Mental Health Index Assessment Method and System

Abstract

The present invention provides a method and system for evaluating a mental health index. The method includes constructing an evaluation model for a mental health risk index; collecting original pulse wave data of a user in real time based on the PPG method; processing the original pulse wave data to obtain physiological data of the user; inputting the physiological data of the user into the evaluation model for the mental health risk index to obtain the mental health risk index of the user; and converting the mental health risk index of the user based on a preset conversion rule for the mental health risk index to obtain the mental health index of the user.

Description

Technical Field

[0001] The present invention relates to the technical field of mental health, and particularly to a method and system for evaluating mental health index. Background Art

[0002] Mental health refers to a continuous and actively developing mental state, in which the subject can make good adaptations and give full play to its physical and mental potential. How to maintain people's mental health and actively intervene and guide those with mental problems is of great importance to the entire family, society, and company.

[0003] Currently, most existing mental health assessment methods use questionnaires or sand trays to evaluate the mental health of users. For example, the patent application with the publication number CN110279425A: A Psychological Assessment Method and System Based on Intelligent Analysis evaluates the mental health of users in the form of a sand tray. However, this method requires users to actively cooperate in psychological assessments. Most users basically do not think they have mental health risks, so they rarely actively cooperate in psychological assessments. Once users notice that they have mental health risks, the risks may have developed from the early stage to the middle or late stage. Therefore, there is an urgent need for a method and system for evaluating mental health index. The system corresponding to this method is set in an intelligent wearable device to continuously monitor the user's physiological data and evaluate the user's mental health index based on this, solving the problem in the prior art that users cannot understand their own mental health assessment index in real time. Summary of the Invention

[0004] In view of the deficiencies of the prior art, the present invention provides a method and system for evaluating mental health index. The system corresponding to this method is set in an intelligent wearable device to continuously monitor the user's physiological data and evaluate the user's mental health index based on this, solving the problem in the prior art that users cannot understand their own mental health assessment index in real time.

[0005] A method for evaluating mental health index includes:

[0006] Constructing a mental health risk index evaluation model;

[0007] Based on the PPG method, continuously collecting the user's original pulse wave data in real time;

[0008] Processing the original pulse wave data to obtain the user's physiological data;

[0009] Inputting the user's physiological data into the mental health risk index evaluation model to obtain the user's mental health risk index;

[0010] Based on a preset conversion rule for the mental health risk index, the mental health risk index of the user is converted to obtain the user's mental health index.

[0011] As an embodiment of the present invention, constructing a mental health risk index assessment model includes:

[0012] Construct an initial prediction model;

[0013] Based on a large-scale medical record database collected and configured in advance, a number of problem physiological data are obtained; among them, the problem physiological data includes physiological state data and mental state data, the physiological state data includes the problem heart rate, problem blood pressure, and problem blood oxygen of the problem user every day, and the mental state data includes the problem heart rate variability parameter;

[0014] Based on the big data AI algorithm, physiological data curves and mental data curves of the state data and time changes are constructed respectively according to each physiological state data and mental state data in the same medical record, and user physiological parameter curves and user mental parameter curves of a number of the same medical records are obtained;

[0015] Curve fitting is performed on a number of user physiological parameter curves and a number of user mental parameter curves respectively to obtain a physiological data reference curve and a mental data reference curve after curve fitting;

[0016] Based on the preset risk values corresponding to each medical record, numerical marking is performed on the physiological data reference curve and the mental data reference curve to obtain a marked physiological data reference curve and a marked mental data reference curve;

[0017] The training physiological data curve and training mental data curve of the state data and time changes constructed by any problem physiological data in the large-scale medical record database collected and configured in advance, the marked physiological data reference curve and the marked mental data reference curve are input into the initial prediction model, and the initial prediction model is trained to obtain a mental health risk index assessment model.

[0018] As an embodiment of the present invention, the user's physiological data includes heart rate, blood pressure, blood oxygen, and heart rate variability parameter.

[0019] As an embodiment of the present invention, inputting the user's physiological data into the mental health risk index assessment model to obtain the user's mental health risk index includes:

[0020] The user's physiological data is divided to obtain the user's physiological state data and the user's mental state data; among them, the user's physiological state data includes heart rate, blood pressure, and blood oxygen, and the user's mental state data includes heart rate variability parameter;

[0021] Construct a curve graph of the user's physiological data showing the change of the state data over time according to the user's physiological state data;

[0022] Construct a curve graph of the user's psychological data showing the change of the state data over time according to the user's psychological state data;

[0023] Input the curve graph of the user's physiological data into the mental health risk index assessment model. The mental health risk index assessment model outputs the risk value corresponding to the marked physiological data reference curve graph in the mental health risk index assessment model whose similarity to the curve graph of the user's physiological data is greater than the preset first similarity threshold, as the first risk value;

[0024] Input the curve graph of the user's psychological data into the mental health risk index assessment model. The mental health risk index assessment model outputs the risk value corresponding to the marked psychological data reference curve graph in the mental health risk index assessment model whose similarity to the curve graph of the user's psychological data is greater than the preset second similarity threshold, as the second risk value;

[0025] Based on the preset weight value, perform a weighted sum calculation on the first risk value and the second risk value to obtain the user's mental health risk index.

[0026] As an embodiment of the present invention, a mental health index assessment method further includes: after a preset time, based on the user's physiological data collected in a preset time period, construct a personalized mental health risk index assessment model to replace the original mental health risk index assessment model;

[0027] The construction process of the personalized mental health risk index assessment model includes:

[0028] After receiving the construction instruction, construct an initial prediction model;

[0029] Receive the user's physiological data collected in a preset time period and construct personalized user physiological data; wherein, the personalized user physiological data includes a number of user physiological data and their corresponding timestamps;

[0030] According to the personalized user physiological data, predict the current user's heart age and the abnormal emotions that occur frequently every day; wherein, predict the current user's heart age through the personalized heart rate, personalized blood pressure, and personalized blood oxygen in the personalized user physiological data, and predict the abnormal emotions that occur frequently every day for the current user through the personalized heart rate variability parameters in the personalized user physiological data;

[0031] According to the current user's heart age and the abnormal emotions that occur frequently every day, determine the problem physiological data of the medical records that match the current user's heart age and the abnormal emotions that occur frequently every day from the pre-collected and configured large-scale medical record database as the first training data;

[0032] Obtain the problem physiological data of the medical records corresponding to the user's mental health risk index output by the mental health risk index evaluation model within a preset time period as the second training data;

[0033] Based on the construction method of the mental health risk index evaluation model, obtain the second user physiological parameter curve graph and the second user psychological parameter curve graph corresponding to a number of identical medical records according to the first training data and the second training data;

[0034] Obtain the second physiological data reference curve graph and the second psychological data reference curve graph after curve fitting according to the second user physiological parameter curve graph and the second user psychological parameter curve graph;

[0035] Based on the preset risk value corresponding to each medical record, perform numerical marking on the second physiological data reference curve graph and the second psychological data reference curve graph to obtain the second marked physiological data reference curve graph and the second marked psychological data reference curve graph;

[0036] Input the state data constructed by any problem physiological data in the first training data and the second training data, the second training physiological data curve graph and the second training psychological data curve graph, the second marked physiological data reference curve graph and the second marked psychological data reference curve graph into the initial prediction model, and train the initial prediction model to obtain a personalized mental health risk index evaluation model.

[0037] A mental health index evaluation system includes:

[0038] A construction module for constructing a mental health risk index evaluation model;

[0039] An acquisition module for real-time acquiring the user's original pulse wave data based on the PPG method;

[0040] A processing module for processing the original pulse wave data to obtain the user's physiological data;

[0041] The processing module is further configured to input the user's physiological data into the mental health risk index evaluation model to obtain the user's mental health risk index;

[0042] A calculation module for converting the user's mental health risk index based on a preset mental health risk index conversion rule to obtain the user's mental health index.

[0043] As an embodiment of the present invention, the construction module performs the following operations:

[0044] Construct an initial prediction model;

[0045] Based on a large-scale medical record database collected and configured in advance, a number of problematic physiological data are obtained; among them, the problematic physiological data includes physiological state data and psychological state data, the physiological state data includes the problematic heart rate, problematic blood pressure, and problematic blood oxygen of the problematic user every day, and the psychological state data includes problematic heart rate variability parameters;

[0046] Based on big data AI algorithms, physiological data curves and psychological data curves showing the changes of state data over time are respectively constructed according to each piece of physiological state data and psychological state data in the same medical record, and a number of user physiological parameter curves and user psychological parameter curves of the same medical record are obtained;

[0047] Curve fitting is respectively performed on a number of user physiological parameter curves and a number of user psychological parameter curves to obtain a physiological data reference curve and a psychological data reference curve after curve fitting;

[0048] Based on preset dangerous values corresponding to each medical record, numerical markings are made on the physiological data reference curve and the psychological data reference curve to obtain a marked physiological data reference curve and a marked psychological data reference curve;

[0049] The training physiological data curve and training psychological data curve showing the changes of state data constructed from any problematic physiological data in the large-scale medical record database collected and configured in advance, the marked physiological data reference curve, and the marked psychological data reference curve are input into the initial prediction model, and the initial prediction model is trained to obtain a mental health risk index evaluation model.

[0050] As an embodiment of the present invention, the user physiological data includes heart rate, blood pressure, blood oxygen, and heart rate variability parameters.

[0051] As an embodiment of the present invention, the processing module is further configured to perform operations including the following:

[0052] The user physiological data is divided to obtain user physiological state data and user psychological state data; among them, the user physiological state data includes heart rate, blood pressure, and blood oxygen, and the user psychological state data includes heart rate variability parameters;

[0053] A user physiological data curve showing the changes of state data over time is constructed according to the user physiological state data;

[0054] A user psychological data curve showing the changes of state data over time is constructed according to the user psychological state data;

[0055] Input the user's physiological data curve into the mental health risk index assessment model. The mental health risk index assessment model outputs the risk value corresponding to the marked physiological data reference curve in the mental health risk index assessment model whose similarity to the user's physiological data curve is greater than a preset first similarity threshold, and use it as the first risk value;

[0056] Input the user's psychological data curve into the mental health risk index assessment model. The mental health risk index assessment model outputs the risk value corresponding to the marked psychological data reference curve in the mental health risk index assessment model whose similarity to the user's psychological data curve is greater than a preset second similarity threshold, and use it as the second risk value;

[0057] Based on a preset weight value, perform a weighted sum calculation on the first risk value and the second risk value to obtain the user's mental health risk index.

[0058] As an embodiment of the present invention, a mental health index assessment system further includes: a personalized construction module, which is used to construct a personalized mental health risk index assessment model based on the user's physiological data collected during a preset time period after a preset time to replace the original mental health risk index assessment model;

[0059] The construction process of the personalized mental health risk index assessment model includes:

[0060] After receiving the construction instruction, construct an initial prediction model;

[0061] Receive the user's physiological data collected during the preset time period and construct personalized user physiological data; among them, the personalized user physiological data includes a number of user physiological data and their corresponding timestamps;

[0062] According to the personalized user physiological data, predict the current user's heart age and frequently occurring abnormal emotions per day; among them, predict the current user's heart age through the personalized heart rate, personalized blood pressure, and personalized blood oxygen in the personalized user physiological data, and predict the frequently occurring abnormal emotions of the current user per day through the personalized heart rate variability parameter in the personalized user physiological data;

[0063] According to the current user's heart age and frequently occurring abnormal emotions per day, determine the problem physiological data of the medical record that matches the current user's heart age and frequently occurring abnormal emotions from the pre-collected and configured large-scale medical record database as the first training data;

[0064] Obtain the problem physiological data of the medical record corresponding to the user's mental health risk index output by the mental health risk index assessment model during the preset time period as the second training data;

[0065] A construction method of a mental health risk index evaluation model, according to the first training data and the second training data, obtains the second user physiological parameter curve graph and the second user psychological parameter curve graph corresponding to a number of identical medical records;

[0066] According to the second user physiological parameter curve graph and the second user psychological parameter curve, obtain the second physiological data reference curve graph and the second psychological data reference curve graph after curve fitting;

[0067] Based on the preset risk values corresponding to each medical record, numerically mark the second physiological data reference curve graph and the second psychological data reference curve graph to obtain the second marked physiological data reference curve graph and the second marked psychological data reference curve graph;

[0068] Input the state data constructed by any problem physiological data in the first training data and the second training data, the second training physiological data curve graph and the second training psychological data curve graph, the second marked physiological data reference curve graph and the second marked psychological data reference curve graph into the initial prediction model, and train the initial prediction model to obtain a personalized mental health risk index evaluation model.

[0069] The beneficial effects of the present invention are as follows: The present invention is set in an intelligent wearable device, combines multiple current physiological parameters of the user to evaluate the mental health state of the user, is used to monitor the physiological data of the user in real time and evaluate the mental health index of the user based on this, and solves the problem that users in the prior art cannot understand their own mental health evaluation index in real time.

[0070] Other features and advantages of the present invention will be described in the following description, and in part, will be obvious from the description, or will be understood by implementing the present invention. The objectives and other advantages of the present invention can be achieved and obtained by the structures specifically pointed out in the written description and the drawings.

[0071] The technical solution of the present invention will be further described in detail below through the drawings and embodiments. Description of the Drawings

[0072] The drawings are used to provide a further understanding of the present invention, and constitute a part of the description. They are used together with the embodiments of the present invention to explain the present invention, and do not constitute a limitation to the present invention. In the drawings:

[0073] Figure 1 It is a method flow chart of a mental health index evaluation method and system in an embodiment of the present invention;

[0074] Figure 2 It is a system schematic diagram in the initial state of a mental health index evaluation method and system in an embodiment of the present invention;

[0075] Figure 3 This is a schematic diagram of the system after a preset time in a mental health index evaluation method and system according to an embodiment of the present invention. Specific implementation manners

[0076] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only for the purpose of illustrating and explaining the present invention, and are not intended to limit the present invention.

[0077] Please refer to Figure 1 , an embodiment of the present invention provides a mental health index evaluation method, including:

[0078] Construct a mental health risk index evaluation model;

[0079] Based on the PPG method, collect the user's original pulse wave data in real time;

[0080] Perform data processing on the original pulse wave data to obtain the user's physiological data;

[0081] Input the user's physiological data into the mental health risk index evaluation model to obtain the user's mental health risk index;

[0082] Based on a preset mental health risk index conversion rule, convert the user's mental health risk index to obtain the user's mental health index;

[0083] The working principle of the above technical solution is as follows: This invention is mainly applied to intelligent wearable devices and is used to double-evaluate the mental health index of the current user based on the user's physiological state and mental state. The evaluation method is as follows: First, a mental health risk index evaluation model is constructed. This model is preferably constructed during the production of the corresponding device. Then, based on the corresponding device on the intelligent wearable device, in the PPG (photoplethysmographic) mode, the user's original pulse wave data is collected in real time. After obtaining the original pulse wave data, the original pulse wave data is processed to obtain the user's physiological data. The user's physiological data preferably includes heart rate, blood pressure, blood oxygen, and heart rate variability parameters. Among them, the method of extracting the user's physiological data based on the original pulse wave data belongs to the methods well-known to those skilled in the art and will not be elaborated here one by one. After obtaining the user's physiological data, this user's physiological data is used as input and transmitted to the previously constructed mental health risk index evaluation model to obtain the user's mental health risk index regarding this user. It should be noted that the user's physiological data input here is usually all the user's physiological data within one day of this user. Thus, the obtained user's mental health risk index is used to predict the mental health risk state of the user on the same day. Finally, based on the preset mental health risk index conversion rule, the user's mental health risk index is converted to obtain the user's mental health index. Among them, the preset mental health risk index conversion rule is preferably: HEALTH = ALL - DANGEROUS; where HEALTH is the user's mental health index, ALL is the preset total balance value, and DANGEROUS is the user's mental health risk index;

[0084] The beneficial effects of the above technical solution are as follows: This invention is set in an intelligent wearable device, combines multiple current physiological parameters of the user to evaluate the user's mental health state, improves the objectivity of the evaluation by detecting physiological parameters, and evaluates the user's mental health problems in real time through the wearability of the intelligent wearable device, improving the evaluation efficiency.

[0085] In one embodiment, constructing the mental health risk index evaluation model includes:

[0086] Construct an initial prediction model;

[0087] Based on a pre-collected and configured large-scale medical record database, a number of problem physiological data are obtained; among them, the problem physiological data includes physiological state data and mental state data. The physiological state data includes the problem heart rate, problem blood pressure, and problem blood oxygen of the problem user every day, and the mental state data includes the problem heart rate variability parameter;

[0088] Based on the big data AI algorithm, physiological data curves and psychological data curves showing the changes of state data over time are constructed respectively according to each physiological state data and psychological state data in the same medical records, and user physiological parameter curves and user psychological parameter curves of a number of the same medical records are obtained;

[0089] Curve fitting is performed respectively on a number of user physiological parameter curves and a number of user psychological parameter curves to obtain a physiological data benchmark curve and a psychological data benchmark curve after curve fitting;

[0090] Based on the preset risk values corresponding to each medical record, numerical markings are made on the physiological data benchmark curve and the psychological data benchmark curve to obtain a marked physiological data benchmark curve and a marked psychological data benchmark curve;

[0091] The training physiological data curve and training psychological data curve showing the changes of state data over time constructed from any problem physiological data, the marked physiological data benchmark curve and the marked psychological data benchmark curve in a large-scale medical record database collected and configured in advance are input into the initial prediction model, and the initial prediction model is trained to obtain a mental health risk index assessment model;

[0092] The working principle of the above technical solution is as follows: It is preferably to complete the construction of the mental health risk index assessment model before the corresponding device leaves the factory. Among them, the construction process is as follows: First, construct an initial prediction model, which preferably adopts an initialized classification prediction model, and then based on a large-scale medical record database collected and configured in advance, obtain a number of problem physiological data; among them, the problem physiological data includes physiological state data and mental state data. The physiological state data includes the problem heart rate, problem blood pressure, and problem blood oxygen of the problem user every day, and the mental state data includes the problem heart rate variability parameter; then based on the big data AI algorithm, respectively construct a physiological data curve graph and a mental data curve graph of the state data and time change according to each physiological state data and mental state data in the same medical record, and obtain a number of user physiological parameter curve graphs and user mental parameter curve graphs of the same medical record; among them, the unit of the abscissa of the curve graph is preferably minutes, and the value of the ordinate is preferably the average value of the corresponding physiological / mental state data within 1 minute; then respectively perform curve fitting on a number of user physiological parameter curve graphs and a number of user mental parameter curve graphs to obtain a physiological data reference curve graph and a mental data reference curve graph after curve fitting; The method of curve fitting can quickly analyze the relationship between a certain physiological / mental state data and time change, and accurately formulate a reference graph; Based on the preset risk value corresponding to each medical record, numerically mark the physiological data reference curve graph and the mental data reference curve graph to obtain a marked physiological data reference curve graph and a marked mental data reference curve graph. This step is used to mark the curve graph according to the risk coefficient of the medical record corresponding to the curve graph; Input the training physiological data curve graph and training mental data curve graph of the state data and time change constructed by any problem physiological data, the marked physiological data reference curve graph and the marked mental data reference curve graph in the large-scale medical record database collected and configured in advance into the initial prediction model, and train the initial prediction model to obtain a mental health risk index assessment model; Preferably, the input of the model is the training physiological data curve graph and training mental data curve graph of the state data and time change constructed by any problem physiological data, the training condition is that the similarity between the training physiological data curve graph and training mental data curve graph and the corresponding marked physiological data reference curve graph and marked mental data reference curve graph reaches a preset threshold, and the output is the risk value marked by the corresponding marked physiological data reference curve graph and marked mental data reference curve graph after reaching the preset threshold. Train until the accuracy rate reaches the preset accuracy rate to complete the training;

[0093] The beneficial effect of the above technical solution is that by using a large-scale medical record database collected and configured in advance and the big data AI algorithm to train the mental health risk index assessment model, it is beneficial to improve the objectivity of the output result of the mental health risk index assessment model.

[0094] In one embodiment, the user's physiological data includes heart rate, blood pressure, blood oxygen, and heart rate variability parameters;

[0095] The beneficial effects of the above technical solution are as follows: Based on the heart rate, blood pressure, blood oxygen, and heart rate variability parameters, the mental health index of the current user is evaluated from the dual perspectives of physiological state and mental state, which is beneficial to improving the accuracy of the evaluation.

[0096] In one embodiment, the user's physiological data is input into a mental health risk index evaluation model to obtain the user's mental health risk index, including:

[0097] The user's physiological data is divided to obtain the user's physiological state data and the user's mental state data; among them, the user's physiological state data includes heart rate, blood pressure, and blood oxygen, and the user's mental state data includes heart rate variability parameters;

[0098] A curve graph of the user's physiological data with the change of state data and time is constructed according to the user's physiological state data;

[0099] A curve graph of the user's mental data with the change of state data and time is constructed according to the user's mental state data;

[0100] The user's physiological data curve graph is input into the mental health risk index evaluation model, and the mental health risk index evaluation model outputs the risk value corresponding to the marked physiological data reference curve graph in the mental health risk index evaluation model with a similarity greater than a preset first similarity threshold as the first risk value;

[0101] The user's mental data curve graph is input into the mental health risk index evaluation model, and the mental health risk index evaluation model outputs the risk value corresponding to the marked mental data reference curve graph in the mental health risk index evaluation model with a similarity greater than a preset second similarity threshold as the second risk value;

[0102] Based on a preset weight value, the first risk value and the second risk value are weighted and summed to obtain the user's mental health risk index;

[0103] The working principle of the above technical solution is as follows: After the system corresponding to this method is used in conjunction with a smart wearable device, the user's physiological data is input into the mental health risk index assessment model to obtain the user's mental health risk index, including: dividing and processing the user's physiological data to obtain the user's physiological state data and the user's mental state data; among them, the user's physiological state data includes heart rate, blood pressure, and blood oxygen, and the user's mental state data includes heart rate variability parameters; furthermore, the user's physiological data includes all the user's physiological data collected within a day of the user; then, a physiological data curve graph of the state data and the time-varying user physiological data is constructed based on the user's physiological state data; and a mental data curve graph of the state data and the time-varying user mental data is constructed based on the user's mental state data; among them, the construction methods of the user's physiological data curve graph and the user's mental data curve graph are the same as those of the training physiological data curve graph and the training mental data curve graph in the mental health risk index assessment model; input the user's physiological data curve graph into the mental health risk index assessment model, and the mental health risk index assessment model outputs the risk value corresponding to the marked physiological data reference curve graph in the mental health risk index assessment model whose similarity to the user's physiological data curve graph is greater than the preset first similarity threshold, as the first risk value; input the user's mental data curve graph into the mental health risk index assessment model, and the mental health risk index assessment model outputs the risk value corresponding to the marked mental data reference curve graph in the mental health risk index assessment model whose similarity to the user's mental data curve graph is greater than the preset second similarity threshold, as the second risk value; based on the preset weight values, perform a weighted summation calculation on the first risk value and the second risk value to obtain the user's mental health risk index; the calculation method is preferably: DANGEROUS = α * dangerous 1 + β * dangerous 2 , where DANGEROUS is the user's mental health risk index, dangerous 1 is the first risk value, dangerous 2 is the second risk value, α and β are preset weight values, and α + β = 1, and β is much greater than α;

[0104] The beneficial effects of the above technical solution are as follows: By setting up a mental health risk index assessment model, the corresponding module set on the smart wearable device collects the user's physiological data every day for corresponding processing to obtain the user's mental health risk index, effectively improving the efficiency of the user's mental health index assessment.

[0105] In one embodiment, a mental health index assessment method further includes: after a preset time, based on the user's physiological data collected within a preset time period, constructing a personalized mental health risk index assessment model to replace the original mental health risk index assessment model;

[0106] The construction process of the personalized mental health risk index assessment model includes:

[0107] After receiving the construction instruction, construct an initial prediction model;

[0108] Receive the user's physiological data collected during a preset time period and construct personalized user physiological data; among them, the personalized user physiological data includes a number of user physiological data and their corresponding timestamps;

[0109] Based on the personalized user physiological data, predict the current user's heart age and the abnormal emotions that occur frequently every day; among them, predict the current user's heart age through the personalized heart rate, personalized blood pressure, and personalized blood oxygen in the personalized user physiological data, and predict the abnormal emotions that occur frequently every day for the current user through the personalized heart rate variability parameters in the personalized user physiological data;

[0110] Based on the current user's heart age and the abnormal emotions that occur frequently every day, determine the problem physiological data of the medical records that match the current user's heart age and the abnormal emotions that occur frequently every day from the large-scale medical record database collected and configured in advance as the first training data;

[0111] Obtain the problem physiological data of the medical records corresponding to the user's mental health risk index output by the mental health risk index assessment model during a preset time period as the second training data;

[0112] Based on the construction method of the mental health risk index assessment model, obtain the second user physiological parameter curve graph and the second user psychological parameter curve graph corresponding to a number of the same medical records according to the first training data and the second training data;

[0113] Obtain the second physiological data reference curve graph and the second psychological data reference curve graph after curve fitting according to the second user physiological parameter curve graph and the second user psychological parameter curve;

[0114] Based on the preset risk value corresponding to each medical record, perform numerical marking on the second physiological data reference curve graph and the second psychological data reference curve graph to obtain the second marked physiological data reference curve graph and the second marked psychological data reference curve graph;

[0115] Input the state data constructed from any problem physiological data in the first training data and the second training data, the second training physiological data curve graph and the second training psychological data curve graph, the second marked physiological data reference curve graph and the second marked psychological data reference curve graph into the initial prediction model, and train the initial prediction model to obtain the personalized mental health risk index assessment model;

[0116] The working principle of the above technical solution is as follows: The mental health risk index assessment model is preferably constructed during the factory period of the corresponding device. In order to better adapt to each user, reduce the assessment time during the assessment, and improve the assessment efficiency, after a preset time, based on the user's physiological data collected during a preset time period, a personalized mental health risk index assessment model is constructed to replace the original mental health risk index assessment model to achieve personalized assessment. Among them, the construction process of the personalized mental health risk index assessment model includes the following processes. First, when the activation time of the product corresponding to this method reaches the preset time, after the construction module corresponding to the personalized mental health risk index assessment model receives the construction instruction sent by other modules, it starts to construct an initial prediction model. Similarly, the initial prediction model preferably adopts an initialized prediction classification model. At the same time, it starts to receive the user's physiological data collected during the preset time period to construct personalized user physiological data. Among them, the personalized user physiological data includes several user physiological data and their corresponding timestamps. The timestamps are used to sort the same type of data in the personalized user physiological data to facilitate improving the subsequent call time. Then, based on the personalized user physiological data, predict the current user's heart age and the abnormal emotions that occur frequently every day, including: predicting the current user's heart age through the personalized heart rate, personalized blood pressure, and personalized blood oxygen in the personalized user physiological data, and predicting the abnormal emotions that occur frequently every day for the current user through the personalized heart rate variability parameters in the personalized user physiological data. The prediction method is preferably to bring the personalized heart rate variability parameters into a preset formula set to obtain the corresponding emotion value, and obtain the corresponding emotion according to the emotion value. Among them, the abnormal emotions include abnormal emotions such as anger, palpitation, and depression. The setting method of the formula set is preferably: First, perform numerical exponential processing on the heart rate variability parameters for training. Then, combine various actual medical records in the pre-configured large-scale medical record database and the emotional stability to establish a formula set for evaluating each of the emotion indicators for each of the heart rate variability parameters. After obtaining the current user's heart age and the abnormal emotions that occur frequently every day, according to the current user's heart age and the abnormal emotions that occur frequently every day, determine the problem physiological data of the medical record that matches the current user's heart age and the abnormal emotions that occur frequently every day from the pre-collected and configured large-scale medical record database as the first training data. Then, obtain the problem physiological data of the medical record corresponding to the user's mental health risk index output by the mental health risk index assessment model during the preset time period as the second training data. Based on the construction method of the mental health risk index assessment model, according to the first training data and the second training data, obtain the second user physiological parameter curve graph and the second user psychological parameter curve graph corresponding to several identical medical records. And according to the second user physiological parameter curve graph and the second user psychological parameter curve, obtain the second physiological data reference curve graph and the second psychological data reference curve graph after curve fitting.Then, based on the preset risk values corresponding to each medical record, numerical markings are made on the second physiological data baseline curve graph and the second psychological data baseline curve graph to obtain the second marked physiological data baseline curve graph and the second marked psychological data baseline curve graph. Then, among the first training data and the second training data, the state data constructed from any problem physiological data, the second training physiological data curve graph and the second training psychological data curve graph that change with time, the second marked physiological data baseline curve graph and the second marked psychological data baseline curve graph are input into the initial prediction model. Finally, the initial prediction model is trained to obtain a personalized mental health risk index evaluation model;

[0117] The beneficial effects of the above technical solution are as follows: By setting up a personalized mental health risk index evaluation model, the number of classification objects in the model is reduced compared to the mental health risk index evaluation model, and it better meets the current user's mental health assessment needs, which is beneficial to reducing the assessment time and improving the assessment efficiency during the assessment.

[0118] A mental health index evaluation system, comprising:

[0119] A construction module for constructing a mental health risk index evaluation model;

[0120] An acquisition module for real-time acquiring the user's original pulse wave data based on the PPG method;

[0121] A processing module for processing the original pulse wave data to obtain the user's physiological data;

[0122] The processing module is also used to input the user's physiological data into the mental health risk index evaluation model to obtain the user's mental health risk index;

[0123] A calculation module for converting the user's mental health risk index based on a preset mental health risk index conversion rule to obtain the user's mental health index.

[0124] In one embodiment, the construction module performs the following operations:

[0125] Construct an initial prediction model;

[0126] Based on a large-scale pre-collected and configured medical record database, a number of problem physiological data are obtained; among them, the problem physiological data include physiological state data and psychological state data, the physiological state data include the problem heart rate, problem blood pressure, and problem blood oxygen of the problem user every day, and the psychological state data include problem heart rate variability parameters;

[0127] Based on big data AI algorithms, physiological data curves and psychological data curves showing the variation of state data with time are constructed respectively according to each physiological state data and psychological state data in the same medical records, and user physiological parameter curves and user psychological parameter curves of a number of the same medical records are obtained;

[0128] Curve fitting is performed respectively on a number of user physiological parameter curves and a number of user psychological parameter curves to obtain a physiological data reference curve and a psychological data reference curve after curve fitting;

[0129] Based on preset risk values corresponding to each medical record, numerical markings are made on the physiological data reference curve and the psychological data reference curve to obtain a marked physiological data reference curve and a marked psychological data reference curve;

[0130] The state data and time-varying training physiological data curves and training psychological data curves constructed from any problem physiological data in a pre-collected and configured large-scale medical record database, the marked physiological data reference curve and the marked psychological data reference curve are input into an initial prediction model, and the initial prediction model is trained to obtain a mental health risk index assessment model.

[0131] In one embodiment, the user physiological data includes heart rate, blood pressure, blood oxygen and heart rate variability parameters.

[0132] In one embodiment, the processing module is further configured to perform operations including the following:

[0133] The user physiological data is divided to obtain user physiological state data and user psychological state data; among them, the user physiological state data includes heart rate, blood pressure and blood oxygen, and the user psychological state data includes heart rate variability parameters;

[0134] A user physiological data curve showing the variation of state data with time is constructed according to the user physiological state data;

[0135] A user psychological data curve showing the variation of state data with time is constructed according to the user psychological state data;

[0136] The user physiological data curve is input into the mental health risk index assessment model, and the mental health risk index assessment model outputs the risk value corresponding to the marked physiological data reference curve in the mental health risk index assessment model whose similarity to the user physiological data curve is greater than a preset first similarity threshold as the first risk value;

[0137] Input the user's psychological data curve into the mental health risk index assessment model, and the mental health risk index assessment model outputs the risk value corresponding to the marked psychological data reference curve in the mental health risk index assessment model where the similarity with the user's psychological data curve is greater than the preset second similarity threshold, as the second risk value;

[0138] Based on the preset weight value, perform a weighted sum calculation on the first risk value and the second risk value to obtain the user's mental health risk index.

[0139] In one embodiment, a mental health index assessment system further includes a personalized construction module, which is used to, after a preset time, construct a personalized mental health risk index assessment model based on the user's physiological data collected during a preset time period to replace the original mental health risk index assessment model;

[0140] The construction process of the personalized mental health risk index assessment model includes:

[0141] After receiving the construction instruction, construct an initial prediction model;

[0142] Receive the user's physiological data collected during the preset time period and construct personalized user physiological data; wherein, the personalized user physiological data includes a number of user physiological data and their corresponding timestamps;

[0143] According to the personalized user physiological data, predict the current user's heart age and the abnormal emotions that occur frequently every day; wherein, predict the current user's heart age through the personalized heart rate, personalized blood pressure, and personalized blood oxygen in the personalized user physiological data, and predict the abnormal emotions that occur frequently every day for the current user through the personalized heart rate variability parameter in the personalized user physiological data;

[0144] According to the current user's heart age and the abnormal emotions that occur frequently every day, determine the problem physiological data of the medical records that match the current user's heart age and the abnormal emotions that occur frequently every day from the large-scale medical record database collected and configured in advance as the first training data;

[0145] Obtain the problem physiological data of the medical records corresponding to the user's mental health risk index output by the mental health risk index assessment model during the preset time period as the second training data;

[0146] Based on the construction method of the mental health risk index assessment model, obtain the second user physiological parameter curve and the second user psychological parameter curve corresponding to a number of the same medical records according to the first training data and the second training data;

[0147] Obtain the second physiological data reference curve and the second psychological data reference curve after curve fitting according to the second user physiological parameter curve and the second user psychological parameter curve;

[0148] Based on the preset risk values corresponding to each medical record, numerical markings are made on the second physiological data reference curve graph and the second psychological data reference curve graph to obtain a second marked physiological data reference curve graph and a second marked psychological data reference curve graph;

[0149] Among the first training data and the second training data, the state data constructed from any problem physiological data, the second training physiological data curve graph and the second training psychological data curve graph that change with time, the second marked physiological data reference curve graph and the second marked psychological data reference curve graph are input into the initial prediction model, and the initial prediction model is trained to obtain a personalized mental health risk index assessment model.

[0150] For the working principles and beneficial effects of different internal functional modules in this mental health index assessment system, reference can be made to the corresponding working principles and beneficial effects mentioned in the above-mentioned mental health index assessment method, and no repeated elaboration will be made here.

[0151] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and its equivalent technologies, the present invention also intends to include these changes and modifications.

Claims

1. A method for evaluating mental health index, characterized in that, it includes: Constructing a mental health risk index evaluation model; Based on the PPG method, real-time collecting the user's original pulse wave data; Processing the original pulse wave data to obtain the user's physiological data; Inputting the user's physiological data into the mental health risk index evaluation model to obtain the user's mental health risk index; Based on a preset mental health risk index conversion rule, converting the user's mental health risk index to obtain the user's mental health index; The constructing of the mental health risk index evaluation model includes: Constructing an initial prediction model; Based on a pre-collected and configured large-scale medical record database, obtaining a number of problem physiological data; wherein, the problem physiological data includes physiological state data and mental state data, the physiological state data includes the problem heart rate, problem blood pressure, and problem blood oxygen of the problem user every day, and the mental state data includes problem heart rate variability parameters; Based on the big data AI algorithm, respectively constructing a physiological data curve graph and a mental data curve graph of the state data and time change according to each piece of the physiological state data and mental state data in the same medical record, to obtain a number of user physiological parameter curve graphs and user mental parameter curve graphs of the same medical record; Respectively performing curve fitting on a number of the user physiological parameter curve graphs and a number of the user mental parameter curve graphs to obtain a physiologically data benchmark curve graph and a mentally data benchmark curve graph after curve fitting; Based on a preset risk value corresponding to each medical record, numerically marking the physiologically data benchmark curve graph and the mentally data benchmark curve graph to obtain a marked physiologically data benchmark curve graph and a marked mentally data benchmark curve graph; Inputting the training physiological data curve graph and training mental data curve graph of the state data and time change constructed by any problem physiological data in the pre-collected and configured large-scale medical record database, the marked physiologically data benchmark curve graph and the marked mentally data benchmark curve graph into the initial prediction model, and training the initial prediction model to obtain the mental health risk index evaluation model; It also includes: after a preset time, based on the user's physiological data collected in a preset time period, constructing a personalized mental health risk index evaluation model to replace the original mental health risk index evaluation model; The constructing process of the personalized mental health risk index evaluation model includes: After receiving the construction instruction, constructing an initial prediction model; Receiving the user's physiological data collected in a preset time period, and constructing personalized user physiological data; wherein, the personalized user physiological data includes a number of user physiological data and their corresponding timestamps; According to the personalized user physiological data, predicting the current user's heart age and daily frequent abnormal emotions; wherein, predicting the current user's heart age through the personalized heart rate, personalized blood pressure, and personalized blood oxygen in the personalized user physiological data, and predicting the current user's daily frequent abnormal emotions through the personalized heart rate variability parameters in the personalized user physiological data; According to the heart age of the current user and the abnormal emotions that occur frequently daily, determine the problem physiological data of the medical records that match the heart age of the current user and the abnormal emotions that occur frequently daily from the large-scale medical record database collected and configured in advance, as the first training data; Obtain the problem physiological data of the medical records corresponding to the user mental health risk index output by the mental health risk index assessment model within a preset time period, as the second training data; Based on the construction method of the mental health risk index assessment model, obtain the second user physiological parameter curve graph and the second user psychological parameter curve graph corresponding to several identical medical records according to the first training data and the second training data; Obtain the second physiological data reference curve graph and the second psychological data reference curve graph after curve fitting according to the second user physiological parameter curve graph and the second user psychological parameter curve; Based on the preset risk value corresponding to each medical record, perform numerical marking on the second physiological data reference curve graph and the second psychological data reference curve graph to obtain the second marked physiological data reference curve graph and the second marked psychological data reference curve graph; Input the state data constructed from any problem physiological data in the first training data and the second training data, the second training physiological data curve graph and the second training psychological data curve graph that change with time, the second marked physiological data reference curve graph and the second marked psychological data reference curve graph into the initial prediction model, and train the initial prediction model to obtain the personalized mental health risk index assessment model.

2. A mental health index assessment method according to claim 1, wherein, The user physiological data includes heart rate, blood pressure, blood oxygen and heart rate variability parameters.

3. A mental health index assessment method according to claim 1, wherein, The step of inputting the user physiological data into the mental health risk index assessment model to obtain the user mental health risk index includes: Perform division processing on the user physiological data to obtain user physiological state data and user psychological state data; wherein, the user physiological state data includes heart rate, blood pressure and blood oxygen, and the user psychological state data includes heart rate variability parameters; Construct a user physiological data curve graph of the state data changing with time according to the user physiological state data; Construct a user psychological data curve graph of the state data changing with time according to the user psychological state data; Input the user physiological data curve graph into the mental health risk index assessment model, and the mental health risk index assessment model outputs the risk value corresponding to the marked physiological data reference curve graph in the mental health risk index assessment model whose similarity to the user physiological data curve graph is greater than a preset first similarity threshold, as the first risk value; Input the user psychological data curve graph into the mental health risk index assessment model, and the mental health risk index assessment model outputs the risk value corresponding to the marked psychological data reference curve graph in the mental health risk index assessment model with a similarity greater than the preset second similarity threshold as the second risk value; Based on the preset weight value, perform a weighted summation calculation on the first risk value and the second risk value to obtain the user mental health risk index.

4. A mental health index assessment system, characterized in that, it includes: A construction module for constructing a mental health risk index assessment model; An acquisition module for real-time acquisition of the user's original pulse wave data based on the PPG method; A processing module for processing the original pulse wave data to obtain the user's physiological data; The processing module is further configured to input the user's physiological data into the mental health risk index assessment model to obtain the user's mental health risk index; A calculation module for converting the user's mental health risk index based on a preset mental health risk index conversion rule to obtain the user's mental health index; The operations performed by the construction module include: Construct an initial prediction model; Based on a pre-collected and configured large-scale medical record database, obtain a number of problem physiological data; wherein, the problem physiological data includes physiological state data and psychological state data, the physiological state data includes the problem heart rate, problem blood pressure, and problem blood oxygen of the problem user every day, and the psychological state data includes the problem heart rate variability parameter; Based on the big data AI algorithm, respectively construct a physiological data curve graph and a psychological data curve graph of the state data and time change according to each piece of the physiological state data and psychological state data in the same medical record, and obtain a number of user physiological parameter curve graphs and user psychological parameter curve graphs of the same medical record; Perform curve fitting on a number of the user physiological parameter curve graphs and a number of the user psychological parameter curve graphs respectively to obtain a physiological data reference curve graph and a psychological data reference curve graph after curve fitting; Based on the preset risk value corresponding to each medical record, perform numerical marking on the physiological data reference curve graph and the psychological data reference curve graph to obtain a marked physiological data reference curve graph and a marked psychological data reference curve graph; Input the training physiological data curve graph and training psychological data curve graph of the state data and time change constructed by any problem physiological data in the pre-collected and configured large-scale medical record database, the marked physiological data reference curve graph and the marked psychological data reference curve graph into the initial prediction model, and train the initial prediction model to obtain the mental health risk index assessment model; The operations performed by the processing module further include: Perform division processing on the user's physiological data to obtain the user's physiological state data and the user's psychological state data; wherein, the user's physiological state data includes heart rate, blood pressure, and blood oxygen, and the user's psychological state data includes heart rate variability parameters; Construct a curve graph of user physiological data with the change of state data and time; Construct a curve graph of user psychological data with the change of state data and time according to the user psychological state data; Input the user physiological data curve graph into the mental health risk index evaluation model, and the mental health risk index evaluation model outputs the risk value corresponding to the marked physiological data reference curve graph in the mental health risk index evaluation model whose similarity to the user physiological data curve graph is greater than a preset first similarity threshold, as the first risk value; Input the user psychological data curve graph into the mental health risk index evaluation model, and the mental health risk index evaluation model outputs the risk value corresponding to the marked psychological data reference curve graph in the mental health risk index evaluation model whose similarity to the user psychological data curve graph is greater than a preset second similarity threshold, as the second risk value; Based on a preset weight value, perform a weighted summation calculation on the first risk value and the second risk value to obtain the user mental health risk index.

5. A mental health index evaluation system according to claim 4, wherein, the user physiological data includes heart rate, blood pressure, blood oxygen and heart rate variability parameters.

6. A mental health index evaluation system according to claim 4, wherein, the processing module is further configured to perform operations including the following: Perform a partitioning process on the user physiological data to obtain user physiological state data and user psychological state data; wherein, the user physiological state data includes heart rate, blood pressure and blood oxygen, and the user psychological state data includes heart rate variability parameters; Construct a curve graph of user physiological data with the change of state data and time according to the user physiological state data; Construct a curve graph of user psychological data with the change of state data and time according to the user psychological state data; Input the user physiological data curve graph into the mental health risk index evaluation model, and the mental health risk index evaluation model outputs the risk value corresponding to the marked physiological data reference curve graph in the mental health risk index evaluation model whose similarity to the user physiological data curve graph is greater than a preset first similarity threshold, as the first risk value; Input the user psychological data curve graph into the mental health risk index evaluation model, and the mental health risk index evaluation model outputs the risk value corresponding to the marked psychological data reference curve graph in the mental health risk index evaluation model whose similarity to the user psychological data curve graph is greater than a preset second similarity threshold, as the second risk value; Based on a preset weight value, perform a weighted summation calculation on the first risk value and the second risk value to obtain the user mental health risk index.

Images