TWI697912B - System and method for evaluating the risk of physiological status and electronic device - Google Patents

Abstract

A system, for evaluating a risk of a physiological status, which comprises a physiological feature sensor, a storage device and a processor. The physiological feature sensor is configured to obtain physiological data which includes a first physiological data obtained at a first time point and a second physiological data obtained at a second time point. The storage device is configured to store the physiological data. The processor obtains at least one feature parameter which reflects a statistics feature of the physiological data according to the physiological data. The processor analyzes the feature parameter by at least one prediction model to obtain risk evaluation information and/or outpatient recommendation information.

Description

Physiological state risk assessment system, method and electronic device

The invention relates to a risk assessment system, method and electronic device of a physiological state.

Generally speaking, people go to hospitals or clinics only when they have abnormal physical conditions. When in a hospital or clinic, people's blood pressure and other physiological data may be checked. Doctors can assess the health of the people based on the physiological data measured at the moment. However, for some people who are prone to stress, or people who have not had enough rest before the measurement, the blood pressure measured at a single point in the hospital is easily distorted, which may lead to inaccurate assessment of the health status of the people by doctors.

The present invention provides a physiological state risk assessment system, method and electronic device, which can provide appropriate risk assessment information and/or medical reminder information based on the physiological data of a user continuously measured within a preset time range, thereby improving the above problem.

The embodiment of the present invention provides a physiological state risk assessment system, which includes a physiological characteristic sensor, a storage device, and a processor. The physiological characteristic sensor is used to obtain a plurality of physiological data of the user within a preset time range. The physiological data includes first physiological data measured at a first time point within the preset time range and second physiological data measured at a second time point within the preset time range, and The first time point is different from the second time point. The storage device is coupled to the physiological characteristic sensor and used for storing the physiological data. The processor is coupled to the physiological characteristic sensor and the storage device. The processor obtains at least one characteristic parameter according to the physiological data, and the characteristic parameter reflects the statistical characteristics of the physiological data. The processor analyzes the characteristic parameters by at least one predictive model to obtain at least one of risk assessment information and medical reminder information. The risk assessment information reflects the risk of the user having a predetermined physiological state.

An embodiment of the present invention further provides an electronic device, which includes a physiological characteristic sensor, a storage device, and a processor. The physiological characteristic sensor is used to obtain a plurality of physiological data of the user within a preset time range. The physiological data includes first physiological data measured at a first time point within the preset time range and second physiological data measured at a second time point within the preset time range, and The first time point is different from the second time point. The storage device is coupled to the physiological characteristic sensor and used for storing the physiological data. The processor is coupled to the physiological characteristic sensor and the storage device. The processor obtains at least one characteristic parameter according to the physiological data. The characteristic parameters reflect the statistical characteristics of the physiological data. The processor analyzes the characteristic parameters by at least one predictive model to obtain at least one of risk assessment information and medical reminder information. The risk assessment information reflects the risk of the user having a predetermined physiological state.

An embodiment of the present invention further provides a method for risk assessment of a physiological state, which includes: obtaining a plurality of physiological data of a user within a preset time range by a physiological characteristic sensor, wherein the physiological data is included in the prediction It is assumed that the first physiological data measured at the first time point within the time range and the second physiological data measured at the second time point within the preset time range, and the first time point is different from the first physiological data Two time points; obtaining at least one characteristic parameter according to the physiological data, wherein the characteristic parameter reflects the statistical characteristics of the physiological data; and analyzing the characteristic parameter by at least one predictive model to obtain risk assessment information and medical reminder information At least one of, wherein the risk assessment information reflects the risk of the user having a predetermined physiological state.

Based on the above, after obtaining the physiological data of the user at different time points by the physiological characteristic sensor, at least one characteristic parameter can be obtained and the characteristic parameter can reflect the statistical characteristics of the physiological data. Then, through the analysis of at least one predictive model, the risk assessment information and/or medical reminder information corresponding to the user can be obtained. By considering the user's physiological data continuously measured within a preset time range, embodiments of the present invention can provide appropriate risk assessment information and/or medical reminder information to assist the general public in managing and/or understanding their physiological state.

In order to make the above-mentioned features and advantages of the present invention more comprehensible, the following specific embodiments are described in detail in conjunction with the accompanying drawings.

Fig. 1 is a schematic diagram of a risk assessment system for a physiological state according to an embodiment of the present invention. Please refer to FIG. 1, the system (also referred to as the risk assessment system of the physiological state) 10 includes a physiological characteristic sensor 11, a storage device 12, a processor 13 and an input/output interface 14. The physiological characteristic sensor 11 may include various sensors for sensing the physiological data of the user, such as a blood pressure sensor, a blood glucose sensor and/or a brain wave detector, and the physiological characteristic sensor 11 The type is not limited to this. In addition, the number of the physiological characteristic sensor 11 may be one or more.

The storage device 12 is coupled to the physiological characteristic sensor 11 and used to store data. For example, the storage device 12 can store the physiological data measured by the physiological characteristic sensor 11. The storage device 12 may include volatile storage media and non-volatile storage media. Among them, the volatile storage medium can be a random access memory, and the non-volatile storage medium can be a read-only memory, a solid state drive, or a traditional hard drive. In addition, the number of storage devices 12 may be one or more.

The processor 13 is coupled to the physiological characteristic sensor 11 and the storage device 12. The processor 13 may include a central processing unit, a graphics processing unit (GPU) or other programmable general-purpose or special-purpose microprocessors, digital signal processors, programmable controllers, and special application products. Body circuits, programmable logic devices or other similar devices or combinations of these devices.

The input/output interface 14 is coupled to the processor 13. The input/output interface 14 is used to output signals or receive signals. For example, the input/output interface 14 may include a screen, a touch screen, a touchpad, a mouse, a keyboard, a physical button, a speaker, a microphone, a wired communication interface, and/or a wireless communication interface, and the type of the input/output interface 14 is different. Limited to this.

In one embodiment, the system 10 may be configured in a single electronic device. The electronic device can be a smart bracelet, smart watch, smart blood pressure meter, smart blood glucose meter, or brain wave detector. The electronic device can sense the physiological data of the user through the physiological characteristic sensor 11. The electronic device can store the measured physiological data through the storage device 12. The electronic device can process the physiological data through the processor 13 and output the processing result through the input/output interface 14. In addition, the electronic device can also send the processing result of the physiological data to the cloud database or other electronic devices (such as a smart phone, a tablet computer, or a notebook computer) through the input/output interface 14.

In one embodiment, the system 10 may be configured in at least two independent electronic devices. For example, the physiological characteristic sensor 11, the storage device 12, and the input/output interface 14 can be configured in a first electronic device (such as a smart bracelet, a smart watch, a smart blood pressure meter, a smart blood glucose meter, or a brain wave detector), and The processor 13 may be configured in a second electronic device (such as a smart phone, a tablet computer, a notebook computer, an industrial computer, or a server). The first electronic device can communicate with the second electronic device via the input/output interface 14 to transmit the measured physiological data to the second electronic device. The second electronic device can process the physiological data from the first electronic device and process the physiological data. In addition, the second electronic device may also output the processing result through an input/output interface and/or store the processing result in a cloud database, etc.

The physiological characteristic sensor 11 can obtain a plurality of physiological data of the user within a predetermined time range. For example, the physiological data may include the physiological data measured at a certain time point (also referred to as the first time point) within the preset time range (also referred to as the first physiological data) and the physiological data measured within the preset time range. Physiological data (also called second physiological data) measured at another time point (also called second physiological data). The first time point is different from the second time point.

FIG. 2 is a schematic diagram of physiological data drawn according to an embodiment of the invention. It should be noted that in the embodiment of FIG. 2, a blood pressure sensor is used as an example of the physiological characteristic sensor 11 in FIG. 1, but it is not intended to limit the present invention.

Please refer to FIG. 1 and FIG. 2, the physiological data 201 is measured by the physiological characteristic sensor 11. The physiological data 201 reflects blood pressure information. For example, the physiological data 201 may reflect the blood pressure information measured by the user at predetermined time intervals within a predetermined time range. The time information 202 reflects the measurement time of the physiological data 201. The physiological data 201 and time information 202 can be stored in the storage device 12.

In this embodiment, the preset time range is 24 hours (ie, one day) and the preset time interval is 60 minutes. In this embodiment, the blood pressure information includes systolic blood pressure and diastolic blood pressure. For example, at time 11:10, the measured systolic and diastolic blood pressure of the user are 138 and 79 millimeters of mercury (mmHg), respectively. At 12:10, the measured systolic and diastolic blood pressure of the user were 144 and 79 mmHg, respectively.

It should be noted that, in one embodiment, the preset time range may be other time ranges. For example, in one embodiment, the preset time range may be a multiple of 24 hours, such as 48 hours or 72 hours. In an embodiment, the predetermined time interval may also be longer (for example, 2 hours) or shorter (for example, 30 minutes). In addition, in an embodiment, the physiological data 201 may include only one of the systolic blood pressure and the diastolic blood pressure.

The processor 13 can obtain at least one characteristic parameter according to the physiological data 201. This characteristic parameter reflects the statistical characteristics of the physiological data 201. For example, this statistical characteristic may include the minimum value of at least some of the values in the physiological data 201, the maximum value of at least some of the values in the physiological data 201, the average of at least some of the values in the physiological data 201, and the variance of at least some of the values in the physiological data 201. , The standard deviation of at least a part of the values in the physiological data 201, the descending ratio of at least a part of the values in the physiological data 201, the rising rate of at least a part of the values in the physiological data 201, the falling rate of at least a part of the values in the physiological data 201, and at least in the physiological data 201 The sudden increase value of a part of the numerical value and/or the sudden decrease value of at least a part of the numerical value in the physiological data 201. In another embodiment, this statistical characteristic can also reflect other types of statistical information (such as a weighted average or median, etc.), which is not limited by the present invention.

In an embodiment, the processor 13 may divide the preset time range into two or more time ranges. Any two time ranges can overlap or not. The processor 13 may analyze the physiological data according to the divided time range to obtain the characteristic parameter. The characteristic parameter may include a period-by-period characteristic parameter, a cross-period characteristic parameter and/or a period-independent characteristic parameter. For example, according to the physiological data measured in the divided single time range, the processor 13 can obtain the time-division characteristic parameters. For example, according to the physiological data measured in multiple divided time ranges, the processor 13 can obtain the cross-period characteristic parameters. Or, based on the physiological data measured in all time ranges, the processor 13 can obtain the characteristic parameters regardless of time period.

Taking FIG. 2 as an example, in one embodiment, the processor 13 may divide 24 hours into a day time period, a sleep time period, and a morning time period. The daytime period can be from 06:00 to 22:00. Sleep period can be 22:00~06:00. The morning time can be 06:00~08:00. According to the physiological data measured in the time range covered by a certain time period among the day time period, the sleep time period and the morning time period, the processor 13 can obtain the time-division characteristic parameters. For example, the processor 13 can obtain the minimum value of systolic blood pressure, the maximum value of systolic blood pressure, the minimum value of diastolic blood pressure, the maximum value of diastolic blood pressure, and the systolic blood pressure according to the physiological data measured during the day time, sleep time or morning time. The average value of BP, the average value of diastolic blood pressure, the variance of systolic blood pressure and/or the variance of diastolic blood pressure, etc. This information belongs to the time period characteristic parameters.

According to the physiological data measured in the time range covered by two of the daytime period, the sleep period and the morning period, the processor 13 can obtain the cross-period characteristic parameters. For example, according to the physiological data measured from the sleep period to the morning period, the processor 13 can obtain the swelling rate and/or the value of the systolic blood pressure (or diastolic blood pressure) from the sleep period to the morning period. Or, based on the physiological data measured from the daytime period to the sleep period, the processor 13 can obtain the ratio of the sudden drop in systolic blood pressure (or diastolic blood pressure) from the daytime period to the sleep period. This information is a cross-period characteristic parameter.

According to the physiological data measured in all the time ranges covered by the day time, sleep time, and morning time, the processor 13 can obtain the characteristic parameters regardless of time. For example, based on all physiological data measured within 24 hours, the processor 13 can obtain the average value of the systolic blood pressure, the average value of the diastolic blood pressure, the standard deviation of the systolic blood pressure, and/or the standard deviation of the diastolic blood pressure during these periods. This information is a characteristic parameter regardless of time period.

In an embodiment, the processor 13 may obtain the average value of N physiological data in one or more time periods according to the following equation (1.1).

(1.1)

(1.1)

In equation (1.1), the characteristic parameter AVG corresponds to the average value of N physiological data, and P(k) corresponds to the value of the physiological data (for example, blood pressure).

In an embodiment, the processor 13 can obtain the variance (also referred to as the average true variance) of the N physiological data in one or more time periods according to the following equation (1.2).

(1.2)

(1.2)

In equation (1.2), the characteristic parameter ARV corresponds to the variance of N physiological data.

In an embodiment, the processor 13 can obtain the decreasing ratio of physiological data in a certain period (also referred to as the first period) to another period (also referred to as the second period) according to the following equation (1.3).

(1.3)

(1.3)

In equation (1.3), the characteristic parameter DIP corresponds to the decreasing proportion of physiological data in the first period to the second period, AVG(Q) corresponds to the average of Q physiological data in the first period, and AVG(P) corresponds to The average value of P physiological data in the second period. For example, assuming that the first time period is a day time period and the second time period is a sleep time period, the characteristic parameter DIP may reflect the decreasing proportion of blood pressure (such as systolic blood pressure or diastolic blood pressure) from the day time period to the sleep period.

In an embodiment, the processor 13 may subtract the average value of the physiological data in a certain period from the average value of the physiological data in another period to obtain the swell value of the physiological data. For example, the processor 13 may subtract the minimum value of the systolic blood pressure in the sleep period from the average value of the systolic blood pressure in the morning period to obtain the value of the sudden increase in systolic blood pressure from the sleep period to the morning period.

In an embodiment, the processor 13 may obtain the swell rate of the physiological data according to the regression coefficient of the physiological data to time in certain two time periods. For example, the processor 13 can obtain the swelling rate of the systolic blood pressure (or diastolic blood pressure) from the sleep period to the morning period according to the following equations (1.4) to (1.6).

(1.4)

(1.4)

(1.5)

(1.5)

(1.6)

(1.6)

In equations (1.4)~(1.6), n is the number of time points collected, and x _i is the time point collected from the time point when the smallest systolic blood pressure occurs during the sleep period to the time point when the morning period ends. , And y _i is the systolic blood pressure collected between the time when the smallest systolic blood pressure occurs during the sleep period and the time when the morning period ends. In addition, the processor 13 can obtain corresponding characteristic parameters based on other types of algorithms, and the present invention is not limited.

The processor 13 can analyze the obtained characteristic parameters by at least one predictive model to obtain risk assessment information and/or medical treatment reminder information corresponding to the user. For example, at least one of the aforementioned characteristic parameters may be used to obtain risk assessment information and/or medical treatment reminder information. This risk assessment information can reflect the risk of the user's default physiological state. This risk can be expressed in terms of probability or level. For example, this risk assessment information can reflect the probability of abnormal physiological conditions such as hypertension and diabetes in the future. It should be noted that this risk assessment information is not medical diagnostic information, but only a rough analysis result obtained by analyzing the measured physiological data. In addition, the medical reminder information can remind the user based on the risk assessment information to check and prevent possible preset physiological conditions.

In an embodiment, the number (or type) of prediction models may be one or more. For example, the prediction model may include at least two of a parameter statistical model, an integrated learning model, and a machine learning model. The parameter statistical model may include a Cox model (also known as a Cox regression model or a proportional hazards regression (Proportional Hazards) model). For example, the parameter statistical model may have the characteristic that the ratio of the calculation results between different groups does not change with time. Comparison of parameter statistical models can explain the performance differences of characteristic parameters between different groups. The ensemble learning (Ensemble Learning) model may include a random forest (Random Forest) model. For example, the ensemble learning model can use multiple decision trees and generate prediction results through voting or integration. The integrated learning model can provide more robust prediction results. Machine learning models include Support Vector Machine (SVM) models. For example, a machine learning model can find a non-linear plane to maximize segmentation of different groups and provide prediction results.

In different embodiments, the more predictive models are used, the more accurate the obtained risk assessment information and/or medical reminder information may be. In the following embodiments, the above three prediction models (ie, three types of prediction models) are used simultaneously as an example, but they are not intended to limit the present invention.

In an embodiment, the processor 13 may run the first prediction model, the second prediction model, and the third prediction model. The first prediction model, the second prediction model, and the third prediction model may be a parameter statistical model, an integrated learning model, and a machine learning model, respectively. The processor 13 may analyze the characteristic parameters through the first prediction model, the second prediction model, and the third prediction model to obtain the first evaluation information, the second evaluation information, and the third evaluation information, respectively. The processor 13 may obtain the risk assessment information and/or medical treatment reminder information according to the first assessment information, the second assessment information, and the third assessment information.

In an embodiment, the processor 13 may obtain the risk value corresponding to the preset physiological state according to the following equation (2.1).

(2.1)

(2.1)

In equation (2.1), the risk value RK corresponds to the user's risk of a preset physiological state, RK(1) corresponds to the first assessment information generated by the first prediction model, and RK(2) corresponds to the second prediction model generated by the second prediction model. Evaluation information, and RK(3) corresponds to the third evaluation information generated by the third prediction model. The higher the risk value RK, the greater the risk that the user will have a preset physiological state. The processor 13 can obtain the risk assessment information according to the risk value RK.

In one embodiment, the processor 13 may obtain the prediction error rate of the first prediction model (also referred to as the first prediction error rate), the prediction error rate of the second prediction model (also referred to as the second prediction error rate), and the first prediction error rate. The prediction error rate of the three prediction models (also called the third prediction error rate). The processor 13 can obtain at least one of risk assessment information and medical care reminder information according to the first assessment information, the second assessment information, the third assessment information, the first prediction error rate, the second prediction error rate, and the third prediction error rate . For example, the processor 13 may obtain the risk value corresponding to the preset physiological state according to the following equations (3.1) to (3.4).

(3.1)

(3.1)

(3.2)

(3.2)

(3.3)

(3.3)

(3.4)

(3.4)

In equations (3.1)~(3.4), parameter a corresponds to the prediction error rate of the first prediction model (ie, the first prediction error rate), and parameter b corresponds to the prediction error rate of the second prediction model (ie, the second prediction error rate) , And the parameter c corresponds to the prediction error rate of the third prediction model (ie, the third prediction error rate). In other words, equations (3.1)~(3.4) consider the different prediction error rates of different prediction models. The processor 13 can adjust the weights α, β, and γ of the prediction results generated by different prediction models. In one embodiment, the prediction accuracy using equations (3.1) to (3.4) may be higher than the prediction accuracy using equation (2.1).

In one embodiment, the first evaluation information may reflect that the first prediction model believes that the user needs or does not need medical treatment. The second evaluation information may reflect that the second prediction model believes that the user needs or does not need medical treatment. The third evaluation information may reflect that the third prediction model believes that the user needs or does not need medical treatment.

In one embodiment, the processor 13 may refer to the first evaluation information, the second evaluation information, and the third evaluation information to obtain medical reminder information in a majority decision manner. For example, if two or more of the three prediction models believe that the user needs medical treatment, the processor 13 can generate medical treatment reminder information that reminds the user to seek medical treatment.

In one embodiment, the processor 13 may also use other mechanisms to obtain medical reminder information. For example, in one embodiment, as long as there is a predictive model that the user needs to seek medical treatment, the processor 13 can generate medical treatment reminder information to remind the user to seek medical treatment. Or, in one embodiment, as long as there is a predictive model that the user does not need to seek medical treatment, the processor 13 does not generate medical treatment reminding information to remind the user to seek medical treatment.

Fig. 3 is a flowchart of a method for risk assessment of a physiological state according to an embodiment of the present invention. Referring to FIG. 3, in step S301, a plurality of physiological data of the user within a preset time range is obtained by the physiological characteristic sensor. The physiological data includes first physiological data measured at a first time point within the preset time range and second physiological data measured at a second time point within the preset time range, and The first time point is different from the second time point. In step S302, at least one characteristic parameter is obtained according to the physiological data. The characteristic parameters reflect the statistical characteristics of the physiological data. In step S303, the characteristic parameters are analyzed by at least one predictive model to obtain risk assessment information and/or medical reminder information. The risk assessment information reflects the risk of the user having a predetermined physiological state.

It should be noted that each step in FIG. 3 has been described in detail above, and will not be repeated here. Each step in FIG. 3 can be implemented as multiple program codes or circuits, and the present invention is not limited. In addition, the method in FIG. 3 can be used in conjunction with the above exemplary embodiments, or can be used alone, and the present invention is not limited.

In summary, after obtaining the physiological data of the user at different time points by the physiological characteristic sensor, at least one characteristic parameter can be obtained and the characteristic parameter can reflect the statistical characteristics of the physiological data. Then, through the analysis of at least one predictive model, the risk assessment information and/or medical reminder information corresponding to the user can be obtained. By considering the user's physiological data continuously measured within a preset time range, embodiments of the present invention can provide appropriate risk assessment information and/or medical reminder information to assist the general public in managing and/or understanding their physiological state. In addition, the risk assessment information and/or medical reminder information can also assist doctors or nurses in assessing the physiological state of the user.

Although the present invention has been disclosed in the above embodiments, it is not intended to limit the present invention. Anyone with ordinary knowledge in the technical field can make slight changes and modifications without departing from the spirit and scope of the present invention. The scope of protection of the present invention shall be subject to those defined by the attached patent scope.

10: Physiological state risk assessment system 11: Physiological characteristic sensor 12: Storage device 13: Processor 14: Input/output interface 201: Physiological data 202: Time information S301~S303: Steps

Fig. 1 is a schematic diagram of a risk assessment system for a physiological state according to an embodiment of the present invention. FIG. 2 is a schematic diagram of physiological data drawn according to an embodiment of the invention. Fig. 3 is a flowchart of a method for risk assessment of a physiological state according to an embodiment of the present invention.

S301~S303: steps

Claims (17)

A risk assessment system for a physiological state, comprising: a physiological characteristic sensor for obtaining a plurality of physiological data of a user within a preset time range, wherein the plurality of physiological data are included in the preset time range A first physiological data measured at a first time point and a second physiological data measured at a second time point within the preset time range, and the first time point is different from the second time point A storage device, coupled to the physiological characteristic sensor and used to store the plurality of physiological data; and a processor, coupled to the physiological characteristic sensor and the storage device, wherein the processor according to the plurality of The physiological data obtains at least one characteristic parameter, the at least one characteristic parameter reflects a statistical characteristic of the plurality of physiological data, and the processor analyzes the at least one characteristic parameter by at least one prediction model to obtain at least one evaluation information and at least one prediction error According to the at least one assessment information and the at least one prediction error rate, at least one of a risk assessment information and a medical treatment reminder information is obtained according to the at least one assessment information, wherein the risk assessment information reflects the risk of a predetermined physiological state of the user. For the physiological state risk assessment system described in item 1 of the scope of patent application, the operation of the processor to obtain the at least one characteristic parameter according to the plurality of physiological data includes: dividing the preset time range into at least a first time Range and a second time range, wherein the first time point is included in the first preset time range, and the first Two time points are included in the second preset time range; and a span of the at least one characteristic parameter is obtained according to the first time range, the second time range, the first physiological data, and the second physiological data Characteristic Parameters. For the physiological state risk assessment system described in item 1 of the scope of patent application, the operation of the processor to obtain the at least one characteristic parameter according to the plurality of physiological data includes: dividing the preset time range into at least a first time Range and a second time range, wherein the first time point and the second time point are both included in the first time range; and obtained according to the first time range, the first physiological data, and the second physiological data One of the at least one feature parameter is a time-divided feature parameter. The risk assessment system of the physiological state according to the first item of the patent application, wherein the at least one prediction model includes a first prediction model, a second prediction model, and a third prediction model, and the processor uses the at least The operation of analyzing the at least one characteristic parameter by a predictive model to obtain the at least one of the risk assessment information and the medical treatment reminder information includes: analyzing the at least one characteristic parameter by the first predictive model to obtain first assessment information Analyze the at least one characteristic parameter by the second prediction model to obtain a second evaluation information; analyze the at least one characteristic parameter by the first prediction model to obtain a second evaluation information; and Obtain at least one of the risk assessment information and the medical treatment reminder information according to the first assessment information, the second assessment information, and the third assessment information. For example, the physiological state risk assessment system described in item 4 of the scope of patent application, wherein the processor obtains the risk assessment information and the medical treatment reminder information according to the first assessment information, the second assessment information, and the third assessment information The at least one of the operations includes: obtaining a first prediction error rate of the first prediction model, a second prediction error rate of the second prediction model, and a third prediction error rate of the third prediction model; and According to the first evaluation information, the second evaluation information, the third evaluation information, the first prediction error rate, the second prediction error rate, and the third prediction error rate to obtain the risk assessment information and the medical treatment reminder information At least one of them. According to the first item of the scope of patent application, the risk assessment system for a physiological state, wherein the at least one prediction model includes at least two of a parameter statistical model, an integrated learning model, and a machine learning model. For the physiological state risk assessment system described in item 1 of the scope of patent application, the preset time range is 24 hours. The physiological state risk assessment system described in item 1 of the scope of patent application, wherein the physiological data reflects blood pressure information. An electronic device, comprising: a physiological characteristic sensor for obtaining a plurality of physiological data of a user in a preset time range, wherein the plurality of physiological data are included in the preset time range A first physiological data measured at a first time point and a second physiological data measured at a second time point within the preset time range, and the first time point is different from the second time point A storage device, coupled to the physiological characteristic sensor and used to store the plurality of physiological data; and a processor, coupled to the physiological characteristic sensor and the storage device, wherein the processor according to the plurality of The physiological data obtains at least one characteristic parameter, the at least one characteristic parameter reflects a statistical characteristic of the plurality of physiological data, and the processor analyzes the at least one characteristic parameter by at least one prediction model to obtain at least one evaluation information and at least one prediction error According to the at least one assessment information and the at least one prediction error rate, at least one of a risk assessment information and a medical treatment reminder information is obtained according to the at least one assessment information, wherein the risk assessment information reflects the risk of a predetermined physiological state of the user. A method for risk assessment of a physiological state includes: obtaining a plurality of physiological data of a user within a preset time range by a physiological characteristic sensor, wherein the plurality of physiological data includes the physiological data within the preset time range A first physiological data measured at a first time point and a second physiological data measured at a second time point within the preset time range, and the first time point is different from the second time point; Obtain at least one characteristic parameter based on the plurality of physiological data, wherein the at least one characteristic parameter reflects a statistical characteristic of the plurality of physiological data; and analyze the at least one characteristic parameter by at least one predictive model to obtain at least one evaluation At least one of risk assessment information and medical treatment reminder information is obtained according to the at least one assessment information and the at least one prediction error rate, wherein the risk assessment information reflects the occurrence of a Presuppose the risk of physiological state. According to the method for risk assessment of the physiological state according to item 10 of the scope of patent application, the step of obtaining the at least one characteristic parameter according to the plurality of physiological data includes: dividing the preset time range into at least a first time range and a first time range A second time range, wherein the first time point is included in the first preset time range, and the second time point is included in the second preset time range; and according to the first time range, the second time range The time range, the first physiological data and the second physiological data obtain a cross-period characteristic parameter of the at least one characteristic parameter. According to the method for risk assessment of the physiological state according to item 10 of the scope of patent application, the step of obtaining the at least one characteristic parameter according to the plurality of physiological data includes: dividing the preset time range into at least a first time range and a first time range A second time range, wherein the first time point and the second time point are both included in the first time range; and the at least one is obtained according to the first time range, the first physiological data, and the second physiological data One-time period characteristic parameter in the characteristic parameter. The method for assessing the risk of a physiological state according to claim 10, wherein the at least one prediction model includes a first prediction model, a second prediction model, and a third prediction model, and the at least one prediction model Analyze the at least one feature The step of obtaining the at least one of the risk assessment information and the medical treatment reminder information by using a characteristic parameter includes: analyzing the at least one characteristic parameter by the first prediction model to obtain a first assessment information; and by the second prediction The model analyzes the at least one characteristic parameter to obtain a second evaluation information; analyzes the at least one characteristic parameter by the first prediction model to obtain a second evaluation information; and according to the first evaluation information, the second evaluation information, and The third assessment information obtains the at least one of the risk assessment information and the medical treatment reminder information. For example, the method for risk assessment of a physiological state as described in item 13 of the scope of patent application, wherein the at least one of the risk assessment information and the medical treatment reminder information is obtained according to the first assessment information, the second assessment information, and the third assessment information One of the steps includes: obtaining a first prediction error rate of the first prediction model, a second prediction error rate of the second prediction model, and a third prediction error rate of the third prediction model; and according to the first prediction model One assessment information, the second assessment information, the third assessment information, the first prediction error rate, the second prediction error rate, and the third prediction error rate obtain the at least one of the risk assessment information and the medical treatment reminder information one. According to the method for assessing the risk of a physiological state according to claim 10, the at least one prediction model includes at least two of a parameter statistical model, an integrated learning model, and a machine learning model. As described in item 10 of the scope of patent application, the risk assessment method of the physiological state, wherein the preset time range is 24 hours. As described in item 10 of the scope of patent application, the physiological state risk assessment method, wherein the physiological data reflects blood pressure information.

Images