Purpose-oriented physiological intervention method and system

Abstract

The purpose-oriented physiological intervention method and system include a data evaluation module, an intervention decision-making module, a purpose requirement management module, and an intervention module; through step C of the implementation method planning of data evaluation, step D of the analysis model loading, step E of data preprocessing and diversion, step F of physiological orientation analysis, step G of purpose requirement conversion, step H of purpose-oriented analysis, step I of intervention decision-making, and step J of intervention measure implementation, that is, the intervention management module according to the second intervention plan dataset Sch2 generated by the intervention decision-making module t , to select the current intervention method and intervention mechanism to be executed and implement the intervention measures. Through purpose-oriented analysis, the physiological intervention method has a purpose orientation; and there are clear records and feedback in the process, forming a real-time closed loop between the intervention purpose and measures, greatly improving the real-time performance and intervention efficiency of the physiological intervention method.

Description

Technical Field

[0001] The present invention relates to the technical field of physiological intervention, and particularly to a purpose-oriented physiological intervention method and system. Background Art

[0002] With the continuous improvement of people's living standards and the numerous changes in the global public health situation, people's need to improve their physical and mental health has become increasingly urgent. These needs involve more and more products and industries, such as: medical care, health care, entertainment (games, AR, VR, etc.), sports (sports assessment, first aid, etc.), sleep, quality care (for infants, the elderly, and other persons with mobility impairments), etc. In the prior art, the products and personnel involved in each link of the above process are different. Usually, detection devices are set up independently; evaluation devices are set up independently, and intervention devices are set up independently; the connection between these independently set detection devices, evaluation devices, and intervention devices is usually very weak, or even non-existent. Therefore, it is impossible to organically combine various means such as monitoring, evaluation, and intervention, and it is very difficult to achieve systematic closed-loop regulation of the entire process, and thus it is also very difficult to perform safe, timely, and effective intervention on the target.

[0003] In the patent application with the application number CN201110197578, a general physiological signal converter is proposed. The main problem it aims to solve is that for the same physiological signal acquisition, due to different acquisition hardware, different data values are formed. After passing through the said physiological signal converter and then through an inference engine, they become relatively unified data values. Its essence is a numerical correction scheme for the same physiological signal.

[0004] In the utility model patent application with the application number CN201220476675, a health detection physiotherapy instrument is proposed. Its main method is to collect signals through a temperature, pulse, or blood pressure acquisition unit, store them, and then compare them with preset threshold data to control the formation of pulsed infrared rays with different pulse widths, and then perform physiotherapy on the wearer. It only detects two to three limited physiological signals, and at the same time, it gives a one-way open-loop system.

[0005] Similarly, in the utility model patent application with the application number CN201521040909, an intelligent wearable scenario controller is proposed. Its main method is to detect at least one of three physiological information (body temperature, respiration, heart function) of the human body and environmental information, and then convert them into digital signals to control a virtual scene. Although the controlled scenarios it proposes are more diverse, firstly, it only detects three specific physiological signals, and at the same time, it only gives an application form of collecting data and implementing intervention, and does not involve the relevant mechanisms, structures, and processes regarding the specific intervention method.

[0006] The present application proposes a goal-oriented physiological intervention method and system. In addition to performing common physiological-oriented analysis in data evaluation, it also introduces goal-oriented analysis. Combining with a prediction module, the intervention module gives an intervention strategy at a certain time point or time period in the form of closed-loop feedback, achieving the goal of dynamically, real-timely, and effectively intervening and regulating a target individual for a specific purpose.

[0007] In the present application, physiological information refers to various signals generated by the physiological activities of a living individual itself that can be digitized. It not only includes physiological signals detected by common medical devices, such as body temperature, blood pressure, blood oxygen, electroencephalogram, electrocardiogram, electromyogram, nasal airflow, respiration, etc., but also includes waves formed by vibrating air generated by the active or passive activities of the living individual, heat release generated by physiological activities, and thermal radiation generated thereby.

[0008] Intervention implementation operation actions, such as the release of odors with different components, air flow with a certain pattern, temperature adjustment at different levels, electrical signals at different levels, etc., various stimuli with a certain pattern, such as vibration, hitting, spatial position transformation, gravity transformation, epidermal or implanted electrical signal stimulation, magnetic field change, taking food, taking or injecting drugs, and can also be various combinations of the above signal changes and operations. Summary of the Invention

[0009] The technical problem to be solved by the present invention is to avoid the deficiencies of the above-mentioned prior art solutions, and propose a goal-oriented physiological intervention method and system, so that the physiological intervention method has goal orientation and can realize a real-time closed-loop physiological intervention control process.

[0010] The technical solution of the present invention to solve the above problems is an objective - oriented physiological intervention method, including the following steps: Step C of data evaluation implementation method planning: Take out the recently collected and classified physiological signal data from the original data pool, and plan the data evaluation implementation method according to the data type information of the current physiological signal data; The data evaluation implementation method planning includes the selection of analysis models; The analysis models include a physiological - oriented analysis model, a purpose - requirement conversion model, an objective - oriented analysis model, an intervention decision - making model, and a security mechanism model; When each analysis model is selected, the input - output data forms of each analysis model are also determined accordingly; Step D of analysis model loading: The physiological - oriented analysis model module loads the physiological - oriented analysis model, the purpose - requirement conversion model module loads the purpose - requirement conversion model, and the objective - oriented analysis model module loads the objective - oriented analysis model; The intervention decision - making model module in the intervention decision - making module loads the intervention decision - making model, and the security mechanism model module in the intervention decision - making module loads the security mechanism model; Step E of data pre - processing and diversion: The pre - processing and diversion module outputs corresponding data to the physiological - oriented analysis model module and the purpose - requirement conversion model module according to the types and contents of the input feature data required by each analysis model given in the data processing planning module; Step F: The physiological - oriented analysis model module conducts physiological - oriented analysis on the data transmitted by the pre - processing and diversion module. The analysis result data of the physiological - oriented analysis model module expresses the physiological signal information, characteristics, and / or trends of the target individual in medicine; Denote the data set output by the physiological - oriented analysis model module as the physiological - oriented data set Med t ; The physiological - oriented data set Med t is output to the objective - oriented analysis model module; Step G: Use the purpose - requirement conversion model module to perform purpose - requirement conversion on the data type transmitted by the pre - processing and diversion module; The purpose - requirement conversion model module obtains the current purpose - requirement through the purpose - requirement management module. The purpose - requirement conversion model module converts the purpose - requirement information into a data expression with the same dimension as the physiological - oriented data set Med t , denoted as the target - oriented data set Targ t ; Step H: The objective - oriented analysis model module analyzes the physiological - oriented data set Med t and the target - oriented data set Targ t according to the intervention methods provided by the intervention management module in the external intervention module, and outputs the first intervention plan data set, denoted as Sch1 t， The first intervention plan data set Sch1 t includes at least one basic intervention method and basic intervention rule information; Each intervention method is assigned an initial weight to indicate the application proportion of the intervention method; Step I: The physiological - oriented analysis model module outputs the physiological - oriented data set Medt To the intervention decision-making module, the purpose-oriented analysis model module outputs the first intervention plan dataset Sch1 t To the intervention decision-making module; the purpose requirement conversion model module outputs the target-oriented data set Targ t To the intervention decision-making module; the predictive analysis model module outputs the predictive physiological-oriented change amount data set ΔMed t Output to the intervention decision-making module; the intervention decision-making module outputs the second intervention plan dataset Sch2 according to the intervention decision-making model and the security mechanism model t To the intervention module; Step J: The intervention module selects the current intervention method and intervention mechanism according to the intervention strategy generated by the intervention decision-making module, that is, the second intervention plan dataset Sch2 t to select the current intervention method and intervention mechanism to implement the intervention measures.

[0011] The described purpose-oriented physiological intervention method further includes Step K and Step L. In Step K, a prediction module is set up. The prediction module includes a historical prediction data module, a prediction model management module, a predictive analysis model module, and a data matching and extraction module; in Step C of the data evaluation implementation method planning, the analysis model selection also includes the predictive analysis model selection; in Step D of the analysis model loading, the predictive analysis model module loads the predictive analysis model; in Step F: the physiological-oriented data set Med t is also output to the prediction module; in Step H: the first intervention plan data set Sch1 t is also output to the prediction module; in Step L, the physiological-oriented data set Med output by the physiological-oriented analysis model module t , the first intervention plan data set Sch1 output by the purpose-oriented analysis model module t are respectively input into the predictive analysis model module and the data matching and extraction module in the prediction module; the historical prediction data module stores the evaluation data information, data evaluation implementation methods, and intervention implementation plan information of each target individual after the system starts running. The historical prediction data module also stores the evaluation data information of the target individual obtained at the (m + 1)-th time after the analysis and intervention are performed according to the data information of the target individual obtained at the m-th time. Here, m is a certain time point or time period earlier than t and does not include t, that is ; the data matching and extraction module compares the historical prediction data with the physiological-oriented data set Med t , the first intervention plan data set Sch1 t and extracts the data generated by the same data analysis model from the historical prediction data module; that is, the data matching and extraction module extracts the data matched as the same type from the historical prediction data module; the historical data set corresponding to the time interval [t - i, t) can be obtained from the historical prediction data, and it is defined represents the time series value of each historical record between the i-th time before the current time mark t and the current time, and does not include t. The historical data set includes the historical physiological guidance data set Med output by the physiological guidance analysis model j and the historical first intervention program data set Sch1 j , and the historical second intervention plan dataset Sch2 output by the intervention decision module j ; After extracting historical data and Med t ,Sch1 t The prediction analysis model module performs prediction analysis based on a set of time series data sets. The set of time series data sets includes a physiological guidance data set Med at the current time point or time period. t , the first intervention plan data set Sch1 at the current time point or time period t , historical physiologically oriented data set Med j , the first historical intervention plan data set Sch1 j and the historical second intervention program data set Sch2 j ; Output of the prediction analysis model module, in the first intervention scenario data set Sch1 t Under the given intervention method, the change of each basic intervention rule contains the corresponding predicted physiological guidance change data set ΔMed t , output to the intervention decision module.

[0012] The goal-oriented physiological intervention method further includes step M: the safety mechanism model module in the intervention decision module is responsible for determining the first intervention scheme data set Sch1 to be implemented. t Different intervention methods in the control intervention rules to implement the safety limit; the safety mechanism model module predicts the physiological orientation change data set ΔMed t , physiologically oriented data set Med t and the first intervention scenario data set Sch1 t , output the safety limit data set Thres of the intervention rule t ; Safety limit data set Thres t Includes static security boundary data set Thres t , also includes real-time dynamic security boundary data set Thres t The intervention decision model module in the intervention decision module is used to predict the input physiological orientation change data set ΔMed t , physiologically oriented data set Med t and the first intervention scenario data set Sch1 t, comprehensively analyze the current purpose requirements provided by the purpose requirements manager, and use the intervention rule safety limit data set Thres t as the safety threshold of the intervention rule, and adjust the intervention rules in the first intervention plan data set Sch1 t to output the second intervention plan data set Sch2 t； Linearly adjust all the intervention rules in the first intervention plan data set Sch1 t . The adjustment method is expressed by the formula:

[0013] ,

[0014] Non-linearly adjust all the intervention rules in the first intervention plan data set Sch1 t ; The second intervention plan data set Sch2 t , on the one hand, output it to the intervention module to execute the actual intervention, and on the other hand, output it to the prediction module and save it in the historical prediction data module.

[0015] Before step C of the described purpose-oriented physiological intervention method, it also includes: Step A: Collect the physiological signal data of the target individual; Step B: Classify and save the collected physiological signal data in the original data pool.

[0016] The technical solution of the present invention to solve the above problems can also be a purpose-oriented physiological intervention system, including a data evaluation module, an intervention decision module, a purpose requirements management module, and an intervention module; the purpose requirements management module is used to save and maintain the content of one or more purpose requirements expected to be achieved through intervention. The purpose requirements management module selects the current purpose requirement information of the system in a static or dynamic manner, and the purpose requirements management module transmits the purpose requirement information to the data evaluation module; the data evaluation module uses the purpose requirement information as one of the input parameters for data evaluation; the intervention decision module uses the purpose requirement information as one of the input parameters for formulating intervention decisions; the intervention decision module combines the information provided by the data evaluation module for analysis and outputs specific intervention strategies to the intervention module; the intervention module executes corresponding intervention measures according to the intervention strategies output by the intervention decision module.

[0017] The data evaluation module includes an original data pool, a data processing planning module, a preprocessing and diversion module, a physiological orientation model management module, a physiological orientation analysis model module, a purpose orientation model management module, a purpose orientation analysis model module, and a purpose requirement conversion model module; the physiological orientation model management module is used to manage and maintain the physiological orientation analysis models used by the physiological orientation analysis model module; the purpose orientation model management module is used to manage and maintain the purpose orientation analysis models used by the purpose orientation analysis model module; in the physiological orientation analysis model and the purpose orientation analysis model, the model information at least includes the model type, the types of input parameter features, and the model configuration information; the data processing planning module is used to plan the data evaluation implementation method, and the data evaluation implementation method planning includes the selection of the analysis model to be used; the analysis model selection includes the selection of the physiological orientation analysis model, the purpose requirement conversion model, and the purpose orientation analysis model in the data evaluation module; the preprocessing and diversion module is used to obtain the preprocessing instruction information from the data processing planning module, the preprocessing and diversion module preprocesses the information in the original data pool with the obtained data analysis model information, and outputs the corresponding data analysis model information according to the data processing planning module to output the corresponding data to the physiological orientation analysis model module and the purpose requirement conversion model module; the physiological orientation analysis model module loads the selected analysis model through the physiological orientation model management module, and performs physiological orientation analysis on the data transmitted by the preprocessing and diversion module. The analysis data result of the physiological orientation analysis model module expresses the physiological signal information, characteristics, and / or trends of the target individual in medicine; the purpose orientation model management module is used to save and maintain one or more purpose requirement conversion model information and purpose orientation analysis model information; the purpose requirement conversion model information and the purpose orientation analysis model information at least include the model type, the types of input parameter features, and the model configuration information; the data update module, through the purpose orientation model management module, adds, deletes, or modifies the purpose requirement conversion model and the purpose orientation analysis model maintained by the purpose orientation model management module to the data evaluation module; the purpose requirement conversion model module loads the selected evaluation model through the purpose orientation model management module, and performs purpose requirement conversion on the data transmitted by the preprocessing and diversion module; the purpose requirement conversion model module obtains the current purpose requirement through the purpose requirement management module, and then converts the purpose information into a physiological orientation data set Med t of the same dimension data expression, denoted as the target orientation data set Targ t .

[0018] The intervention decision-making module includes an intervention and security mechanism management module, a security mechanism model module, and an intervention decision-making model module; when planning the implementation method of data evaluation in the data processing planning module, the selection of the analysis model also includes the selection of the intervention decision-making model and the security mechanism model in the intervention decision-making module; the intervention and security mechanism management module is used to save and maintain information on one or more intervention decision-making models and security mechanism models; the intervention and security mechanism management module selects the currently loaded intervention decision-making model and security mechanism model according to the notification information of the data processing planning module in the data evaluation module; the security mechanism model module controls the security boundary for implementing the intervention rules according to different upcoming intervention methods, and outputs the intervention rule security boundary Thres t , the physiological orientation data set Med t and the first intervention plan data set Sch1 t , and outputs the intervention rule security boundary Thres t ; the security mechanism model includes a static security mechanism model and a dynamic security mechanism model; the intervention decision-making model module is used to comprehensively analyze the input predicted physiological orientation change data set ΔMed t , the physiological orientation data set Med t and the first intervention plan data set Sch1 t with the current purpose requirements provided by the purpose requirements manager, outputs the security boundary data set Thres t , and uses the security boundary data set Thres t as the security threshold for the intervention rules to adjust the intervention rules in the first intervention plan data set Sch1 t , and outputs the second intervention plan data set Sch2 t ; the intervention decision-making model module converts the purpose requirements obtained from the purpose requirements management module into information with the same dimension as the physiological orientation data set Med t , that is, the target orientation data set Targ t , then compares it with the physiological orientation data set Med t and the predicted physiological orientation change data set ΔMed t , and adjusts according to the rules in the first intervention plan data set Sch1 t to obtain the second intervention plan data set Sch2 t .

[0019] The described goal-oriented physiological intervention system further includes a prediction module. The prediction module is used to predict the effects of the upcoming intervention methods, providing reference data for the adjustment of the intervention strategies output by the intervention decision-making module. At the same time, the prediction module will also save the intervention strategies that have been implemented and the intermediate data information generated before obtaining these intervention strategies, as reference data for prediction analysis. The prediction module includes a historical prediction data module, a data matching and extraction module, a prediction analysis model module, and a prediction model management module. The physiological orientation data set Med output by the data evaluation module t , the first intervention plan data set Sch1 t , on the one hand, will be input into the prediction analysis model module in the prediction module, and on the other hand, will be input into the data matching and extraction module in the prediction module together with the historical prediction data. When planning the implementation method of data evaluation in the data processing and planning module, the selection of the analysis model also includes notifying the prediction model management module in the prediction module to select the corresponding prediction analysis model module. The historical prediction data module stores the information of each data evaluation, intervention implementation plan, and the evaluation data information of the target individual obtained again after implementing this intervention since the system started running. From the historical prediction data, it includes the historical data set corresponding to the time interval [t-i, t). Define to represent the time series value of each historical record from the i-th time before the current time mark t to the current time, and does not include t. That is, the historical data set includes the historical physiological orientation data set Med output by the physiological orientation analysis model j and the historical first intervention plan data set Sch1 j , as well as the historical second intervention plan data set Sch2 output by the intervention decision-making module j ; The data matching and extraction module compares the historical prediction data with the physiological orientation data set Med t , the first intervention plan data set Sch1 t , and extracts the data generated by the same data analysis model from the historical prediction data module. That is, the data matching and extraction module extracts the data matched as "the same type" from the historical prediction data module; The prediction analysis model module conducts prediction analysis based on a set of time series data sets. This set of time series data sets includes the historical physiological orientation data set Med output by the physiological orientation analysis model j and the historical first intervention plan data set Sch1 j , as well as the historical second intervention plan data set Sch2 output by the intervention decision-making module j , the current physiological orientation data set Med t , the current first intervention plan data set Sch1 t; The output of the predictive analysis model module, under the intervention method given in the current first intervention plan data set Sch1 t The changes of each basic intervention rule included under the given intervention method, generating the corresponding physiological guidance data set Med t The predictive physiological guidance change amount data set ΔMed t ; The predictive physiological guidance change amount data set ΔMed t , on the one hand, output to the intervention decision-making module, and on the other hand, output to the historical prediction data module for storage.

[0020] The intervention module includes an intervention management module, an intervention driving module, and an intervention execution module; the intervention management module is used to manage one or more intervention method data used in the system; the intervention management module selects the current intervention method and intervention mechanism to be executed according to the intervention strategy generated by the intervention decision-making module; the intervention management module updates the corresponding intervention method data from the external data update module according to the actual configuration of the intervention devices in the system.

[0021] The described goal-oriented physiological intervention system further includes: a data acquisition module and a data update module; the data acquisition module is used to acquire physiological signal data of a target individual and transmit the acquired physiological signal data to the data evaluation module; the data evaluation module is used to classify and save the physiological signal data and perform evaluation operations; the data update module is used to interface with an external expert system. On the one hand, the data update module is used to obtain the required analysis model information from the external expert system. On the other hand, the data update module is used to transmit various historical data stored in the goal-oriented physiological intervention system to the outside; the data update module is used to connect to an external expert system to obtain various analysis models required by the goal-oriented physiological intervention system; the data update module is also used to transmit various historical data stored in the goal-oriented physiological intervention system to the outside; the data update module is connected to the data processing and planning module, and the data update module is also respectively connected to the physiological guidance model management module, the goal-oriented model management module, the intervention and safety mechanism management module, and the prediction model management module; the data update module, through the data processing and planning module or the physiological guidance model management module, adds, deletes, or modifies one or more physiological guidance analysis models maintained by the physiological guidance model management module to the data evaluation module; the data update module, through the data processing and planning module or the goal-oriented model management module, adds, deletes, or modifies one or more goal-oriented models maintained by the goal-oriented model management module to the data evaluation module; the data update module, through the data processing and planning module or the intervention and safety mechanism management module, adds, deletes, or modifies one or more safety mechanism models and intervention decision models maintained by this module to the intervention decision module; the data update module, through the data processing and planning module or the prediction model management module, adds, deletes, or modifies one or more prediction analysis models maintained by the prediction model management module to the data evaluation module.

[0022] Compared with the prior art, one of the beneficial effects of this application is that: through the goal-oriented analysis model module, the goal demand and physiological information characteristics are directly associated using artificial intelligence algorithms, so that the physiological intervention method has goal orientation and can achieve real-time closed-loop physiological intervention control. The system has the closed-loop control ability to establish a feedback loop, so that the user's needs can be imported into the system in real time through the goal demand management module, and the system can perform real-time response operations. And the intervention process from data acquisition, data analysis, data prediction, intervention decision-making, intervention, and then to data acquisition has clear records and feedback, so that a real-time closed-loop is formed between the intervention goal and the intervention measures, greatly improving the self-adaptability and intervention efficiency of the physiological intervention method.

[0023] Compared with the prior art, the second beneficial effect of the present application is that the data evaluation implementation method planning can effectively interact with an external expert system to obtain the AI algorithm analysis models required for each link in the physiological intervention method, enabling the physiological intervention method to update the algorithm modules, maintaining the operation of the system and method under the latest and most effective models, and loading the same or different models under different data acquisitions and different intervention purposes.

[0024] Compared with the prior art, the third beneficial effect of the present application is that the data evaluation implementation method planning module can plan the upcoming data processing flow according to the information of each AI model in the system, meeting the data input type requirements of each AI model, that is, realizing the data flow matching between multiple AI models, enabling the system to accommodate and coordinate more types of AI models, thereby improving the adaptability of the system to data processing from various sources, achieving the adaptive matching between the diversification of physiological parameter acquisition devices and the diversification of AI analysis modules, improving the compatibility of the system, and broadening the application scope of the system.

[0025] Compared with the prior art, the fourth beneficial effect of the present application is that through the purpose requirement conversion model module, using artificial intelligence technology, from the perspective of data processing, one or more physiological signal data are more directly associated with the purposes to be achieved. For a mature artificial intelligence algorithm in a design framework, changes in multiple purpose requirements within a certain range usually only require the process of collecting corresponding training data for training for the artificial intelligence algorithm, and the algorithm adjustment cost due to changes in purpose requirements is low; while the traditional method usually designs a specific classical processing algorithm according to the specific indicators of one or several specified purpose requirements to achieve. Once the specific indicators of the purpose requirements change, a new algorithm needs to be redesigned specifically, and the algorithm adjustment cost required for the change is high, and the greater the complexity of the purpose requirements, the greater the gap in input costs between the two. The technical solution in the present application is an efficient purpose-oriented physiological intervention method and system.

[0026] Compared with the prior art, the fifth beneficial effect of the present application is that the intervention decision model and the safety mechanism model module in the intervention decision module cooperate together to output the safety limit data set Thres t and serve as an important basis for outputting the second intervention plan data set Sch2 t to ensure the safety of the intervention.

[0027] Compared with the prior art, the sixth beneficial effect of the present application is as follows: A prediction module is provided, which can make full use of historical data, use a prediction analysis model to predict the effect of intervention, and can adjust the intervention plan in real time according to the prediction result, further improving the real-time performance and adaptability of the system and method, and also increasing the reliability of the system and method, and avoiding some abnormal interventions. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 is a schematic block diagram of a preferred embodiment of an objective-oriented physiological intervention system;

[0029] Figure 2 is a functional schematic block diagram of a data evaluation module;

[0030] Figure 3 is a functional schematic block diagram of an intervention decision module;

[0031] Figure 4 is a functional schematic block diagram of an intervention module;

[0032] Figure 5 is a functional schematic block diagram of a prediction module. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0033] The following further details the content of the present application in conjunction with the accompanying drawings.

[0034] As Figures 1 to 5 In an embodiment of an objective-oriented physiological intervention system shown, it includes a data acquisition module, a data evaluation module, a prediction module, an intervention decision module, an objective requirement management module, an intervention module, and a data update module. The data acquisition module is used to acquire physiological signal data of a target individual and transmit the acquired physiological signal data to the data evaluation module; the data evaluation module is used to classify and save the physiological signal data and perform evaluation operations.

[0035] The principle of classifying and saving physiological signal data is that the physiological signal data collected by different signal sensors are each separately classified into a category and additional sensor-related parameter information, such as hardware parameters, working mode, sampling frequency, and detection site information, is used as the label of this category of data, and it is saved in the form of a time series. The data storage method can be non-volatile storage (referring to a storage method that can continue to retain data after power failure), volatile storage (referring to a storage method that will clear the saved data after power failure), or a combination of both.

[0036] As Figures 1 to 5In an embodiment of a goal - oriented physiological intervention system shown, the goal requirement management module is used to save and maintain the content of one or more goal requirements to be achieved through intervention. For example, a goal requirement content regarding sleep is that when the intervened individual has been in the sleep state for a certain period of time, to keep them in the light sleep state, and this goal requirement content is digitized as [Goal Requirement Content 0001]. The goal requirement management module selects the current goal requirement information of the system in a static or dynamic manner. For example, the goal requirement management module digitizes the current goal requirement information, [Goal Requirement Content 0001, maintenance duration x]; the specific change mechanism of the goal requirement management module is maintained by the goal requirement management module and can be managed by the data update module for updating management rules and goal change mechanisms; this mechanism enables the system to have the closed - loop control ability to establish a positive feedback loop, so that the user's needs can be imported into the system in real - time through the goal requirement management module, allowing the system to perform real - time response operations. The goal requirement management module transmits the goal requirement information to the data evaluation module.

[0037] As Figures 1 to 5 In an embodiment of a goal - oriented physiological intervention system shown, the data evaluation module includes a raw data pool, a data processing planning module, a pre - processing and diversion module, a physiological orientation model management module, a physiological orientation analysis model module, a goal orientation model management module, a goal orientation analysis model module, and a goal requirement conversion model module; the data evaluation module uses the goal requirement information as one of the input parameters for data evaluation.

[0038] The principle of classifying and saving physiological signal data in the raw data pool is as follows: Physiological signal data collected by different signal sensors are each classified separately and attached with sensor - related parameter information and detection site information as labels for this type of data, and are saved in the form of a time series. The sensor - related parameter information includes hardware parameters, working modes, and sampling frequencies.

[0039] The physiological orientation model management module is used to manage and maintain the physiological orientation analysis models used by the physiological orientation analysis model module; the goal orientation model management module is used to manage and maintain the goal orientation analysis models used by the goal orientation analysis model module; in the physiological orientation analysis model and the goal orientation analysis model, the model information includes at least model type, input parameter feature types, and model configuration information.

[0040] The data processing planning module is used to plan the implementation method of data evaluation. The implementation method of data evaluation includes the selection of analysis models; the selection of analysis models includes the selection of the physiological orientation analysis model, the goal requirement conversion model, and the goal orientation analysis model in the data evaluation module.

[0041] The preprocessing and diversion module is used to obtain preprocessing instruction information from the data processing planning module. The preprocessing and diversion module preprocesses the information in the original data pool with the obtained data analysis model information, and outputs the corresponding data analysis model information according to the data processing planning module to output the corresponding data to the physiological orientation analysis model module and the purpose requirement conversion model module.

[0042] The physiological orientation analysis model module loads the selected analysis model through the physiological orientation model management module, and conducts physiological orientation analysis on the data transmitted by the preprocessing and diversion module. The data analysis results of the physiological orientation analysis model module express the physiological signal information, characteristics, and trends of the target individual in medicine. Denote the data set output by the physiological orientation analysis model module as the physiological orientation data set Med t , t represents the current time point or time period, and the physiological orientation data set Med t represents the digital interpretation of the physiological performance of the target individual in medicine within the current time point or time period t. For example, Med t is [current physiological signal or physiological signal time series 1,…,n, [current physiological performance 1]].

[0043] The purpose-oriented model management module is used to save and maintain one or more purpose requirement conversion model information and purpose-oriented analysis model information; the purpose requirement conversion model information and purpose-oriented analysis model information at least include model type, input parameter feature types, and model configuration information; the data update module, through the purpose-oriented model management module, adds, deletes, or modifies the purpose requirement conversion model and purpose-oriented analysis model maintained by the purpose-oriented model management module to the data evaluation module.

[0044] The purpose requirement conversion model module loads the selected evaluation model through the purpose-oriented model management module, and conducts purpose requirement conversion on the data transmitted by the preprocessing and diversion module; the purpose requirement conversion model module obtains the current purpose requirement through the purpose requirement management module, and then converts the purpose information into a data expression of the same dimension as the physiological orientation data set Med t , denoted as the target-oriented data set Targ t , for example, Targ t is [target physiological signal or physiological signal sequence 1,…,n, [target physiological performance 1]].

[0045] Such as Figures 1 to 5In an embodiment of a goal - oriented physiological intervention system shown, the intervention decision - making module includes an intervention and safety mechanism management module, a safety mechanism model module, and an intervention decision - making model module; the intervention decision - making module takes the goal - demand information as one of the input parameters for formulating intervention decisions; the intervention decision - making module analyzes by combining the information provided by the data evaluation module and outputs specific intervention strategies to the intervention module; the intervention module executes corresponding intervention measures according to the intervention strategies output by the intervention decision - making module. When planning the implementation method of data evaluation in the data processing and planning module, the selection of the analysis model also includes the selection of the intervention decision - making model and the safety mechanism model in the intervention decision - making module.

[0046] The intervention and safety mechanism management module is used to save and maintain information of one or more intervention decision - making models and safety mechanism models; the intervention and safety mechanism management module selects the currently to - be - loaded intervention decision - making model and safety mechanism model according to the notification information of the data processing and planning module in the data evaluation module.

[0047] The safety mechanism model module controls the safety boundary for implementing intervention rules according to different upcoming intervention methods, and outputs the intervention rule safety boundary Thres t based on the input physiological - orientation change - amount data set ΔMed t , physiological - orientation data set Med t and the first intervention plan data set Sch1 t ; the safety mechanism model includes a static safety mechanism model and a dynamic safety mechanism model.

[0048] The intervention decision - making model module is used to comprehensively analyze the input predicted physiological - orientation change - amount data set ΔMed t , physiological - orientation data set Med t and the first intervention plan data set Sch1 t with the current goal - demand provided by the goal - demand manager, outputs the safety - boundary data set Thres t , and uses the safety - boundary data set Thres t as the safety threshold of the intervention rule to adjust the intervention rules in the first intervention plan data set Sch1 t and outputs the second intervention plan data set Sch2 t .

[0049] The intervention decision - making model module first converts the goal - demand obtained from the goal - demand management module into information with the same dimension as the physiological - orientation data set Med t , that is, the target - orientation data set Targ t , then combines it with the physiological - orientation data set Med t and the predicted physiological - orientation change - amount data set ΔMedt Compare and adjust according to the rules in the first intervention plan data set Sch1 t to obtain the second intervention plan data set Sch2 t .

[0050] As Figures 1 to 5 shown in an embodiment of a goal-oriented physiological intervention system, the prediction module includes a historical prediction data module, a data matching and extraction module, a prediction analysis model module, and a prediction model management module; when planning the implementation method of data evaluation in the data processing and planning module, the selection of the analysis model also includes notifying the prediction model management module in the prediction module to select the corresponding prediction analysis model module. The physiological orientation data set Med t output by the data evaluation module, the first intervention plan data set Sch1 t , on the one hand, will be input into the prediction analysis model module in the prediction module, and on the other hand, will be input into the data matching and extraction module in the prediction module together with the historical prediction data.

[0051] As Figures 1 to 5 shown in an embodiment of a goal-oriented physiological intervention system, the historical prediction data module stores the evaluation data information, data evaluation implementation method, and intervention implementation plan information of each target individual after the system starts running. The historical prediction data module also stores the evaluation data information of the target individual obtained at the (m + 1)-th time after the analysis and intervention are implemented based on the data information of the target individual obtained at the m-th time, where m is a certain time point or time period earlier than t and does not include t, that is ; The historical data set corresponding to the time interval [t - i, t) can be obtained from the historical prediction data. Define to represent the time series value of each historical record from the i-th time before the current time mark t to the current time and does not include t. Then the historical data set includes the historical physiological orientation data set Med j output by the physiological orientation analysis model and the historical first intervention plan data set Sch1 j , and the historical second intervention plan data set Sch2 j output by the intervention decision module; after extracting the historical data, it is used as input data together with Med t , Sch1 t and input into the prediction analysis model module.

[0052] As Figures 1 to 5In an embodiment of a goal-oriented physiological intervention system shown, the intervention module includes an intervention management module, an intervention driving module, and an intervention execution module; the intervention management module is used to manage data of one or more intervention methods used in the system; the intervention management module selects the current intervention method and intervention mechanism to be executed according to the intervention strategy generated by the intervention decision module; the intervention management module updates the corresponding intervention method data from the external data update module according to the actual configuration of the intervention devices in the system.

[0053] As Figures 1 to 5 In an embodiment of a goal-oriented physiological intervention system shown, it further includes: a data acquisition module and a data update module; the data acquisition module is used to acquire physiological signal data of a target individual and transmit the acquired physiological signal data to the data evaluation module; the data evaluation module is used to classify and save the physiological signal data and perform evaluation operations; the data update module is used to interface with an external expert system. On the one hand, the data update module is used to obtain the required analysis model information from the external expert system; on the other hand, the data update module is used to transmit various historical data saved in the goal-oriented physiological intervention system to the outside; the data update module is used to connect with the external expert system to obtain various analysis models required by the goal-oriented physiological intervention system; the data update module is also used to transmit various historical data saved in the goal-oriented physiological intervention system to the outside; the data update module is connected to the data processing and planning module, and the data update module is also respectively connected to the physiological orientation model management module, the goal orientation model management module, the intervention and safety mechanism management module, and the prediction model management module; the data update module adds, deletes, or modifies one or more physiological orientation analysis models maintained by the physiological orientation model management module to the data evaluation module through the data processing and planning module or the physiological orientation model management module; the data update module adds, deletes, or modifies one or more goal orientation models maintained by the goal orientation model management module to the data evaluation module through the data processing and planning module or the goal orientation model management module; the data update module adds, deletes, modifies one or more safety mechanism models and intervention decision models maintained by the intervention decision module to the intervention decision module through the data processing and planning module or the intervention and safety mechanism management module; the data update module adds, deletes, or modifies one or more prediction analysis models maintained by the prediction model management module to the data evaluation module through the data processing and planning module or the prediction model management module.

[0054] As Figures 1 to 5 In an embodiment of a goal-oriented physiological intervention method shown, it includes the following steps: Step A: Acquire physiological signal data of a target individual.

[0055] Step B: Classify and save the acquired physiological signal data in the original data pool;

[0056] Step C of the implementation method planning for data evaluation: Retrieve the recently collected and classified physiological signal data from the original data pool, and plan the data evaluation implementation method according to the data type information of the current physiological signal data. The data evaluation implementation method planning includes the selection of analysis models; the analysis models include a physiology-oriented analysis model, a purpose requirement conversion model, a purpose-oriented analysis model, an intervention decision model, and a security mechanism model.

[0057] Step D of the analysis model loading: The physiology-oriented analysis model module loads the physiology-oriented analysis model, the purpose requirement conversion model module loads the purpose requirement conversion model, and the purpose-oriented analysis model module loads the purpose-oriented analysis model; the intervention decision model module in the intervention decision module loads the intervention decision model, and the security mechanism model module in the intervention decision module loads the security mechanism model.

[0058] Step E of data preprocessing and diversion: The preprocessing and diversion module outputs corresponding data to the physiology-oriented analysis model module and the purpose requirement conversion model module according to the types and contents of input feature data required by the subsequent analysis models given in the data processing planning module.

[0059] Step F: The physiology-oriented analysis model module performs a physiology-oriented analysis on the data transmitted by the preprocessing and diversion module. The analysis data result of the physiology-oriented analysis model module expresses the physiological signal information, characteristics, and / or trends of the target individual in medicine; record the data set output by the physiology-oriented analysis model module as the physiology-oriented data set Med t , t represents the current time point or time period, and the physiology-oriented data set Med t represents the digital interpretation of the physiological performance of the target individual in medicine within the current time point or time period t; within the data evaluation module, the physiology-oriented data set Med t is output to the purpose-oriented analysis model module; at the same time, the data evaluation module outputs the physiology-oriented data set Med t to the prediction module and the intervention decision module respectively.

[0060] In this application, the physiology-oriented analysis model, the purpose requirement conversion model, and the purpose-oriented analysis model are generated based on supervised deep learning through pre-learning and training.

[0061] Step G: Use the purpose requirement conversion model module to perform purpose requirement conversion on the data type transmitted by the preprocessing and diversion module; the purpose requirement conversion model module obtains the current purpose requirement through the purpose requirement management module, and the purpose requirement conversion model module converts the purpose requirement information into a form related to the physiology-oriented data set Med according to the purpose requirement conversion model tData representation in the same dimension, denoted as the target-oriented data set Targ t ; The target-oriented data set Targ t includes multiple physiological-oriented data sets Med t ; The purpose requirement conversion model module compares the target-oriented data set Targ t with the physiological-oriented data set Med t to evaluate how much deviation remains from completion, and this deviation data set is denoted as the target-oriented deviation data set ΔTarg t and output to the purpose-oriented analysis model module.

[0062] For example, if the content of the physiological-oriented data set Med t is:

[0063] [Current physiological signal 1,…,n,[Current physiological manifestation 1]];

[0064] then the purpose information is converted into the target-oriented data set Targ t :

[0065] [Target physiological signal 1,…,n,[Target physiological manifestation 1]].

[0066] Then the target-oriented data set Targ t is compared with the physiological-oriented data set Med t to evaluate how much deviation remains from completion, and this data set is denoted as the target-oriented deviation data set ΔTarg t and output to the purpose-oriented analysis model.

[0067] Step H: The purpose-oriented analysis model module analyzes the physiological-oriented data set Med t , the target-oriented deviation data set ΔTarg t according to the intervention method provided by the intervention management module in the external intervention module, and outputs the first intervention plan data set denoted as Sch1 t， The first intervention plan data set Sch1 t includes the basic intervention method, basic intervention rule information and at least one intervention method; See the specification: Each intervention method will be assigned an initial weight to identify the application proportion of this intervention method.

[0068] The intervention plan data set Sch1 t is expressed as:

[0069] [Intervention method 1: Weight 1: [Intervention rules 1,..,n],…, Intervention method m: Weight m: [Intervention rules 1,..,n]]; The first intervention plan data set Sch1 tOutput to the prediction module and the intervention decision-making module respectively. Both n and m are natural numbers, and the natural numbers between 1 and n are omitted.

[0070] Step I: The physiological orientation analysis model module outputs the physiological orientation data set Med t to the intervention decision-making module, and the goal orientation analysis model module outputs the first intervention plan data set Sch1 t to the intervention decision-making module; the goal requirement conversion model module outputs the goal orientation data set Targ t to the intervention decision-making module; the prediction analysis model module outputs the predicted physiological orientation change amount data set ΔMed t Output to the intervention decision-making module; the intervention decision-making module outputs the second intervention plan data set Sch2 t based on the intervention decision-making model and the safety mechanism model to the intervention module. The intervention decision-making module also outputs Sch2 t to the historical data prediction module of the prediction module for storage as historical data.

[0071] Step J: The intervention module implements the intervention measures according to the second intervention plan data set Sch2 t

[0072] Step K: Set the prediction module, which includes a historical prediction data module, a prediction model management module, a prediction analysis model module, and a data matching and extraction module; in step C of the data evaluation implementation method planning, the analysis model selection also includes the prediction analysis model selection; in step D of the analysis model loading, the prediction analysis model module loads the prediction analysis model; in step F, the physiological orientation data set Med t is also output to the prediction module; in step H: the first intervention plan data set Sch1 t is also output to the prediction module.

[0073] Step L: The physiological orientation data set Med output by the physiological orientation analysis model module t , and the first intervention plan data set Sch1 output by the goal orientation analysis model module t are both input to the prediction analysis model module and the data matching and extraction module in the prediction module; the historical prediction data module stores the evaluation data information, data evaluation implementation methods, and intervention implementation plan information of each target individual after the system starts running, and the historical prediction data module also stores the evaluation data information of the target individual obtained again after the current intervention is implemented; the data matching and extraction module matches the historical prediction data with the physiological orientation data set Med t , the first intervention plan data set Sch1 tFor comparison, the data generated by the same data analysis model is extracted from the historical prediction data module; that is, the data matching and extraction module extracts the data matched as "the same type" from the historical prediction data module;

[0074] The historical data set corresponding to the time interval [t - i, t) can be obtained from the historical prediction data, and it is defined as representing the time series values of each historical record from the i-th time before the current time mark t to the current time. Then the historical data set includes the historical physiological orientation data set Med j output by the physiological orientation analysis model and the historical first intervention plan data set Sch1 j , as well as the historical second intervention plan data set Sch2 j output by the intervention decision-making module; the prediction analysis model module takes the historical data set and the current data set as input data, that is, Med t , Sch1 t , Med t-1 , Sch1 t-1 , Sch2 t-1 ,…, Med t-i , Sch1 t-i , Sch2 t-i for prediction analysis; output the changes in each basic intervention rule included under the intervention method given by Sch1 t , and the predicted physiological orientation change data set ΔMed t of the corresponding Med t change amount. There are many reference implementation methods for predicting and analyzing a set of time series data, which will not be elaborated here. The predicted physiological orientation change data set ΔMed t is output to the intervention decision-making module, and the predicted physiological orientation change data set ΔMed t is also output to the historical prediction data module for storage as historical data.

[0075] In a specific scenario application, taking sports fitness as an example:

[0076] Purpose requirement: Moderate exercise for 30 minutes;

[0077] Intervention method: Adjust the treadmill speed;

[0078] Data collection: Electrocardiogram, body temperature;

[0079] Data planning: Load the models of each module and determine the input data requirements of each module.

[0080] The data processing planning module first determines the type range of each analysis module that needs to be loaded based on the current system's available (already mounted in the system and effective) intervention methods and data acquisition sensor information, and then further screens out the specific modules that need to be loaded based on the purpose requirements to ensure that the input data set of each module formed during the operation of the system is less than or equal to the data set required by each module.

[0081] For example, if the data input requirement of the loaded physiological-oriented analysis model is the most recently collected x serial data, that is, [ECG, blood oxygen, EEG, and body temperature after filtering and noise reduction], and the system actually only collects ECG and body temperature, the missing data items are replaced by zero.

[0082] According to the information provided by the data processing planning module, extract the most recent data sequence collected, [original ECG sequence, original body temperature sequence], and record the corresponding position of the current data sequence in time as t.

[0083] During data preprocessing and diversion, according to the planning of the data processing planning module, it is known that the input data required by the subsequent physiological guidance analysis model module is [ECG sequence after filtering and denoising, body temperature sequence after filtering and denoising], and the input data required by the target demand conversion analysis module is [ECG sequence after filtering and denoising, body temperature sequence after filtering and denoising, current target demand].

[0084] After the physiological guidance analysis model module is run, the output data Med t , that is, [current heart rate sequence, current heart rate variability sequence, current body warm-up evaluation index], and the target demand conversion analysis module outputs the same dimension target-oriented data set Targ t , that is, [current target heart rate sequence, current target heart rate variability sequence, current target body warm-up assessment index]. Although the target demand is given within a certain period of time, the intervention process will not be achieved in one go, so the target data here is also adjusted each time as the state of the intervened object changes.

[0085] The purpose-oriented analysis model inputs data, and the physiological-oriented analysis model outputs Med t , [Current heart rate sequence, current heart rate variability sequence, current body warm-up evaluation index], target-oriented dataset Targ output by the target demand conversion analysis model t , [current target heart rate sequence, current target heart rate variability sequence, current target body warm-up evaluation index], and current intervention method, output the first intervention plan data set Sch1 t , that is, [the first intervention plan adjusts the treadmill speed to k meters / second].

[0086] Prediction module. For the historical data extraction module, the input data is Med t , Sch1 t , assuming that the current historical data has saved two pieces of the same type of data [Med t-1 , Sch1 t-1 , Sch2 t-1 , [Med t-2 , Sch1 t-1 , Shc2 t-2 , then the input for the predictive analysis model is [[Med t , Sch1 t , 0], [Med t-1 , Sch1 t-1 , Sch2 t-1 , [Med t-2 , Sch1 t-1 , Shc2 t-2 . Then the input for the predictive analysis model is that if Sch1 t is directly implemented, the predicted change in physiological signal ΔMed t , [the predicted sequence of heart rate changes, the predicted sequence of heart rate variability changes, the predicted change in the body preheat assessment index].

[0087] Intervention decision-making module. If the safety threshold at this time is static, Thres t is a safety value [allowing adjustment of the maximum treadmill speed], then the calculation of Sch2 t is performed:

[0088] ,

[0089] At this time, because Med t , ΔMed t , Targ t is a set of time-series data, it can be merged into a single value, for example, by taking the mean respectively, and then calculating the one-dimensional data Sch2 t , [the second intervention plan adjusts the treadmill speed to k m / s]; it can also be calculated directly in the form of sequence data in turn to form the sequence data Sch2 t , [the second intervention plan adjusts the treadmill speed to k1 m / s, the second intervention plan adjusts the treadmill speed to k2 m / s,..., the second intervention plan adjusts the treadmill speed to kn m / s,]. In addition to performing the intervention according to Sch2 t , the current information of each analysis model and [Med t , Sch1 t , Sch2 tSend it as historical data to the prediction module for data storage. After completing the data analysis and corresponding intervention operations at the t-th time, the system process returns to the initial description and performs the data analysis operation at the (t + 1)-th time.

[0090] The above application scenario is only an extremely simple application scenario to which the technical solution of this application can be applied; the technical solution of this application can also be applied to more complex application scenarios with multi-dimensional interventions. In the technical solution of this application, the purpose-oriented physiological intervention method and system include a data evaluation module, an intervention decision module, a purpose requirement management module, and an intervention module; through step C of the implementation method planned by data evaluation, step D of loading the analysis model, step E of data preprocessing and diversion, step F of physiological orientation analysis, step G of purpose requirement conversion, step H of purpose-oriented analysis, step I of intervention decision-making, and step J of implementing intervention measures, that is, the intervention management module selects the current intervention method and intervention mechanism to be executed according to the second intervention plan dataset Sch2 generated by the intervention decision module t , to select the current intervention method and intervention mechanism to be executed and implement the intervention measures. Through purpose-oriented analysis, the physiological intervention method has a purpose orientation; and there are clear records and feedback in the process, forming a real-time closed loop between the intervention purpose and measures, greatly improving the real-time performance and intervention efficiency of the physiological intervention method.

[0091] The above are only embodiments of the present invention, and do not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of the invention, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present invention.

Claims

1. An objective - oriented physiological intervention method, characterized in that, it includes the following steps: Step C of planning the data evaluation implementation method: Take out the recently collected and classified physiological signal data from the original data pool, and plan the data evaluation implementation method according to the data type information of the current physiological signal data; The planning of the data evaluation implementation method includes the selection of analysis models; The analysis models include physiological - oriented analysis models, objective - requirement conversion models, objective - oriented analysis models, intervention decision - making models, and safety mechanism models; When each analysis model is selected, the input - output data forms of each analysis model are also determined accordingly; Step D of loading the analysis models: The physiological - oriented analysis model module loads the physiological - oriented analysis model, the objective - requirement conversion model module loads the objective - requirement conversion model, and the objective - oriented analysis model module loads the objective - oriented analysis model; In the intervention decision - making module, the intervention decision - making model module loads the intervention decision - making model, and the safety mechanism model module in the intervention decision - making module loads the safety mechanism model; Step E of data pre - processing and diversion: The pre - processing and diversion module outputs corresponding data to the physiological - oriented analysis model module and the objective - requirement conversion model module according to the types and contents of the input feature data required by each analysis model given in the data processing planning module; Step F: The physiological orientation analysis model module performs a physiological orientation analysis on the data transmitted by the preprocessing and diversion module. The analysis data results of the physiological orientation analysis model module express the physiological signal information, characteristics, and / or trends of the target individual in medicine; Denote the data set output by the physiological orientation analysis model module as the physiological orientation data set Med t ; The physiological orientation data set Med t is output to the goal orientation analysis model module; Step G: Use the objective - requirement conversion model module to perform objective - requirement conversion on the data type transmitted by the pre - processing and diversion module; The purpose requirement conversion model module obtains the current purpose requirements through the purpose requirement management module. Based on the purpose requirement conversion model, the purpose requirement conversion model module converts the purpose requirement information into data expressions in the same dimension as the physiological orientation data set Med t and records it as the target orientation data set Targ t ; Step H: The purpose-oriented analysis model module analyzes the physiological-oriented data set Med t and the target-oriented data set Targ t according to the intervention methods provided by the intervention management module in the external intervention module, and outputs the first intervention plan data set denoted as Sch1 t， The first intervention plan data set Sch1 t includes at least one basic intervention method and basic intervention rule information; Step I: The physiological guidance analysis model module outputs the physiological guidance data set Med t to the intervention decision-making module, and the goal-oriented analysis model module outputs the first intervention plan data set Sch1 t to the intervention decision-making module; The purpose requirement conversion model module outputs the target-oriented data set Targ t to the intervention decision-making module; the prediction analysis model module outputs the predicted physiological orientation change amount data set ΔMed t Output to the intervention decision-making module; the intervention decision-making module outputs the second intervention plan data set Sch2 according to the intervention decision-making model and the safety mechanism model t to the intervention module; Step J: The intervention module selects the current intervention method and intervention mechanism according to the intervention strategy generated by the intervention decision-making module, i.e., the second intervention plan dataset Sch2 t , and implements the intervention measures.

2. The objective - oriented physiological intervention method according to claim 1, characterized in that, it further includes step K and step L, In step K, a prediction module is set up, and the prediction module includes a historical prediction data module, a prediction model management module, a prediction analysis model module, and a data matching and extraction module; In step C of the implementation method planning for data evaluation, the analysis model selection also includes the predictive analysis model selection; in step D of the analysis model loading, the predictive analysis model module loads the predictive analysis model; in step F: the physiological orientation data set Med t is also output to the prediction module; in step H: the first intervention plan data set Sch1 t is also output to the prediction module; In step L, the physiological orientation data set Med output by the physiological orientation analysis model module t , and the first intervention plan data set Sch1 output by the purpose orientation analysis model module t , are respectively input into the prediction analysis model module and the data matching and extraction module in the prediction module; The historical prediction data module stores the evaluation data information of each target individual, the data evaluation implementation method, and the intervention implementation plan information after the system starts running. The historical prediction data module also stores the evaluation data information of the target individual obtained at the (m + 1)-th time after analysis and intervention are implemented based on the data information of the target individual obtained at the m-th time, where m is a certain time point or time period earlier than t and does not include t, that is ; Data matching and extraction module, which compares historical prediction data with the physiological orientation data set Med t , the first intervention plan data set Sch1 t and extracts the data generated by the same data analysis model from the historical prediction data module; that is, the data matching and extraction module extracts the data matched as the same category from the historical prediction data module; Historical data sets corresponding to the time interval [t-i, t) can be obtained from historical prediction data, and it is defined that represents the time series values of each historical record from the i-th time before the current time mark t to the current time, and does not include t. Then the historical data set includes the historical physiological orientation data set Med output by the physiological orientation analysis model j and the historical first intervention plan data set Sch1 j , as well as the historical second intervention plan data set Sch2 output by the intervention decision module j ; After extracting the historical data, it is used together with Med t , Sch1 t as input data and input into the prediction analysis model module; A predictive analysis model module that performs predictive analysis based on a set of time series data; the set of time series data includes a physiological orientation data set Med at the current time point or time period t 、a first intervention plan data set Sch1 at the current time point or time period t 、a historical physiological orientation data set Med j 、a historical first intervention plan data set Sch1 j and a historical second intervention plan data set Sch2 j ; Output of the predictive analysis model module, in the first intervention plan data set Sch1 t Under the intervention method given, the changes of each basic intervention rule included, generating the corresponding predicted physiological orientation change amount data set ΔMed t , and output to the intervention decision-making module.

3. The objective - oriented physiological intervention method according to claim 2, characterized in that, Step M: The safety mechanism model module in the intervention decision-making module is responsible for controlling the safety boundaries for the implementation of intervention rules according to different intervention methods in the first intervention plan data set Sch1 t ; The safety mechanism model module outputs the safety boundary data set Thres t according to the predicted physiological orientation change data set ΔMed t , the physiological orientation data set Med t , and the first intervention plan data set Sch1 t ; The safety boundary data set Thres t includes the static safety boundary data set Thres t , and also includes the real-time dynamic safety boundary data set Thres t ; The intervention decision-making model module in the intervention decision-making module comprehensively analyzes the input prediction physiological orientation change amount data set ΔMed t , physiological orientation data set Med t and the first intervention plan data set Sch1 t , and uses the current purpose requirements provided by the purpose requirements manager for comprehensive analysis, and uses the intervention rule safety limit data set Thres t as the safety threshold of the intervention rule to adjust the intervention rule in the first intervention plan data set Sch1 t , and outputs the second intervention plan data set Sch2 t ; For all intervention rules in the first intervention plan data set Sch1 t perform a linear adjustment, and the adjustment method is expressed by the formula as follows: . Non-linearly adjust all intervention rules in the first intervention plan data set Sch1 t ; The second intervention plan dataset Sch2 t On the one hand, it is output to the intervention module to execute the actual intervention. On the other hand, it is output to the prediction module and saved in the historical prediction data module.

4. The objective - oriented physiological intervention method according to claim 1, characterized in that, Before step C, it further includes: Step A: Collect the physiological signal data of the target individual; Step B: Classify and save the collected physiological signal data in the original data pool.

5. An objective - oriented physiological intervention system, characterized in that, used to implement the objective - oriented physiological intervention method according to claim 1, including, a data evaluation module, an intervention decision - making module, an objective - requirement management module, and an intervention module; The objective - requirement management module is used to save and maintain the content of one or more objective requirements expected to be achieved through intervention. The objective - requirement management module selects the current objective - requirement information of the system in a static or dynamic manner, and the objective - requirement management module transmits the objective - requirement information to the data evaluation module; The data evaluation module includes an original data pool, a data processing planning module, a pre - processing and diversion module, a physiological - oriented model management module, a physiological - oriented analysis model module, an objective - oriented model management module, an objective - oriented analysis model module, and an objective - requirement conversion model module; The data evaluation module takes the objective - requirement information as one of the input parameters for data evaluation; The intervention decision - making module takes the objective - requirement information as one of the input parameters for formulating intervention decisions; It further includes a prediction module, which is used to predict the effects generated by the upcoming intervention method and provide reference data for the adjustment of the intervention strategy output by the intervention decision-making module; The intervention decision-making module analyzes by combining the information provided by the data evaluation module and outputs a specific intervention strategy to the intervention module; The intervention module executes corresponding intervention measures according to the intervention strategy output by the intervention decision-making module.

6. The goal-oriented physiological intervention system according to claim 5, wherein, The physiological orientation model management module is used to manage and maintain the physiological orientation analysis model used by the physiological orientation analysis model module; The goal-oriented model management module is used to manage and maintain the goal-oriented analysis model used by the goal-oriented analysis model module; in the physiological orientation analysis model and the goal-oriented analysis model, the model information at least includes the model type, the type of input parameter characteristics, and the model configuration information; The data processing planning module is used to plan the implementation method of data evaluation. The implementation method of data evaluation planning includes the selection of the analysis model used; the selection of the analysis model includes the selection of the physiological orientation analysis model, the goal demand conversion model, and the goal-oriented analysis model in the data evaluation module; The preprocessing and diversion module is used to obtain preprocessing instruction information from the data processing planning module. The preprocessing and diversion module preprocesses the information in the original data pool with the obtained data analysis model information and outputs the corresponding data to the physiological orientation analysis model module and the goal demand conversion model module according to the data analysis model information output by the data processing planning module; The physiological orientation analysis model module loads the selected analysis model through the physiological orientation model management module and conducts physiological orientation analysis on the data transmitted by the preprocessing and diversion module. The analysis data result of the physiological orientation analysis model module expresses the medical physiological signal information, characteristics, and / or trends of the target individual; The goal-oriented model management module is used to save and maintain one or more goal demand conversion model information and goal-oriented analysis model information; The goal demand conversion model information and the goal-oriented analysis model information at least include the model type, the type of input parameter characteristics, and the model configuration information; The data update module, through the goal-oriented model management module, adds, deletes, or modifies the goal demand conversion model and the goal-oriented analysis model maintained by the goal-oriented model management module in the data evaluation module; The goal demand conversion model module loads the selected evaluation model through the goal-oriented model management module and conducts goal demand conversion on the data transmitted by the preprocessing and diversion module; The purpose requirement conversion model module obtains the current purpose requirements through the purpose requirement management module, and then converts the purpose information into data expressions in the same dimension as the physiological orientation data set Med through the purpose requirement conversion model, denoted as the target orientation data set Targ t t .​ 7. The goal-oriented physiological intervention system according to claim 6, wherein, The intervention decision-making module includes an intervention and safety mechanism management module, a safety mechanism model module, and an intervention decision-making model module; When planning the implementation method of data evaluation in the data processing planning module, the selection of the analysis model also includes the selection of the intervention decision-making model and the safety mechanism model in the intervention decision-making module; An intervention and security mechanism management module is used to save and maintain information of one or more intervention decision models and security mechanism models; the intervention and security mechanism management module selects the currently loaded intervention decision model and security mechanism model according to the notification information of the data processing planning module in the data evaluation module; The safety mechanism model module controls the safety boundaries for the implementation of intervention rules according to different upcoming intervention methods, and based on the input predicted physiological orientation change amount data set ΔMed t , physiological orientation data set Med t and the first intervention plan data set Sch1 t , and outputs the intervention rule safety boundary Thres t ; The safety mechanism model includes a static safety mechanism model and a dynamic safety mechanism model; An intervention decision-making model module for comprehensively analyzing the input prediction physiological orientation change amount data set ΔMed t , physiological orientation data set Med t and the first intervention plan data set Sch1 t , and performing a comprehensive analysis with the current purpose requirement provided by the purpose requirement manager, outputting a safety limit data set Thres t , and using the safety limit data set Thres t as the safety threshold of the intervention rule to adjust the intervention rule in the first intervention plan data set Sch1 t , and outputting a second intervention plan data set Sch2 t ; The intervention decision-making model module first converts the purpose requirements obtained from the purpose requirement management module into information of the same dimension as the physiological orientation data set Med t to obtain the target orientation data set Targ t , and then compares it with the physiological orientation data set Med t and the predicted physiological orientation change amount data set ΔMed t , and adjusts according to the rules in the first intervention plan data set Sch1 t to obtain the second intervention plan data set Sch2 t .

8. The purpose-oriented physiological intervention system according to claim 7, wherein, The simultaneous prediction module also saves the implemented intervention strategies and the intermediate data information generated before obtaining the intervention strategies as reference data for predictive analysis; The prediction module includes a historical prediction data module, a data matching and extraction module, a prediction analysis model module, and a prediction model management module; The physiological guidance data set Med output by the data evaluation module t and the first intervention plan data set Sch1 t will, on the one hand, be input into the prediction analysis model module in the prediction module, and on the other hand, be input into the data matching and extraction module in the prediction module together with the historical prediction data; When planning the implementation method of data evaluation in the data processing planning module, the selection of the analysis model also includes notifying the prediction model management module in the prediction module to select the corresponding prediction analysis model module; The historical prediction data module saves the information of each data evaluation, intervention implementation plan, and the evaluation data information of the target individual obtained again after implementing the intervention since the system started running; From the historical prediction data, including the historical data set corresponding to the time interval [t-i, t), it is defined that represents the time series value of each historical record between the i-th time before the current time mark t and the current time, and does not include t, that is, the historical data set includes the historical physiological orientation data set Med output by the physiological orientation analysis model j and the historical first intervention plan data set Sch1 j , as well as the historical second intervention plan data set Sch2 output by the intervention decision module j ; A data matching and extraction module that matches historical prediction data with the physiological guidance data set Med t and the first intervention plan data set Sch1 t for comparison, and extracts the data generated by the same data analysis model from the historical prediction data module; that is, the data matching and extraction module extracts the data matched as "the same type" from the historical prediction data module. A predictive analysis model module performs predictive analysis based on a set of time series data sets; the set of time series data sets includes a historical physiological orientation data set Med output by a physiological orientation analysis model j and a historical first intervention plan data set Sch1 j , and a historical second intervention plan data set Sch2 output by an intervention decision module j , the current physiological orientation data set Med t , the current first intervention plan data set Sch1 t ; Output of the predictive analysis model module, in the current first intervention plan data set Sch1 t Under the given intervention method, the changes of each basic intervention rule included generate the corresponding physiological guidance data set Med t The predictive physiological guidance change amount data set ΔMed t ; The predictive physiological guidance change amount data set ΔMed t , on the one hand, is output to the intervention decision-making module, and on the other hand, is output to the historical prediction data module for storage.

9. The purpose-oriented physiological intervention system according to claim 5, wherein, The intervention module includes an intervention management module, an intervention driving module, and an intervention execution module; The intervention management module is used to manage data of one or more intervention methods used in the system; The intervention management module selects the currently executed intervention method and intervention mechanism according to the intervention strategy generated by the intervention decision module; The intervention management module updates the corresponding intervention method data from the external data update module according to the actual configuration of the intervention devices in the system.

10. The purpose-oriented physiological intervention system according to claim 6, wherein, It further includes: a data acquisition module and a data update module; The data acquisition module is used to acquire the physiological signal data of the target individual and transmit the acquired physiological signal data to the data evaluation module; the data evaluation module is used to classify and save the physiological signal data and perform evaluation operations; The data update module is used to interface with an external expert system. On the one hand, the data update module is used to obtain the required analysis model information from the external expert system; on the other hand, the data update module is used to transmit various historical data saved in the purpose-oriented physiological intervention system to the outside; The data update module is used to connect with an external expert system to obtain various analysis models required by the purpose-oriented physiological intervention system from the external expert system; the data update module is also used to transmit various historical data saved in the purpose-oriented physiological intervention system to the outside; The data update module is connected to the data processing planning module, and the data update module is also respectively connected to the physiological orientation model management module, the purpose orientation model management module, the intervention and security mechanism management module, and the prediction model management module; The data update module adds, deletes, or modifies one or more physiological orientation analysis models maintained by the physiological orientation model management module to the data evaluation module through the data processing planning module or the physiological orientation model management module; The data update module, through the data processing planning module or the goal-oriented model management module, adds, deletes, or modifies one or more goal-oriented models maintained by the goal-oriented model management module to the data evaluation module; The data update module, through the data processing planning module or the intervention and security mechanism management module, adds, deletes, or modifies one or more security mechanism models and intervention decision models maintained by this module to the intervention decision module; The data update module, through the data processing planning module or the prediction model management module, adds, deletes, or modifies one or more predictive analysis models maintained by the prediction model management module to the data evaluation module.

Images