Hospital intelligent follow-up visit management method and system
Patent Information
- Application Number
- CN202510912174.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-03
- Publication Date
- 2025-10-17
Smart Images
Figure CN120809114A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application belongs to the technical field of intelligent medical treatment, and particularly relates to a hospital intelligent follow-up management method and system. BACKGROUND
[0002] Before the introduction of big data and computer technology, the follow-up work of the hospital mainly relied on manual operation and paper-based management, showing the characteristics of low efficiency, information lag, limited coverage, etc. At that time, the follow-up management completely depended on the personal experience of medical staff and manual recording, forming a primitive working mode with "paper and pen + telephone" as the core. Specifically, the doctor needs to manually flip through the paper medical record book and record the patient list and basic information that need to be followed up with a pen. The follow-up is mainly carried out through three ways: one is to ask the patient's condition orally and record it by hand during the outpatient review; two is that the nurse contacts the patient one by one through the fixed telephone and checks the options on the printed follow-up form; three is to collect patient feedback through the mailing of paper questionnaires. All the collected information is recorded in the paper follow-up book or the file bag, and these materials need to be manually sorted and archived and stored in the medical record room of the hospital. Due to the lack of systematic management, the follow-up records are often lost, the handwriting is illegible, and the data statistics are difficult, etc. In terms of follow-up content, due to the lack of intelligent analysis tools, the doctor can only judge the recovery of the patient according to personal experience. The monitoring of important indicators completely depends on the patient's self-conscious report or regular hospital examination, and it is difficult to find abnormal conditions in time. For the long-term management of chronic patients, it is even more difficult, and often only the passive mode of "contacting when having symptoms" can be achieved. The whole follow-up process is time-consuming and laborious, and a nurse can only complete the follow-up work of 20-30 patients a day, and it is easy to cause information omission or recording errors due to human factors. This traditional mode is not only inefficient, but also difficult to guarantee the quality of follow-up, and it is impossible to carry out large-scale data analysis and experience summary.
[0003] With the development of big data and artificial intelligence technologies, hospital intelligent follow-up management system is one of the core applications in the field of smart healthcare, aiming to achieve continuous monitoring and personalized intervention of patients' health status outside the hospital through artificial intelligence (AI), big data analysis and Internet of Things (IoT) technologies. Traditional follow-up relies on manual phone calls or outpatient reviews, which has low efficiency, insufficient coverage, and fragmented data. Modern intelligent follow-up systems must rely on these technologies to achieve innovation. In terms of artificial intelligence, the system uses natural language processing (NLP) to automatically analyze patients' complaints and electronic medical records (EMR), combined with machine learning models (such as random forest, LSTM time series network) to predict patient risk stratification and dynamically adjust follow-up strategies. For example, by analyzing the vital signs trends of postoperative patients, AI can early warn of infection or complication risks and trigger proactive follow-up. In addition, computer vision technology can process patients' uploaded wound photos to automatically assess healing progress and reduce human error. Big data technology provides the underlying support for the system. By integrating multi-source heterogeneous data (EMR, wearable devices, PROs questionnaires, medical insurance records, etc.), the system constructs a patient's whole life cycle health portrait. Based on distributed computing frameworks (such as Hadoop, Spark), the system analyzes massive data in real time to identify potential correlations (such as the correlation between a certain drug and recovery speed). Data lake technology enables long-term storage and retrospective analysis, supporting scientific research and clinical decision-making. For example, by mining historical data, it is found that the blood glucose fluctuations of diabetic patients are related to seasonal changes, and the follow-up frequency is optimized. In the smart healthcare ecosystem, the system connects intelligent hardware (such as blood pressure monitors, ECG patches) through the Internet of Things (IoT) to achieve automatic data collection and cloud synchronization. Combined with 5G and edge computing, critical values (such as heart attack signs) can be transmitted to doctors' terminals in real time, reducing response time. At the same time, the system is interconnected with hospital HIS and regional health platforms to form a'monitoring-analysis-intervention-feedback' closed loop.
[0004] Specific to intelligent follow-up, patients, especially COPD patients, need long-term self-management after discharge, and follow-up is the key link to link their in-hospital and out-of-hospital health management. However, the current follow-up mechanism does not meet the institutionalization, dynamization and standardization requirements, resulting in insufficient continuity of services for patients in hospitals, communities and families. A precise, dynamic and standardized intelligent follow-up service system for COPD patients is needed. There is a lack of COPD follow-up system with decision support function and intelligent follow-up content pushing according to patient conditions in the research published at home and abroad. Different behavior conversion strategies should be taken for individuals at different stages to promote their transition to the action and maintenance stages. COPD patients at different behavior stages should focus on different follow-up content during follow-up and timely follow-up intervention to promote the positive change of their behavior to the maintenance stage. However, how to use the above advanced technology, use rich server resources and big data resources, and provide follow-up strategies that are dynamically adapted to the current stage without increasing the computing burden of the user terminal itself, make accurate decisions when the follow-up stage and its follow-up strategy are fuzzy, and thus make timely and appropriate follow-up strategy conversion, is a problem to be solved. Based on the above problems, the present application sets up a main thread and a timing thread in the follow-up process based on the characteristics of the long follow-up time of the follow-up stage and the scattering of the follow-up time; the main thread is used to execute the follow-up strategy, and the timing thread is used to execute the wake-up operation of the main thread when the timing arrives, thereby reducing the software and hardware overhead of the follow-up strategy execution; the real-time decision of the follow-up stage and the instantiation of the follow-up process are made based on the patient parameters and the big data patient parameters, which can quickly adapt to the real-time changes of the patients and improve the efficiency of the follow-up service. SUMMARY
[0005] In order to solve the above problems in the prior art, the present application provides a hospital intelligent follow-up management method and system, the method comprising:
[0006] The follow-up process comprises a main thread and a timing thread; the main thread executes the follow-up strategy, and after completion, enters a sleep state, and returns to the follow-up strategy execution state in response to the wake-up of the timing thread; the timing thread wakes up the main thread when the timing arrives;
[0007] The first rule score and the first membership degree of determining that the patient belongs to the current follow-up stage are determined; the second rule score and the second membership degree of determining that the patient belongs to the target follow-up stage are determined; the current patient belongs to the current follow-up stage, the target follow-up stage, or is in a migration state based on the first rule score and the first membership degree, the second rule score and the second membership degree;
[0008] The patient parameter rule set and the feature combination set corresponding to the follow-up stage are obtained; the current patient parameters are matched with the parameter rule set to obtain a rule score; the current patient parameter sequence is compared with each feature combination in the feature combination set to determine the hit feature combination and the number of hits to determine the membership; the first rule score and the first membership are obtained when the follow-up stage is the current follow-up stage; the second rule score and the second membership are obtained when the follow-up stage is the target follow-up stage;
[0009] When the first rule score indicates that the patient is in the current follow-up stage and the first membership indicates that the indication is reliable and the second rule score indicates that the patient is not in the target follow-up stage and the second membership indicates that the indication is reliable, it is determined that the current patient is in the current follow-up stage; when the first rule score indicates that the patient is not in the current follow-up stage and the first membership indicates that the indication is reliable and the second rule score indicates that the patient is in the target follow-up stage and the second membership indicates that the indication is reliable, it is determined that the current patient is in the target follow-up stage; otherwise, it is indicated that the migration state;
[0010] When in the migration state, it is further analyzed whether to migrate the follow-up stage, and if yes, it is determined to enter the target follow-up stage, otherwise, the current follow-up stage is maintained.
[0011] Further, after the patient parameter is updated, the patient parameter is sent to the server through the patient parameter acquisition interface; the server determines whether the follow-up stage changes based on the updated patient parameter, and when the follow-up stage changes, the follow-up strategy corresponding to the changed follow-up stage is re-instantiated based on the updated patient parameter, and the re-instantiated follow-up strategy is sent to the client to replace the current follow-up strategy executed by the main thread in the client.
[0012] Further, when the patient is registered, a follow-up record data block corresponding to the server is created; the patient parameter is stored in time sequence; a patient parameter acquisition interface is created in the server, which is used for real-time acquisition of the patient parameter, and the patient parameter is stored in the follow-up record data block in time sequence or time interval sequence.
[0013] Further, the instantiation sets the main thread and the timing thread, specifically: determining the follow-up stage in which the patient is located, instantiating the follow-up strategy corresponding to the follow-up stage based on the patient parameter; the follow-up strategy corresponding to the current follow-up stage is taken as the current follow-up strategy executed in the main thread, and the follow-up strategy is loaded in the main thread; the timing time points are obtained from the instantiated follow-up strategy, arranged in time sequence to form a timing time sequence; the timing thread is instantiated based on the timing time sequence.
[0014] Further, the follow-up strategy comprises one or more follow-up events, the follow-up event comprises a trigger condition and a follow-up operation executor; when the timing point acquisition is performed, the timing point is determined based on the trigger condition; and when the timing arrives, the follow-up event corresponding to the timing point is woken up, and the main thread is woken up to execute the follow-up event.
[0015] Further, the follow-up operation executor comprises a follow-up operation, a follow-up instruction and / or a follow-up action set.
[0016] Further, when in the migration state, the migration state is analyzed in the follow-up stage based on the patient parameter sequence and the feature combination set in the recent preset time length range, and it is determined whether the migration in the follow-up stage is needed, if yes, it is determined that the current patient belongs to the target follow-up stage, otherwise, the current follow-up stage is kept.
[0017] Further, the first membership degree is determined based on the hit feature combination and the hit times, specifically, the first membership degree is determined based on the hit feature combination and the hit times, so that the more the hit feature combinations and / or the more the hit times are, the higher the first membership degree is, and vice versa.
[0018] A hospital intelligent follow-up management system, the hospital intelligent follow-up management system is used for realizing the hospital intelligent follow-up management method.
[0019] A hospital intelligent follow-up management platform, the hospital intelligent follow-up management platform is used for realizing the hospital intelligent follow-up management method.
[0020] The beneficial effects of the present application include:
[0021] Based on the feature that the long follow-up time of the follow-up stage occurs in a scattered manner, the follow-up process comprises a main thread and a timing thread; the main thread is used for executing the follow-up strategy, and the timing thread performs the wake-up operation of the main thread when the timing arrives, so that the software and hardware overhead of the follow-up strategy execution is reduced; based on the real-time decision of the patient parameter and the instantiation of the follow-up process, the real-time change of the patient can be quickly adapted, and the follow-up service efficiency is improved;
[0022] The big data patient parameter information is introduced, the credibility of the scoring rule is quantitatively described based on big data through the membership degree on the basis of the rule scoring, and the co-occurrence typicality and the time sequence typicality are considered at the same time during the quantitative description; meanwhile, the current and target aspects of the follow-up stage of the patient are simultaneously decided, only when both aspects have clear analysis results, the determined decision is made, and the accuracy of the follow-up stage analysis is improved;
[0023] The uncertain decision state is described by migration state, more decision information is introduced by introducing decision into next stage to form double-layer decision, and the possibility of obtaining discrimination result is improved; the discrimination of complex conditions such as retreat of follow-up stage and no direction conversion of follow-up stage is supported; the accuracy of follow-up strategy is improved by multi-level fine discrimination of follow-up stage. BRIEF DESCRIPTION OF DRAWINGS
[0024] The drawings described herein are used to provide further understanding of the present application, form part of the present application, and do not constitute improper limitation on the present application, in the drawings:
[0025] Figure 1 The hospital intelligent follow-up management method provided by the present application is shown in the figure.
[0026] Figure 2 The client acquisition interface corresponding to the server side for setting the client follow-up process provided by the present application is shown in the figure. DETAILED DESCRIPTION
[0027] The present application will be described in detail below in combination with the drawings and specific embodiments, wherein the illustrative embodiments and the description are only used to explain the present application, but do not constitute limitation on the present application.
[0028] The present application provides a hospital intelligent follow-up management method and system, as shown in the accompanying Figure 1 The method comprises the following steps:
[0029] Step S1: creating a patient follow-up process in a client; the follow-up process comprises a main thread and a timing thread; the main thread is used for executing a follow-up strategy, and enters a sleep state after execution processing, and returns a follow-up strategy execution state in response to a wake-up operation of the timing thread; the timing thread executes the wake-up operation of the main thread when timing arrives;
[0030] The patient follow-up process is created, specifically: a follow-up process is created in a mobile terminal when a patient is discharged, and the follow-up process comprises a main thread and a timing thread; the main thread is used for executing a follow-up event corresponding to a current time point in a follow-up strategy, and enters a sleep state after execution processing, and enters a follow-up strategy execution state in response to a wake-up operation of the timing thread; the timing thread executes the wake-up operation of the main thread when timing arrives based on a timing time point sequence, and the timing thread is a resident thread;
[0031] sending the patient parameters to the server through the patient parameter collection interface after the patient parameters are updated; determining whether a follow-up stage changes based on the updated patient parameters, and when the follow-up stage changes, re-instantiating a follow-up strategy corresponding to the changed follow-up stage based on the updated patient parameters, and sending the re-instantiated follow-up strategy to the client to replace the current follow-up strategy executed by the main thread in the client;
[0032] Preferably, a follow-up record data block corresponding to the server is created when the patient is registered; the patient parameters are stored in chronological order; a patient parameter collection interface is created in the server to collect the patient parameters in real time and store the patient parameters in the follow-up record data block in chronological order or time interval order;
[0033] Preferably, the initial value of the patient parameters includes the current follow-up stage of the patient determined by evaluation; the initial value is used to instantiate the main thread and the timing thread after the follow-up process is created in the client; the patient parameter collection interface is specifically connected to the EMR system to extract data in batches at regular intervals through the HL7 / FHIR standard interface; the wearable device is connected to collect real-time data and set regular push and / or time trigger to collect patient reported outcomes (PROs) data; wherein the EMR data is the electronic medical record of the hospital, which has a high degree of structure, and the parameters of the wearable device are continuous dynamic data that need to be processed after being obtained; the patient reported outcomes (PROs) are pain scores and quality of life questionnaires; the data can be obtained through the user terminal process by using the application program or web form; see the attached Figure 2 The client collection interface corresponding to the server is provided for the user terminal to collect and manage data; the main thread and the timing thread are instantiated, specifically, the follow-up stage of the patient is determined, and the follow-up strategy corresponding to the follow-up stage is instantiated based on the patient parameters; the follow-up strategy corresponding to the current follow-up stage is used as the current follow-up strategy executed in the main thread, and the follow-up strategy is loaded in the main thread; the timing time points are obtained from the instantiated follow-up strategy, arranged in chronological order to form a timing time sequence; the timing thread is instantiated based on the timing time sequence; when re-instantiating, the main thread and the timing thread are re-instantiated based on the updated patient parameters; the timing of re-instantiation is when the follow-up stage changes or the patient parameters change, etc.
[0034] Preferably, the follow-up strategy includes one or more follow-up events, and the follow-up event includes a trigger condition and a follow-up operation execution body; when the timing time points are obtained, the timing time points are determined based on the trigger condition; and when the timing arrives, the follow-up event corresponding to the timing time point is awakened, and the main thread is awakened to execute the follow-up event;
[0035] Preferably, the follow-up operation executor comprises a follow-up operation, a follow-up instruction and / or a follow-up action set;
[0036] Preferably, the patient parameters are collected in real time, the patient parameters are time-aligned according to time intervals in which the collection time is located, and the patient parameters are sorted and stored according to the time interval occurrence sequence to form a patient parameter sequence; wherein: the collection of the patient parameters needs to integrate EMR data, wearable devices and patient-reported outcomes; through the time interval-based alignment, patient parameters close in time are aligned in the same time interval; facilitating subsequent feature combination and comparison of discrimination combination;
[0037] Preferably, a parameter rule set associated with each follow-up stage is provided; the parameter rule set comprises one or more parameter rules, and if the current patient parameters and the parameter rule set associated with a follow-up stage are consistent, it is determined that the patient belongs to the follow-up stage; the consistency refers to that the rule score meets the requirement of the current follow-up stage;
[0038] The follow-up stages include an intentionless stage, an intention stage, a preparation stage, an action stage and a maintenance stage; wherein: the intentionless stage indicates no change intention within 6 months, which is manifested as denial of disease severity / lack of knowledge / attitude of resistance to behavior change; the preparation stage indicates action within 1 month, which is manifested as developing a medication plan / purchasing monitoring equipment / attending a health education course; the action stage indicates sustained behavior change <6 months, which is manifested as regular medication records / using a peak flow meter / abstinence from smoking <6 months / exercise log; the maintenance stage indicates sustained change >6 months, which is manifested as stable medication adherence >80% / continued abstinence / annual acute exacerbation frequency reduction;
[0039] Step S2: determining a first rule score and a first membership degree of the patient belonging to the current follow-up stage; determining a second rule score and a second membership degree of the patient belonging to the target follow-up stage; based on the first rule score and the first membership degree, the second rule score and the second membership degree, determining that the current patient belongs to the current follow-up stage, the target follow-up stage or is in a migration state;
[0040] Preferably, the server executes step S2 based on the real-time collected patient parameters; the target follow-up stage is one of the next stage of the current follow-up stage and a high-probability migration stage;
[0041] Alternative: the target follow-up stage is other follow-up stages except the current follow-up stage; in this way, the follow-up stage can be dealt with the complexity of the patient's follow-up stage regression, no direction conversion, etc.; for example: regarding each non-current follow-up stage as a target follow-up stage and performing rule scoring and membership calculation; the target follow-up stage can be set as the one with the highest second rule score and second membership value, at this time, only the decision between the current follow-up stage and the one with the highest second rule score and second membership value is needed; when the second rule score and the second membership value between non-current follow-up stages cannot be clearly distinguished, it can be pushed into the next decision stage, and when the current patient belongs to the current follow-up stage, the target follow-up stage, or is in the migration state based on the first rule score and the first membership, the second rule score and the second membership, the analysis is performed;
[0042] The first rule score and the first membership of the determination that the patient belongs to the current follow-up stage are specifically: obtaining a patient parameter rule set and a feature combination set corresponding to a follow-up stage; obtaining the current patient parameter, matching the current patient parameter and the parameter rule set to obtain a rule score; comparing the sequence of the current patient parameter in the recent preset time length range with each feature combination in the feature combination set to determine the hit feature combination and the number of hits; determining the membership based on the hit feature combination and the number of hits; when the follow-up stage is the current follow-up stage, the first rule score and the first membership are obtained; when the follow-up stage is the target follow-up stage, the second rule score and the second membership are obtained;
[0043] Preferably: when matching the patient parameter and the parameter rule set, each parameter rule in the parameter rule set needs to be matched one by one, and the rule score consistent with the matching result is obtained based on the score after each parameter rule is matched; the simplest comprehensive way is addition, weighted summation, average, etc.;
[0044] The current patient parameter sequence in the recent preset time length range and each feature combination in the feature combination set are compared to determine the hit feature combination and the number of hits; Specifically, the feature combination set associated with the follow-up stage is pre-constructed; The feature combinations in a feature combination set are obtained in turn, compared along the time sequence and the current patient parameter sequence, and it is determined whether there is a feature combination (similar) in the patient parameter sequence. If so, it is determined to be hit; The hit number of the feature combination is incremented after the hit; Repeat the step until all feature combinations are compared; For example: the patient parameter sequence contains <(A1, A2, A3), (B1, B2, B3) (C1, C2, C3) (B1, B2, C3)>; The patient parameter sequence contains 4 time interval patient parameter groups; For feature combination (B1, B2), the sequence hit number is 2;
[0045] The first membership degree is determined based on the hit feature combination and the hit number; Specifically, the first membership degree is determined based on the hit feature combination and the hit number, so that the more the hit feature combinations and / or the hit numbers are, the higher the first membership degree is; Otherwise, it is lower; The membership degree belongs to the [0, 1] numerical space;
[0046] Preferably, the correspondence between the feature combination, the hit number and the membership degree is set, and the first membership degree corresponding to the hit feature combination and the hit number is obtained by querying the correspondence;
[0047] The feature combination set associated with the follow-up stage is pre-constructed; Specifically, the patient parameter big data record is obtained; The typical co-occurring patient parameters in the follow-up stage with high rule score are constructed and the feature combinations associated with the follow-up stage are put into the feature combination set;
[0048] The discrimination combination set associated with the follow-up stage is pre-constructed; Specifically, the patient parameter big data record is obtained; The typical first-occurring patient parameters in the first time length before entering the follow-up stage with high rule score are constructed and the discrimination combinations associated with the follow-up stage are put into the discrimination combination set; The first time length is a preset length, which can be defined by the number of contained time intervals;
[0049] Preferably, all patient parameters in the same feature combination belong to one time interval; All patient parameters in the same discrimination combination do not belong to one time interval, and the two patient parameters in the discrimination combination that do not belong to one time interval are respectively located in the time intervals that occur in time; Therefore, the discrimination combination is based on time sequence;
[0050] Alternatively, the feature combination set associated with the pre-construction and follow-up stage; Specifically: obtain patient parameter big data records; Process each patient parameter big data record in turn; For each follow-up stage, obtain the patient parameters corresponding to when the rule score is greater than or equal to the high rule score threshold; Use the patient parameters that appear at the same time to form a patient parameter combination; If the same patient parameter combination frequently appears in the patient parameter big data records of multiple patients, and / or frequently appears in the patient parameter big data records of the same patient, then the patient parameter combination is a co-occurring typical patient parameter combination, called a feature combination, and it is put into the feature combination set; Wherein: one or more patient parameters in the feature combination are collected at the same time interval;
[0051] Alternatively, the discrimination combination set associated with the pre-construction and follow-up stage; Specifically: obtain patient parameter big data records; Process each patient parameter big data record in turn; For each follow-up stage, obtain the patient parameters before their rule scores are greater than or equal to the high rule score threshold; Use the patient parameters that appear in sequence to form a patient parameter combination; If the same patient parameter combination frequently appears in the patient parameter big data records of multiple patients, and / or frequently appears in the patient parameter big data records of the same patient, then the patient parameter combination is a time sequence typical patient parameter combination, called a discrimination combination, and it is put into the discrimination combination set; Wherein: one or more patient parameters in the discrimination combination are collected at the same or sequentially adjacent time intervals;
[0052] Based on the first rule score and the first membership degree, the second rule score and the second membership degree to determine that the current patient belongs to the current follow-up stage, the target follow-up stage, or is in a migration state; Specifically: when the first rule score indicates that the patient is in the current follow-up stage (the first rule score meets the requirements of the current follow-up stage) and the first membership degree indicates that the indication is reliable (the first membership degree is greater than or equal to the membership threshold), and the second rule score indicates that the patient is not in the target follow-up stage (the second rule score does not meet the requirements of the target follow-up stage) and the second membership degree indicates that the indication is reliable (the second membership degree is greater than or equal to the membership threshold), it is determined that the current patient is in the current follow-up stage; When the first rule score indicates that the patient is not in the current follow-up stage and the first membership degree indicates that the indication is reliable, and the second rule score indicates that the patient is in the target follow-up stage and the second membership degree indicates that the indication is reliable, it is determined that the current patient is in the target follow-up stage; Other cases indicate that the patient is in a migration state; The migration state is an ambiguous state, and it is not possible to provide a follow-up strategy suitable for the follow-up stage for the patient, and the migration state needs to be clearly discriminated;
[0053] Preferably, when there are multiple target follow-up stages, that is, when the second rule score and the second membership value between the non-current follow-up stages cannot be clearly distinguished, the determination of the analysis result that the current patient belongs to the current follow-up stage, the target follow-up stage, or is in the migration state is sequentially based on the first rule score and the first membership, the second rule score and the second membership for each target follow-up stage, and the final analysis result is further determined based on multiple analysis results corresponding to multiple target follow-up stages; in this case, when one analysis result indicates a clear follow-up stage and the other analysis results point to the migration state, the final analysis result can be determined as the clear follow-up stage; and when multiple analysis results point to the current follow-up stage, the final analysis result can be determined as the current follow-up stage; in this way, the decision is introduced into the next step, thereby introducing more decision information and improving the possibility of obtaining the analysis result;
[0054] Preferably, when in the migration state, the follow-up stage of the migration state is analyzed based on the patient parameter sequence and the analysis feature combination set in the preset time length range, and it is determined whether the follow-up stage migration is needed, if yes, it is determined that the current patient belongs to the target follow-up stage, otherwise, the current follow-up stage is maintained; specifically, an analysis combination set associated with the follow-up stage is constructed in advance; an analysis combination in the analysis combination set is sequentially obtained and compared with the current patient parameter sequence, and the hit number of the analysis combination is incremented each time a hit is made; the step is repeated until all analysis combinations are compared; if the coverage rate of the hit analysis combination to the analysis combination set is greater than the coverage rate threshold, and the hit number is greater than or equal to the hit number threshold, it is determined that the follow-up stage migration is needed.
[0055] Preferably, if the coverage rate of the hit analysis combination is greater than or equal to the coverage rate of the analysis combination in the previous time interval, and the hit number is greater than or equal to the hit number of the analysis combination in the previous time interval, compared with the previous time interval, it is determined that the follow-up stage migration is needed.
[0056] Alternatively, the current patient belongs to the current follow-up stage, the target follow-up stage, or is in the migration state is determined based on the first rule score and the first membership, the second rule score and the second membership; specifically, a corresponding table between the first rule score and the first membership, the second rule score and the second membership, and the follow-up stage (and the migration state) is constructed in advance; the follow-up stage to which the current patient belongs is obtained by querying the corresponding table; the corresponding table is obtained by statistical analysis of historical data, at this time, the migration stage can no longer be set, and the corresponding table can be directly set according to the follow-up stage to which the current patient belongs indicated by the historical data;
[0057] Alternatively, the first rule score and the first membership degree, the second rule score and the second membership degree determine that the current patient belongs to the current follow-up stage, the target follow-up stage, or is in a migration state; specifically, the first rule score and the first membership degree, the second rule score and the second membership degree are input into a discrimination model to obtain the follow-up stage to which the current patient belongs; the discrimination model is a pre-trained artificial intelligence model;
[0058] Preferably, the artificial intelligence model is a supervised model, a decision model, a neural network model, etc.
[0059] A computer program (also known as a program, software, software application, script, or code) can be written in any form of programming language, including compiled or interpreted languages, declarative or procedural languages, and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, object, or other unit suitable for use in a computing environment. A computer program may, but need not, correspond to a file in a file system. A program can be stored in a portion of a file that holds other programs or data (e.g., one or more scripts stored in a markup language document), in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files that store one or more modules, sub programs, or code portions). A computer program can be deployed to be executed on one computer or on multiple computers that are located at one site or distributed across multiple sites and interconnected by a communication network.
[0060] Those skilled in the art will appreciate that embodiments of the present application can be provided as methods, systems, or computer program products. Accordingly, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage media, etc.) having computer usable program code embodied in the medium.
[0061] The present application is described in reference to the flowcharts and / or block diagrams of the methods, apparatus (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general purpose computer, a special purpose computer, an embedded processor, or other programmable data processing apparatus to produce a machine, so that the instructions that are executed by the processor of the computer or other programmable data processing apparatus generate means for implementing the functions specified in the flowchart and / or block diagram block or blocks. Figure 1 The functions specified in a flow or multiple flows and / or blocks Figure 1 The means for implementing the functions specified in a flow or multiple flows and / or blocks
[0062] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the flow Figure 1 The functions of a flow or multiple flows and / or a block or multiple blocks in conjunction with the disclosed aspects. Figure 1
[0063] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flow Figure 1 The functions of a flow or multiple flows and / or a block or multiple blocks in conjunction with the disclosed aspects. Figure 1
[0064] Finally, it should be noted that the above-mentioned embodiments are merely intended to illustrate the technical solutions of the present application, but not to limit it. Although the present application has been described in detail with reference to the above-mentioned embodiments, it should be understood by those skilled in the art that the specific embodiments of the present application can be modified or replaced equivalently without departing from the spirit and scope of the present application, and any modifications or equivalent replacements thereof shall be included in the protection scope of the claims of the present application.
Claims
1. A hospital intelligent follow-up management method, characterized in that: Include: The follow-up process includes a main thread and a timer thread; the main thread executes the follow-up strategy, enters a dormant state after completion, and returns to the follow-up strategy execution state in response to the wake-up of the timer thread; the timer thread wakes up the main thread when the timer arrives; Determining a first rule score and a first membership that the patient belongs to the current follow-up stage; determining a second rule score and a second membership that the patient belongs to the target follow-up stage; and determining, based on the first rule score and the first membership, the second rule score and the second membership, whether the current patient belongs to the current follow-up stage, the target follow-up stage, or is in a migration state; Wherein: obtaining a patient parameter rule set and a feature combination set corresponding to a follow-up stage; matching the current patient parameter with the parameter rule set to obtain a rule score; comparing the current patient parameter sequence with each feature combination in the feature combination set, determining the hit feature combination and the number of hits to determine the membership; when the follow-up stage is the current follow-up stage, obtaining a first rule score and a first membership; when the follow-up stage is the target follow-up stage, obtaining a second rule score and a second membership; When the first rule score indicates that the patient is in the current follow-up stage and the first membership shows that the indication is credible and the second rule score indicates that the patient is not in the target follow-up stage and the second membership shows that the indication is credible, it is determined that the current patient is in the current follow-up stage; when the first rule score indicates that the patient is not in the current follow-up stage and the first membership shows that the indication is credible and the second rule score indicates that the patient is in the target follow-up stage and the second membership shows that the indication is credible, it is determined that the current patient is in the target follow-up stage; otherwise, it indicates a migration state; When in the migration state, it is further analyzed whether to migrate to the follow-up stage. If so, it will be determined to enter the target follow-up stage, otherwise it will maintain the current follow-up stage.
2. The hospital intelligent follow-up management method according to claim 1, characterized in that: After the patient parameters are updated, the patient parameters are sent to the server through the patient parameter collection interface; the server determines whether a follow-up stage change occurs based on the updated patient parameters, and when the follow-up stage changes, the server re-instantiates the follow-up strategy corresponding to the changed follow-up stage based on the updated patient parameters, and sends the re-instantiated follow-up strategy to the client to replace the current follow-up strategy executed by the main thread in the client.
3. The hospital intelligent follow-up management method according to claim 2, characterized in that: When a patient registers, a corresponding follow-up record data block is created on the server side; it is used to store patient parameters in chronological order; A patient parameter collection interface is created on the server side for real-time collection of patient parameters, and the patient parameters are stored in a follow-up record data block in chronological order or time interval order.
4. The hospital intelligent follow-up management method according to claim 3, characterized in that: The instantiation and setting of the main thread and the timing thread specifically include: determining the follow-up stage of the patient, and instantiating a follow-up strategy corresponding to the follow-up stage based on the patient parameters; using the follow-up strategy corresponding to the current follow-up stage as the current follow-up strategy executed in the main thread, and loading the follow-up strategy in the main thread; The timing time points are obtained from the instantiated follow-up strategy and arranged in chronological order to form a timing time sequence; and a timing thread is instantiated based on the timing time sequence.
5. The hospital intelligent follow-up management method according to claim 4, characterized in that: The follow-up strategy includes one or more follow-up events, each of which includes a trigger condition and a follow-up operation execution body. When obtaining a timing time point, the timing time point is determined based on the trigger condition. When the timing arrives, the follow-up event corresponding to the timing time point is awakened, and the main thread executes the follow-up event after being awakened.
6. The hospital intelligent follow-up management method according to claim 5, characterized in that: The follow-up operation execution body includes a follow-up operation, a follow-up instruction and / or a follow-up action set.
7. The hospital intelligent follow-up management method according to claim 6, characterized in that: When in the migration state, the migration state is analyzed for the follow-up stage based on the patient parameter sequence and the identification feature combination set within the most recent preset time length to determine whether follow-up stage migration is required. If so, the current patient will be determined to belong to the target follow-up stage, otherwise, the current follow-up stage will be maintained.
8. The hospital intelligent follow-up management method according to claim 7, characterized in that: The first membership is determined based on the hit feature combination and its hit times; specifically: the first membership is determined based on the hit feature combination and its hit times, so that the more hit feature combinations and / or the more hit times, the higher the first membership; otherwise, the lower the first membership.
9. A hospital intelligent follow-up management system, characterized in that: The hospital intelligent follow-up management system is used to implement the hospital intelligent follow-up management method described in any one of claims 1 to 8.
10. A hospital intelligent follow-up management platform, characterized by: The hospital intelligent follow-up management platform is used to implement the hospital intelligent follow-up management method described in any one of claims 1 to 8.