Vectorizing the symptom entity sample to obtain a symptom entity sample vector; and mixing the symptom entity sample vector into a sample vector with a label for clustering to obtain a plurality of clusters, wherein the label is used to indicate the symptom type; Using the labels of the sample vectors in the same cluster as the labels of the symptom entity sample vectors to obtain a plurality of labeled triage sample data; Divide the labeled multiple triage sample data based on the labels to obtain multiple data units; Calculate the number of triage results in each data unit Probability , ,in, For triage results The number of corresponding triage sample data, is the total number of triage sample data for the data unit, is the serial number of the triage result; A symptom-department relationship data table is constructed based on multiple labels and the probabilities of triage results corresponding to the multiple labels.

3. A method for analyzing medical consultation strategies based on waiting time estimation according to claim 2, characterized in that: Determining the target department with the highest probability for the current user based on the symptom entity and the symptom-department relationship data table includes: Labeling the symptom entity to obtain a label of the symptom entity; Match the label of the symptom entity with the symptom-department relationship data table to obtain the probability of multiple triage results ; When there is only one symptom entity, the department in the triage result with the highest probability is extracted as the target department; when there are multiple symptom entities, the total probability of each triage result is calculated, and the department in the triage result with the highest total probability is taken as the target department.

4. The method for analyzing medical consultation strategies based on waiting time estimation according to claim 1, characterized in that: The method for constructing the waiting time relationship curve includes: Obtain multiple historical medical records from multiple departments of multiple hospitals, where the historical medical records include the registration time point and time of visit ; Based on the registration time point of each historical medical record and time of visit Calculate the waiting time for each visit ; Calculate the first average waiting time of multiple historical consultation records within each natural day ; Based on the first average waiting time Cluster multiple historical medical records based on the registration time point to obtain multiple sample clusters; The sample cluster whose number of records is greater than a preset number threshold is taken as the target sample cluster, and the regular time period corresponding to the target sample cluster is extracted; Divide each natural day into a plurality of sub-time periods, and divide the historical medical records in the target sample cluster into a plurality of sub-time periods based on the registration time points, to obtain a plurality of record units; Calculate the second average waiting time of multiple historical medical records in each record unit and standard deviation ; Based on the second average waiting time and standard deviation Construct a reference range of waiting time for the corresponding sub-time period; A waiting time relationship curve of a regular time period is constructed based on the waiting time reference ranges of multiple sub-time periods.

5. A method for analyzing medical consultation strategies based on waiting time estimation according to claim 4, characterized in that: The waiting time relationship curve includes a minimum reference curve, a typical reference curve and a maximum reference curve. The waiting time relationship curve of a regular time period is constructed based on the waiting time reference ranges of multiple sub-time periods, including: Mapping the waiting time reference ranges of the multiple sub-time periods into a two-dimensional coordinate system to obtain a maximum waiting time line segment, an average waiting time line segment, and a minimum waiting time line segment of each sub-time period, wherein the horizontal axis of the two-dimensional coordinate system is the time axis, and the vertical axis of the two-dimensional coordinate system is the waiting time axis; Take the midpoint of the maximum waiting time segment, the midpoint of the average waiting time segment, and the midpoint of the minimum waiting time segment respectively to obtain the minimum value point, typical value point, and maximum value point of each sub-time period; The minimum value points, typical value points and maximum value points corresponding to multiple sub-time periods are fitted respectively to obtain a minimum reference curve, a typical reference curve and a maximum reference curve.

6. A method for analyzing medical consultation strategies based on waiting time estimation according to claim 5, characterized in that: The estimated waiting time for target departments in multiple candidate hospitals is determined based on the estimated arrival time and the pre-built waiting time relationship curves of multiple hospitals, including: Determine the waiting time relationship curve of the target department of each hospital in regular time periods based on the estimated arrival time; Map the estimated arrival time to the two-dimensional coordinate system to obtain the intersection of the estimated arrival time with the minimum reference curve, the typical reference curve and the maximum reference curve to obtain the minimum reference duration. , Typical reference duration and the maximum reference duration .

7. A method for analyzing medical consultation strategies based on waiting time estimation according to claim 6, characterized in that: The medical consultation decision is made based on the multiple candidate time points, the estimated arrival time of the multiple candidate hospitals, and the estimated waiting time of the target departments of the multiple candidate hospitals, including: Construct candidate plans by starting from any candidate time point and reaching the target department of any candidate hospital; Calculate the time cost of each candidate solution and risk costs ; Based on the time cost and the risk cost Calculate the evaluation value of each candidate solution ,in, , is the serial number of the candidate solution; The multiple candidate solutions with the highest evaluation values ​​are used as target solutions, and the target solutions are returned to the current user.

8. The method for analyzing medical consultation strategies based on waiting time estimation according to claim 7, characterized in that: Calculate the time cost of each candidate solution and risk costs ,include: Calculate the total reference duration for each candidate solution, which is the estimated arrival time and the typical reference time. The estimated arrival time is the difference between the estimated arrival time and the candidate time point; Calculate the waiting time floating ratio for each candidate solution , ; Normalize the total reference duration of multiple candidate solutions to obtain the time cost of multiple candidate solutions ; Floating ratio of waiting time for multiple candidate solutions Perform normalization to obtain the risk costs of multiple candidate solutions .

9. The method for analyzing medical consultation strategies based on waiting time estimation according to claim 1, characterized in that: Taking the multiple candidate time points as departure time points, estimating the estimated arrival time of the current user at the multiple candidate hospitals, including: Call the external navigation API interface, use multiple candidate time points as departure time points, the user's current location information as the starting point, and multiple candidate hospitals as the end point for navigation planning, and obtain the estimated arrival time of multiple travel modes for multiple options, where each option includes a candidate time point and a candidate hospital; When the travel mode selection information from the current user is received, the estimated arrival time of the current user at multiple candidate hospitals when the current user departs at multiple candidate time points is obtained.

10. A medical consultation strategy analysis system based on waiting time estimation, characterized in that: include: An acquisition module is used to acquire the current user's medical consultation demand information and the user's current location information, wherein the medical consultation demand information includes a medical consultation time period and a symptom description; A department determination module, used to extract symptom entities from the symptom description, and determine the target department with the highest probability for the current user based on the symptom entities and a symptom-department relationship data table, wherein the symptom-department relationship data table represents the corresponding probabilities between multiple symptom types and departments; A preliminary screening module, used to screen candidate hospitals based on the target department and the current location information; A duration estimation module is used to sample the consultation time period to obtain multiple candidate time points; using the multiple candidate time points as departure time points, estimating the estimated arrival time of the current user at multiple candidate hospitals, and determining the estimated waiting time of the target departments of the multiple candidate hospitals based on the estimated arrival time and the pre-constructed waiting time relationship curves of the multiple hospitals; The decision-making module is used to make a medical consultation decision based on the multiple candidate time points, the estimated arrival time of the multiple candidate hospitals, and the estimated waiting time of the target departments of the multiple candidate hospitals.

Images