An intelligent generation method and system of an individualized nursing plan based on big data

Abstract

This application provides a method and system for intelligent generation of individualized nursing plans based on big data, belonging to the field of intelligent generation technology for individualized nursing plans. It involves collecting multi-source health data of the target nursing subject, performing feature encoding on the multi-source health data to generate a structured health record; performing similarity mapping on the health status of the target nursing subject according to a diagnostic mapping table and the structured health record to obtain a set of health statuses, and determining an individualized status index based on the health status set; detecting status differences based on real-time monitoring data and the structured health record to generate a status change trend vector; performing fusion analysis on the status change trend vector and the individualized status index to generate a dynamic health intervention index; and generating an individualized nursing plan for the target nursing subject based on the dynamic health intervention index. This application can achieve dynamic adaptation of nursing plans to the patient's real-time health status, improving the accuracy and timeliness of nursing plans.

Description

Technical Field

[0001] This application relates to the field of intelligent generation technology for individualized care plans, and more specifically, to a method and system for intelligent generation of individualized care plans based on big data. Background Technology

[0002] Currently, in the field of healthcare, with the rapid development of big data technology and IoT devices, data-driven nursing decision support systems are increasingly being applied in clinical practice. In existing technologies, nursing plans are typically formulated based on clinical guidelines and the professional experience of nursing staff. This involves collecting patients' health data and combining it with a standardized nursing knowledge base to recommend appropriate nursing interventions. Some technological solutions manage patient information uniformly by constructing electronic health records, using diagnostic mapping tables to match and analyze patients' symptoms and signs, or using wearable devices to monitor changes in patients' physiological parameters in real time. Furthermore, some systems employ health status assessment models to quantitatively score patients' health conditions and retrieve corresponding nursing plan templates from the nursing knowledge base based on the scoring results to assist nursing staff in making clinical decisions.

[0003] In existing technologies, nursing care plans are typically developed based on clinical guidelines and the personal experience of nurses, selecting standardized nursing interventions after a static assessment of the patient's health data. However, existing solutions suffer from several drawbacks: fragmented and heterogeneous multi-source health data, making effective integration and unified utilization difficult; health status assessments are often based on single, static data points, lacking the ability to perceive and quantify dynamic changes in the patient's health status in real time; nursing plans are disconnected from real-time patient monitoring data, making timely adjustments in response to changes in the patient's condition difficult; and standardized interventions lack individualized context, resulting in a low degree of matching between recommended nursing plans and actual patient needs, leading to confusion among nurses when faced with a large number of intervention options. Therefore, achieving dynamic adaptation of nursing plans to the patient's real-time health status and improving the accuracy and timeliness of nursing plans are challenges facing the industry. Summary of the Invention

[0004] This application provides a method and system for intelligent generation of individualized nursing plans based on big data, which can realize the dynamic adaptation of nursing plans to the real-time health status of patients, thereby improving the accuracy and timeliness of nursing plans.

[0005] Firstly, this application provides a method for intelligently generating individualized care plans based on big data, the intelligent generation method comprising the following steps:

[0006] Collect multi-source health data of the target care subjects, perform feature encoding on the multi-source health data, and generate structured health records;

[0007] Obtain a diagnostic mapping table, perform similarity mapping on the health status of the target care object based on the diagnostic mapping table and the structured health record, obtain a set of health status of the target care object, and determine the individualized status index of the target care object based on the set of health status.

[0008] The system acquires real-time monitoring data from the smart chip worn by the target care subject, performs state difference detection based on the real-time monitoring data and the structured health record, generates a state change trend vector, and performs fusion analysis based on the state change trend vector and the individualized state index to generate a dynamic health intervention index.

[0009] Based on the dynamic health intervention index, an individualized care plan is generated for the target care recipient.

[0010] In this embodiment, the multi-source health data includes static attribute data and clinical diagnosis and treatment data.

[0011] In this embodiment, the process of feature encoding the multi-source health data to generate a structured health record specifically includes:

[0012] The multi-source health data is classified according to data type, and key feature fields are extracted from each type of data.

[0013] For text-based data, clinical indicator descriptions and status labels are extracted through keyword matching and semantic rule parsing.

[0014] For numerical time series data, calculate its statistical characteristics within a preset time window, including mean, maximum, minimum, rate of change, and fluctuation amplitude;

[0015] The extracted and calculated feature fields are combined according to a preset data structure template to generate a structured health record.

[0016] In this embodiment, the set of health statuses of the target care recipients is obtained by performing a similarity mapping based on the diagnostic mapping table and the structured health record, specifically including:

[0017] Extract the current state feature vector of the target care subject from the structured health record;

[0018] Obtain the standard feature vector corresponding to each diagnosis name from the diagnosis mapping table;

[0019] Based on the current state feature vector and the standard feature vector corresponding to each diagnostic name, the state response is performed respectively to obtain the corresponding diagnostic state response degree;

[0020] Threshold screening is performed based on the responsiveness of each diagnostic state to obtain the set of health statuses of the target care subjects.

[0021] In this embodiment, determining the individualized status index of the target care subject based on the set of health statuses specifically includes:

[0022] The state weight of each health state is determined based on the set of health states.

[0023] Obtain the pre-configured baseline values ​​for the status scores corresponding to each health status;

[0024] The individualized status index of the target care subject is determined based on the status weight of each health status and the corresponding status score benchmark value.

[0025] In this embodiment, the individualized state index is an index used to quantitatively characterize the current overall health level of the target care subject.

[0026] In this embodiment, the generation of a dynamic health intervention index through fusion analysis based on the state change trend vector and the individualized state index specifically includes:

[0027] The health status coordinate system of the target care subject is determined based on the state change trend vector and the individualized state index;

[0028] Obtain a pre-defined health intervention decision map;

[0029] Based on the health status coordinate system of the target nursing object and the health intervention decision map, the health status is located to obtain the target decision area in the health intervention decision map;

[0030] A dynamic health intervention index is generated based on the target decision area.

[0031] In this embodiment, the dynamic health intervention index is used to quantitatively characterize the degree of deviation of the current health status of the target care subject from the baseline point of its decision-making area.

[0032] In this embodiment, the individualized care plan includes: a set of care measures, the frequency of implementation, and the duration of implementation.

[0033] Secondly, this application provides a big data-based intelligent generation system for individualized nursing care plans, used to execute a big data-based intelligent generation method for individualized nursing care plans, the intelligent generation system comprising:

[0034] The data acquisition module is used to collect multi-source health data of the target care object, perform feature encoding on the multi-source health data, and generate a structured health record;

[0035] The index determination module is used to obtain a diagnostic mapping table, perform similar mapping on the health status of the target nursing object according to the diagnostic mapping table and the structured health record, obtain a set of health status of the target nursing object, and determine the individualized status index of the target nursing object based on the set of health status.

[0036] The health intervention module is used to acquire real-time monitoring data from the smart chip worn by the target care subject, perform state difference detection based on the real-time monitoring data and the structured health record, generate a state change trend vector, and perform fusion analysis based on the state change trend vector and the individualized state index to generate a dynamic health intervention index.

[0037] The plan generation module is used to generate individualized care plans for the target care subjects based on the dynamic health intervention index.

[0038] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects:

[0039] Multi-source health data of the target care subject is collected, and the multi-source health data is feature-encoded to generate a structured health record. A diagnostic mapping table is obtained, and the health status of the target care subject is similarly mapped according to the diagnostic mapping table and the structured health record to obtain a set of health statuses of the target care subject. An individualized status index of the target care subject is determined based on the set of health statuses. Real-time monitoring data from the smart chip worn by the target care subject is obtained, and status difference detection is performed based on the real-time monitoring data and the structured health record to generate a status change trend vector. The status change trend vector and the individualized status index are fused and analyzed to generate a dynamic health intervention index. An individualized care plan for the target care subject is generated based on the dynamic health intervention index.

[0040] Therefore, this application firstly classifies and encodes multi-source health data, transforming scattered and heterogeneous raw data into structured health records, providing a standardized data foundation for subsequent health status analysis and nursing plan generation. Secondly, by constructing a diagnostic mapping table and performing similarity mapping on health status, multidimensional health features are quantitatively matched with standard diagnostic features to obtain a set of health statuses, which are then integrated to generate an individualized status index, achieving an objective quantitative expression and comprehensive assessment of health status, providing a precise status benchmark for subsequent nursing decisions. Thirdly, by acquiring real-time monitoring data from a smart chip and performing status difference detection with the structured health records, a status change trend vector reflecting dynamic health changes is generated. This vector is then integrated with the individualized status index to obtain a dynamic health intervention index, achieving real-time perception and quantitative representation of health status change trends, providing timely and accurate decision-making basis for dynamic adjustment of nursing plans. Finally, by matching and personalizing the dynamic health intervention index with a preset nursing plan template, an individualized nursing plan that is adapted to the current health status of the target nursing subject in real time is generated, realizing the transformation of nursing plans from static standardization to dynamic individualization, improving the accuracy and timeliness of nursing interventions.

[0041] In summary, the technical solution adopted in this application can achieve dynamic adaptation between nursing plans and patients' real-time health status, thereby improving the accuracy and timeliness of nursing plans. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this embodiment of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is an exemplary flowchart of a method for intelligently generating individualized care plans based on big data, as provided in this application.

[0044] Figure 2 This is a flowchart illustrating the process for determining the individualized status index of the target care recipient, as provided in this application.

[0045] Figure 3 This is a flowchart illustrating the process of generating dynamic health intervention indices based on the information provided in this application;

[0046] Figure 4 This is a module structure diagram of an intelligent generation system for individualized nursing plans based on big data, provided in this application. Detailed Implementation

[0047] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0048] This application provides a method and system for intelligently generating individualized care plans based on big data. The core of this method involves: collecting multi-source health data of the target care subject; performing feature encoding on the multi-source health data to generate a structured health record; obtaining a diagnostic mapping table; performing similarity mapping on the health status of the target care subject based on the diagnostic mapping table and the structured health record to obtain a set of health statuses for the target care subject; determining an individualized status index for the target care subject based on the set of health statuses; acquiring real-time monitoring data from a smart chip worn by the target care subject; performing status difference detection based on the real-time monitoring data and the structured health record to generate a status change trend vector; performing fusion analysis based on the status change trend vector and the individualized status index to generate a dynamic health intervention index; and generating an individualized care plan for the target care subject based on the dynamic health intervention index.

[0049] Example 1: To better understand the above technical solution, the following will provide a detailed description of the technical solution in conjunction with the accompanying drawings and specific implementation methods. (Refer to...) Figure 1 As shown in the figure, this is an exemplary flowchart of a method for intelligently generating individualized care plans based on big data, according to this embodiment of the present application. The intelligent generation method includes the following steps:

[0050] In step S1, multi-source health data of the target care subject is collected, and the multi-source health data is feature-encoded to generate a structured health record.

[0051] In practical implementation, firstly, multi-source health data of the target nursing subjects can be collected. This multi-source health data includes static attribute data and clinical diagnosis and treatment data. Data collection strategies can be configured according to the nursing scenario of the target nursing subjects. For inpatient nursing scenarios, the data collection system is configured with hospital information system interface adapters, laboratory information system interface adapters, medical image archiving and communication system interface adapters, and bedside medical IoT gateways. For home nursing scenarios, wearable device data receiving adapters, home health monitoring terminal adapters, and mobile health application interface adapters are configured. Each data collection adapter adopts a plug-in architecture design, with each adapter containing a protocol parsing module, a data caching module, and a breakpoint resume module. The obtained data is used as static attribute data. Then, through the hospital information system interface adapter, the demographic information of the target nursing subjects is extracted from the electronic medical record database. This includes name, gender, age, height, weight, ethnicity, and occupation; past medical history information, including previous diagnostic records, surgical history, and trauma history; allergy history information, including drug allergies, food allergies, and other allergens; family genetic history information, including records of hereditary diseases in immediate family members; and, through a laboratory information system interface adapter, the collection of laboratory test results for the target nursing subject, including specific values ​​and reference ranges for tests such as complete blood count, urinalysis, comprehensive biochemistry, coagulation function, and blood gas analysis; through a medical image archiving and communication system interface adapter, the collection of report texts and image metadata from imaging examinations such as X-rays, CT scans, MRIs, and ultrasounds; and through a hospital information system interface adapter, the collection of clinical document data such as medical orders, nursing records, and progress notes. The obtained data is used as clinical diagnosis and treatment data, thus combining static attribute data and clinical diagnosis and treatment data into multi-source health data for the target nursing subject.

[0052] In this embodiment, the generation of structured health records by feature encoding of the multi-source health data can be achieved through the following steps:

[0053] The multi-source health data is classified according to data type, and key feature fields are extracted from each type of data.

[0054] For text-based data, clinical indicator descriptions and status labels are extracted through keyword matching and semantic rule parsing.

[0055] For numerical time series data, calculate its statistical characteristics within a preset time window, including mean, maximum, minimum, rate of change, and fluctuation amplitude;

[0056] The extracted and calculated feature fields are combined according to a preset data structure template to generate a structured health record.

[0057] In practical implementation, firstly, the multi-source health data can be classified according to data type, and key feature fields from each type of data can be extracted. That is, the collected raw multi-source health data can be read and divided into text data and numerical time-series data according to the type identifier in the data header. For text data, its document type is identified, such as medical records, laboratory report text, imaging report text, or nursing record sheet; for numerical time-series data, its corresponding physiological indicator type is identified, such as heart rate sequence, blood pressure sequence, blood oxygen sequence, or blood glucose sequence. After classification, key feature fields are located and extracted from each type of data according to preset feature extraction rules, such as: The process involves extracting chief complaint and physical sign description fields from medical records, and extracting test item names and result values ​​from laboratory reports. Then, for text-based data, clinical indicator descriptions and status labels are extracted through keyword matching and semantic rule parsing. Specifically, a pre-built medical dictionary and semantic rule base can be loaded. The medical dictionary includes clinical terms such as symptom names, physical sign names, disease names, drug names, and test item names, along with their synonyms and near-synonyms. The semantic rule base defines language patterns for identifying clinical status descriptions. For each piece of text data, Chinese word segmentation and part-of-speech tagging are performed first, and then the segmentation results are matched with the medical dictionary to identify... The text identifies clinical terms and then applies rule templates from a semantic rule base to perform pattern matching on the surrounding context, extracting the attribute modification relationships of the terms. For example, from "patient's chief complaint of persistent right lower quadrant pain," structured descriptions such as "pain location = right lower quadrant" and "pain nature = persistent" are extracted. The extracted clinical indicator descriptions and status labels are then standardized according to a preset format to generate structured text feature fields. Furthermore, for numerical time-series data, its statistical characteristics within a preset time window are calculated, including mean, maximum, minimum, rate of change, and fluctuation amplitude. In other words, preset features can be provided for each type of time-series data. The time window length is selected, for example: a 5-minute window for heart rate data, a 24-hour window for blood pressure data, and a 2-hour window for blood glucose data. Continuous time-series data is segmented according to the time window. Statistical calculations are performed on the data sequence within each window: the arithmetic mean of the sequence is calculated as a central tendency feature; the maximum and minimum values ​​in the sequence are identified as extreme values; the slope of the sequence within the window is calculated through linear fitting as a rate of change feature to characterize the speed of change of the indicator; and the standard deviation of the sequence is calculated as a volatility feature to characterize the stability of the indicator. The statistical feature values ​​of each time window are then associated and stored with their corresponding timestamps.Finally, the extracted and calculated feature fields can be combined according to a preset data structure template to generate a structured health record. That is, a preset data structure template for the health record can be used. This template adopts a hierarchical architecture design, including a basic attribute layer and a time-series state layer. The basic attribute layer stores key feature fields extracted from static attribute data, such as age, gender, previous diagnosis name, and allergen name, organized in a key-value pair format. The time-series state layer stores state information extracted from text data and time-series data that changes over time, organized in a time-series database format. Each record contains a timestamp, indicator name, and indicator value. The various feature fields obtained in the above steps are filled according to their corresponding levels and field names in the template. Missing fields are filled with null values ​​or default values. After filling, a unique identifier is assigned to the record and a generation timestamp is added, ultimately generating a structured health record.

[0058] In step S2, a diagnostic mapping table is obtained, and the health status of the target nursing object is similarly mapped according to the diagnostic mapping table and the structured health record to obtain a set of health statuses of the target nursing object. Based on the set of health statuses, the individualized status index of the target nursing object is determined.

[0059] In practical implementation, obtaining the diagnostic mapping table involves extracting various disease names and their typical clinical manifestations from the International Classification of Diseases, clinical practice guidelines, and a verified historical case database. The collected disease diagnostic names are then deduplicated and standardized, unifying different descriptions of the same disease into a single standard diagnostic name. This generates a basic list containing multiple standard diagnostic names. For each standard diagnostic name, typical symptoms and signs corresponding to that diagnosis are compiled from clinical data. Typical symptoms include the patient's chief complaint, and typical signs include the typical ranges of vital signs such as body temperature, heart rate, and blood pressure. A fixed-dimensional feature vector is constructed for each diagnostic name, with each position in the vector corresponding to a symptom or sign indicator. The corresponding feature value is filled in at the typical position of the diagnosis, and 0 is filled in at atypical positions. Each obtained feature vector serves as the standard feature vector for each corresponding diagnostic name. Each diagnostic name is then associated with and stored with its corresponding standard feature vector, and the resulting stored list serves as the diagnostic mapping table.

[0060] In this embodiment, the process of performing a similarity mapping on the health status of the target care subject based on the diagnostic mapping table and the structured health record to obtain the set of health statuses of the target care subject can be achieved through the following steps:

[0061] Extract the current state feature vector of the target care subject from the structured health record;

[0062] Obtain the standard feature vector corresponding to each diagnosis name from the diagnosis mapping table;

[0063] Based on the current state feature vector and the standard feature vector corresponding to each diagnostic name, the state response is performed respectively to obtain the corresponding diagnostic state response degree;

[0064] Threshold screening is performed based on the responsiveness of each diagnostic state to obtain the set of health statuses of the target care subjects.

[0065] In practical implementation, firstly, the current state feature vector of the target care subject can be extracted from the structured health record. That is, the structured health record of the target care subject can be read, and the health indicator data within a preset time window can be extracted from the temporal state layer of the structured health record. The current state feature vector consists of two-dimensional feature components: symptom feature component and vital sign feature component. The symptom feature component is obtained from the text parsing results in the record, including the names of the currently existing symptoms and their severity levels. Each symptom is converted into a numerical feature value according to a preset encoding rule. The vital sign feature component is obtained from the temporal statistical features in the structured health record, including the mean, latest value, and trend of various vital signs within the current time window. The above two types of feature components are concatenated into a numerical vector, which is used as the current state feature vector of the target care subject. Then, the standard feature vector corresponding to each diagnosis name can be obtained from the diagnosis mapping table. That is, the diagnosis mapping table can be loaded, and the standard feature vector corresponding to each diagnosis name can be extracted from the diagnosis mapping table.

[0066] Furthermore, in practical implementation, the state response can be performed based on the current state feature vector and the standard feature vector corresponding to each diagnostic name to obtain the corresponding diagnostic state response degree. That is, the current state feature vector can be compared with the standard feature vector of each diagnostic name in turn. For the symptom feature part, the current state feature vector contains multiple currently existing symptom records. Each symptom record includes a symptom name and a severity level. The system matches each current symptom with the typical symptom set of the diagnosis recorded in the standard feature vector: if the current symptom name appears in the typical symptom set, the matching score of the symptom is calculated based on the similarity between the severity level of the current symptom and the severity level of the typical symptom of the diagnosis. The matching score is 1 when the severity is exactly the same, 0.7 when the severity differs by one level, and 0.3 when the severity differs by two or more levels. If a symptom name does not appear in the typical symptom set, the matching score for that symptom is 0. The matching scores of all current symptoms are summed to obtain the total matching score for the symptom section. For the vital signs section, the current state feature vector contains the values ​​of each currently measured vital sign. Each current vital sign value is compared with the reference range of the corresponding vital sign recorded in the standard feature vector: if the current vital sign value is within the reference range, the matching score for that vital sign is 1; if the current vital sign value deviates from the reference range but does not exceed the critical value (where the critical value can be preset by expert experience), the matching score is calculated based on the degree of deviation, with a higher score for a smaller deviation and a lower score for a larger deviation; if the current vital sign value exceeds the critical value, the matching score for that vital sign is 0. The matching scores of all vital sign indicators are summed to obtain the total matching score for the vital signs section. For example: suppose the normal reference range for a certain vital sign is [L, H], where L is the lower limit and H is the upper limit, and the current measurement value is x.The deviation *e* of *x* from the reference range is calculated as follows: If *L* ≤ *x* ≤ *H*, then *e* = 0, indicating no deviation; if *x* < *L*, then *e* = (*L* - *x*) / *L*, indicating a deviation below the lower limit; if *x* > *H*, then *e* = (*x* - *H*) / *H*, indicating a deviation above the upper limit. The matching score for this trait is calculated based on the deviation *e*: when *e* = 0, the matching score is 1.0; when *0* < *e* ≤ 0.1, the matching score is 0.9; when *0.1* < *e* ≤ 0.2, the matching score is 0.7; when *0.2* < *e* ≤ 0.3, the matching score is 0.4; when *0.3* < *e* ≤ 0.4, the matching score is 0.2; when *e* > *H*, the matching score is 0.2. At a score of 0.4, the matching score is 0. Based on the experimental data analysis, preset symptom and sign weights are used. The total matching score for the symptom portion is multiplied by the symptom weight, and the total matching score for the sign portion is multiplied by the sign weight. These two scores are then added together, and the result is used as the diagnostic status responsiveness corresponding to the diagnostic name. The value ranges from 0 to 1; a higher value indicates a better match between the current health status and the typical manifestations of the diagnosis. The diagnostic status responsiveness quantitatively characterizes the degree of match between the current health status of the target nursing subject and the typical clinical manifestations corresponding to the standard diagnostic name. Then, a threshold screening can be performed based on each diagnostic status responsiveness to obtain the set of health statuses of the target nursing subject. That is, a responsiveness threshold can be preset, and all diagnostic status responsivenesses are compared with the threshold. Diagnostic names with responsivenesses greater than the threshold are extracted and sorted from high to low. The resulting sequence is used as the set of health statuses of the target nursing subject, where the health status set contains one or more candidate diagnostic names and their corresponding diagnostic status responsivenesses.

[0067] Preferably, in this embodiment, the individualized status index of the target care subject is determined based on the set of health statuses, with reference to... Figure 2 As shown in the figure, this is a flowchart illustrating the process of determining the individualized status index of the target care subject in some embodiments of this application. In this embodiment, the determination of the individualized status index of the target care subject can be achieved through the following steps:

[0068] In step S21, the state weight of each health state is determined according to the set of health states;

[0069] In step S22, the pre-configured status score benchmark values ​​corresponding to each health status are obtained;

[0070] In step S23, the individualized state index of the target care subject is determined based on the state weight of each health state and the corresponding state score benchmark value.

[0071] In practical implementation, firstly, the state weights of each health state can be determined based on the set of health states. That is, for each health state in the set, the corresponding diagnostic state responsiveness is used as the initial weight of that health state. Then, the initial weights of all health states are normalized so that the sum of the weights of all health states in the set equals 1. The normalized weight value is the state weight of each health state. Secondly, the pre-configured state scoring benchmark values ​​corresponding to each health state can be obtained. That is, the state scoring benchmark values ​​are pre-set values ​​based on the severity of each health state, the degree of impact on quality of life, and the urgency of clinical intervention. These values ​​can be preset based on experimental data analysis. Finally, the individualized state index of the target nursing subject can be determined based on the state weights of each health state and the corresponding state scoring benchmark values. That is, the state weights of each health state are multiplied by the corresponding state scoring benchmark values, and the results are summed. The summed result is used as the individualized state index of the target nursing subject. The individualized state index is used to quantitatively characterize the current overall health level of the target nursing subject. The higher the value, the worse the health status.

[0072] It should be noted that by constructing a diagnostic mapping table and performing similar mapping on health status, multidimensional health characteristics are quantitatively matched with standard diagnostic characteristics to obtain a set of health statuses. These sets are then integrated to generate individualized status indices, achieving an objective quantitative expression and comprehensive assessment of health status, and providing a precise status benchmark for subsequent nursing decisions.

[0073] In step S3, real-time monitoring data from the smart chip worn by the target care subject is acquired. Based on the real-time monitoring data and the structured health record, state difference detection is performed to generate a state change trend vector. Based on the state change trend vector and the individualized state index, a fusion analysis is performed to generate a dynamic health intervention index.

[0074] In practical implementation, to obtain real-time monitoring data from the smart chip worn by the target care recipient, a wireless connection can be established with the smart chip worn by the target care recipient via Bluetooth. After confirming that the chip is online and can communicate normally, a data reading command is sent to the smart chip according to the preset sampling interval. After receiving the command, the chip returns the sensor measurement value at the current moment. The data returned by the chip, including heart rate, body temperature, blood oxygen saturation, blood pressure and the corresponding collection timestamp, is received and used as real-time monitoring data.

[0075] In practical implementation, state difference detection can be performed based on real-time monitoring data and structured health records to generate a state change trend vector. Specifically, real-time monitoring data can be parsed to extract current physiological indicators, including current heart rate, body temperature, blood oxygen saturation, and blood pressure. These values ​​are then used to form a current monitoring feature vector. Next, the target patient's structured health record is read, and historical monitoring data within a preset time period is extracted from the temporal state layer of the health record. Statistical baseline values ​​for each indicator within the historical time period are calculated, including historical mean, historical median, or historical stable period values. These baseline values ​​are then used to form a historical baseline feature vector. The current monitoring feature vector is compared item by item with the historical baseline feature vector. For each monitoring indicator, the current monitoring feature vector is subtracted from the historical baseline feature vector, and the result is used as the state difference value. These state difference values ​​are then combined into a vector, which is used as the state value. The state difference vector is used to reflect the degree of deviation of each indicator from the historical normal level. Positive values ​​indicate that the indicator is rising, and negative values ​​indicate that the indicator is falling. Based on the normal fluctuation range of each monitoring indicator, the components in the difference vector are normalized. A fixed-length time window is maintained to cache the normalized state difference vectors of the most recent T moments. For each monitoring indicator, its normalized difference value sequence within the time window is linearly fitted, and the slope of the fitted line is calculated. The result is used as the rate of change of the monitoring indicator. The rates of change of each monitoring indicator are combined into a trend vector. The normalized difference vector of the most recent moment is used as the current deviation vector. The trend vector and the current deviation vector are concatenated to obtain the state change trend vector. The state change trend vector is used to characterize the degree of deviation of the target nursing object from the historical baseline and the trend of change of each indicator.

[0076] Preferably, in this embodiment, a dynamic health intervention index is generated by fusing the state change trend vector and the individualized state index, with reference to... Figure 3 As shown in the figure, this is a schematic diagram of the process for generating a dynamic health intervention index in some embodiments of this application. In this embodiment, the generation of the dynamic health intervention index can be achieved by the following steps:

[0077] In step S31, the health status coordinate system of the target care object is determined based on the state change trend vector and the individualized state index;

[0078] In step S32, a preset health intervention decision map is obtained;

[0079] In step S33, the health status is located based on the health status coordinate system of the target nursing object and the health intervention decision map to obtain the target decision area in the health intervention decision map;

[0080] In step S34, a dynamic health intervention index is generated based on the target decision region.

[0081] In practical implementation, firstly, the health status coordinate system of the target nursing subject can be determined based on the state change trend vector and the individualized state index. That is, each rate of change can be extracted from the state change trend vector, and the corresponding influence weights can be preset according to the degree of influence of each monitoring indicator on the health status. Then, the weighted sum of each rate of change and its corresponding influence weight is used as the comprehensive rate of change. The individualized state index is used as the abscissa, and the comprehensive rate of change is used as the ordinate to construct the health status coordinate system of the target nursing subject in a two-dimensional space, and the state coordinate points of the target nursing subject are obtained. Then, a preset health intervention decision map can be obtained, that is, a pre-generated health intervention decision map can be loaded. The health intervention decision map is a two-dimensional planar diagram. Its horizontal axis represents the range of individualized state indices, and its vertical axis represents the range of comprehensive rate of change. The map is divided into multiple decision regions, each corresponding to a polygonal region defined by both horizontal and vertical axis intervals. Each decision region is pre-associated with a type of health intervention strategy, such as observation and waiting strategy, lifestyle intervention strategy, outpatient visit strategy, or emergency treatment strategy. It is also associated with a corresponding guidance parameter template, which defines the value ranges for parameters such as nursing goals, intervention intensity, and examination frequency. The construction of the health intervention decision map is based on expert experience and experimental data analysis; further... The system can locate the health status of the target care subject based on the health status coordinate system and the health intervention decision map, obtaining the target decision area in the health intervention decision map. Specifically, the state coordinate points of the target care subject can be mapped onto the health intervention decision map using the ray method for location. The polygonal area where the state coordinate points are located is then determined as the target decision area, and the corresponding area identifier is obtained. If the state coordinate point falls exactly on the boundary between two decision areas, the decision area with higher priority is selected according to preset boundary processing rules. Finally, a dynamic health intervention index can be generated based on the target decision area; that is, the geometric center coordinates of the target decision area and the index of that area within the individual can be obtained. The boundary intervals on the state index axis and the boundary intervals on the comprehensive rate of change axis are defined. The coordinates of the center point of the region represent the typical health status characteristics of the region. The state coordinates of the target care object are subtracted from the center point of the region, and the result is used as the state offset vector. The state offset vector is used to reflect the degree and direction of deviation of the target care object from the typical state in the current region. The vector magnitude of the state offset vector is then calculated, and the result is normalized. The result is then used as the dynamic health intervention index. It should be noted that the dynamic health intervention index is used to quantitatively characterize the degree of deviation of the target care object's current health status from the baseline point of its decision region.

[0082] It should be noted that by acquiring real-time monitoring data from the smart chip and detecting state differences with structured health records, a state change trend vector reflecting dynamic changes in health is generated. This vector is then fused with an individualized state index to obtain a dynamic health intervention index. This enables real-time perception and quantitative representation of health status change trends, providing timely and accurate decision-making basis for the dynamic adjustment of nursing plans.

[0083] In step S4, an individualized care plan for the target care subject is generated based on the dynamic health intervention index.

[0084] In practical implementation, individualized nursing plans for the target nursing object can be generated based on the dynamic health intervention index. That is, multiple dynamic health intervention index intervals can be preset, each interval corresponding to an intervention level. The dynamic health intervention index is compared with each interval to determine its intervention level. The intervention level includes Level 1 intervention, Level 2 intervention, and Level 3 intervention. The higher the level, the higher the intensity and urgency of the nursing intervention. Then, the basic nursing package corresponding to the intervention level is retrieved from the nursing plan library. The basic nursing package contains a set of standard nursing measures, execution frequency, and execution duration under that level, so that the corresponding basic nursing package is used as the individualized nursing plan for the target nursing object.

[0085] Therefore, this application firstly classifies and encodes multi-source health data, transforming scattered and heterogeneous raw data into structured health records, providing a standardized data foundation for subsequent health status analysis and nursing plan generation. Secondly, by constructing a diagnostic mapping table and performing similarity mapping on health status, multidimensional health features are quantitatively matched with standard diagnostic features to obtain a set of health statuses, which are then integrated to generate an individualized status index, achieving an objective quantitative expression and comprehensive assessment of health status, providing a precise status benchmark for subsequent nursing decisions. Thirdly, by acquiring real-time monitoring data from a smart chip and performing status difference detection with the structured health records, a status change trend vector reflecting dynamic health changes is generated. This vector is then integrated with the individualized status index to obtain a dynamic health intervention index, achieving real-time perception and quantitative representation of health status change trends, providing timely and accurate decision-making basis for dynamic adjustment of nursing plans. Finally, by matching and personalizing the dynamic health intervention index with a preset nursing plan template, an individualized nursing plan that is adapted to the current health status of the target nursing subject in real time is generated, realizing the transformation of nursing plans from static standardization to dynamic individualization, improving the accuracy and timeliness of nursing interventions.

[0086] In summary, the technical solution adopted in this application can achieve dynamic adaptation between nursing plans and patients' real-time health status, thereby improving the accuracy and timeliness of nursing plans.

[0087] Example 2: This application provides an intelligent generation system for individualized nursing plans based on big data, referencing... Figure 4 As shown in the figure, this is a module structure diagram of a big data-based intelligent generation system for individualized nursing plans according to this embodiment of the present application. The intelligent generation system includes:

[0088] The data acquisition module 100 is used to collect multi-source health data of the target care object, perform feature encoding on the multi-source health data, and generate a structured health record;

[0089] The index determination module 200 is used to obtain a diagnostic mapping table, perform similar mapping on the health status of the target nursing object according to the diagnostic mapping table and the structured health record, obtain a set of health status of the target nursing object, and determine the individualized status index of the target nursing object based on the set of health status.

[0090] The health intervention module 300 is used to acquire real-time monitoring data from the smart chip worn by the target care subject, perform state difference detection based on the real-time monitoring data and the structured health record, generate a state change trend vector, and perform fusion analysis based on the state change trend vector and the individualized state index to generate a dynamic health intervention index.

[0091] The plan generation module 400 is used to generate individualized care plans for the target care subjects based on the dynamic health intervention index.

[0092] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations 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, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0093] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-Erasable Programmable Read-Only Memory (EEPROM), compactdisc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

[0094] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

Claims

1. A method for intelligently generating individualized nursing plans based on big data, characterized in that, The intelligent generation method includes the following steps: Collect multi-source health data of the target care subjects, perform feature encoding on the multi-source health data, and generate structured health records; A diagnostic mapping table is obtained. Based on the diagnostic mapping table and the structured health record, the health status of the target nursing object is similarly mapped to obtain a set of health statuses of the target nursing object. The state weight of each health status is determined based on the set of health statuses. The pre-configured state scoring benchmark value corresponding to each health status is obtained. Based on the state weight of each health status and the corresponding state scoring benchmark value, the individualized state index of the target nursing object is determined. The individualized state index is an index used to quantitatively characterize the current overall health level of the target nursing object. The system acquires real-time monitoring data from the smart chip worn by the target care subject, performs state difference detection based on the real-time monitoring data and the structured health record, generates a state change trend vector, determines the health state coordinate system of the target care subject based on the state change trend vector and the individualized state index, obtains a preset health intervention decision map, locates the health state based on the health state coordinate system of the target care subject and the health intervention decision map, obtains the target decision area in the health intervention decision map, and generates a dynamic health intervention index based on the target decision area. The dynamic health intervention index is used to quantitatively characterize the degree of deviation of the target care subject's current health state from the benchmark point of its decision area. Based on the dynamic health intervention index, an individualized care plan is generated for the target care recipient.

2. The method for intelligently generating individualized nursing plans based on big data as described in claim 1, characterized in that, The multi-source health data includes static attribute data and clinical diagnosis and treatment data.

3. The method for intelligently generating individualized nursing plans based on big data as described in claim 1, characterized in that, The process of feature encoding the multi-source health data to generate a structured health record specifically includes: The multi-source health data is classified according to data type, and key feature fields are extracted from each type of data. For text-based data, clinical indicator descriptions and status labels are extracted through keyword matching and semantic rule parsing. For numerical time series data, calculate its statistical characteristics within a preset time window, including mean, maximum, minimum, rate of change, and fluctuation amplitude; The extracted and calculated feature fields are combined according to a preset data structure template to generate a structured health record.

4. The method for intelligently generating individualized nursing plans based on big data as described in claim 1, characterized in that, Based on the diagnostic mapping table and the structured health records, a similarity mapping is performed on the health status of the target care subjects to obtain the set of health statuses of the target care subjects, specifically including: Extract the current state feature vector of the target care subject from the structured health record; Obtain the standard feature vector corresponding to each diagnosis name from the diagnosis mapping table; Based on the current state feature vector and the standard feature vector corresponding to each diagnostic name, the state response is performed respectively to obtain the corresponding diagnostic state response degree; Threshold screening is performed based on the responsiveness of each diagnostic state to obtain the set of health statuses of the target care subjects.

5. The method for intelligently generating individualized nursing plans based on big data as described in claim 1, characterized in that, The individualized nursing care plan includes: a set of nursing measures, frequency of implementation, and duration of implementation.

6. A big data-based intelligent generation system for individualized nursing care plans, used to execute the big data-based intelligent generation method for individualized nursing care plans as described in any one of claims 1 to 5, characterized in that, The intelligent generation system includes: The data acquisition module is used to collect multi-source health data of the target care object, perform feature encoding on the multi-source health data, and generate a structured health record; The index determination module is used to obtain a diagnostic mapping table, perform similar mapping on the health status of the target nursing object according to the diagnostic mapping table and the structured health record, obtain a set of health status of the target nursing object, and determine the individualized status index of the target nursing object based on the set of health status. The health intervention module is used to acquire real-time monitoring data from the smart chip worn by the target care subject, perform state difference detection based on the real-time monitoring data and the structured health record, generate a state change trend vector, and perform fusion analysis based on the state change trend vector and the individualized state index to generate a dynamic health intervention index. The plan generation module is used to generate individualized care plans for the target care subjects based on the dynamic health intervention index.

Images