Apparatus and method for estimating biological information

Abstract

A device and method for estimating biological information are disclosed. According to one aspect of the present invention, the device for estimating biological information includes: a biological signal acquirer configured to acquire a biological signal; a processor configured to extract one or more first eigenvalues ​​from the acquired biological signal, determine a scale factor based on the extracted first eigenvalues, and estimate the biological information based on the determined scale factor and the first eigenvalues.

Description

[0001] This application claims priority from Korean Patent Application No. 10-2018-0019441 filed on February 19, 2018, in the Korean Intellectual Property Office, and Korean Patent Application No. 10-2019-0002484 filed on January 8, 2019, in the Korean Intellectual Property Office, the disclosures of which are incorporated herein in their entirety by reference for all purposes. Technical Field

[0002] The following description relates to a technique for non-invasively estimating bio-information using bio-signals. Background Art

[0003] Recently, with the aging of the population, a surge in medical expenses, and a shortage of medical personnel for professional medical services, research on IT-medical convergence technology in which IT technology and medical technology are combined is being actively conducted.

[0004] Specifically, the monitoring of the health status of the human body is not limited to medical institutions, but is expanding to the field of mobile medical care that can monitor the health status of users anytime and anywhere in daily life at home or in the office.

[0005] Typical examples of biosignals indicating a health condition of an individual include electrocardiogram (ECG) signals, plethysmogram (PPG) signals, electromyogram (EMG) signals, etc., and various biosignal sensors have been developed to measure these signals in daily life.

[0006] For example, according to research on PPG signals, the entire PPG signal is a superposition of a propagation wave starting from the heart to a distal end portion of the body and a reflected wave returning from the distal end portion.

[0007] Furthermore, it is known that information for estimating blood pressure can be obtained by extracting various features associated with propagating waves or reflected waves.

[0008] However, in some cases, due to the decline in biosignal quality, interference from motion noise, etc., the method of estimating bioinformation from biosignals to monitor health conditions in daily life may provide unstable estimation results of the bioinformation, and various studies have been conducted to solve this problem. Summary of the invention

[0009] Provided are an apparatus and method for estimating bio-information, in which bio-information can be accurately and stably estimated even when the bio-signal is measured in an unstable environment by appropriately performing scale conversion of features extracted from the bio-signal.

[0010] In a general aspect, a device for estimating bio-information is provided, the device comprising: a bio-signal acquirer configured to acquire a bio-signal; and a processor configured to extract one or more first eigenvalues ​​from the acquired bio-signal, determine a scaling factor based on the extracted first eigenvalues, and estimate the bio-information based on the determined scaling factor and the first eigenvalues.

[0011] The first characteristic value may include: a characteristic associated with cardiac output (CO), a characteristic associated with total peripheral resistance (TPR), and a combination thereof.

[0012] The processor may calculate a second eigenvalue by combining the first eigenvalues, and may determine the scale factor by adjusting a reference scale factor using a scale control ratio calculated based on the second eigenvalue.

[0013] Furthermore, the processor may calculate the second eigenvalue by combining at least one of the individual variations and the combined variations of the first eigenvalues.

[0014] The processor may calculate the proportional control ratio according to the magnitude of the second eigenvalue by applying the calculated second eigenvalue to the proportional control ratio decision function.

[0015] Here, the proportional control ratio decision function may have a valley shape, wherein in the valley shape, the proportional control ratio has a minimum value at a point referenced to the second eigenvalue, and increases from the referenced second eigenvalue as the second eigenvalue changes, and when the second eigenvalue falls outside a threshold range, the proportional control ratio is saturated to a predetermined proportional control ratio.

[0016] The processor may calculate individual proportional control ratios for the first characteristic value, and may determine the proportional factor by adjusting a reference proportional factor using the proportional control ratios calculated based on statistical values ​​of the individual proportional control ratios.

[0017] Furthermore, the processor may calculate a third eigenvalue by combining the first eigenvalues, and estimate the biometric information based on the calculated third eigenvalue and the determined scale factor.

[0018] Furthermore, the processor may estimate the biometric information by multiplying a difference between the third eigenvalue and the reference third eigenvalue by a determined scale factor, and adding an offset value for the biometric information to the multiplied value.

[0019] In response to the first characteristic value exceeding a predetermined threshold, the processor may determine a reference scale factor as the scale factor.

[0020] The processor may normalize the first eigenvalue based on the reference first eigenvalue.

[0021] The apparatus for estimating bio-information may further include: an output section configured to output the bio-signal, the feature value of the bio-signal, the multiplication coefficient ratio control factor, the multiplication coefficient ratio, and the bio-information.

[0022] Here, the biological information may include blood pressure, pulse, cardiac output, blood sugar, triglyceride, and keratin.

[0023] In another general aspect, a method for estimating biological information is provided, the method comprising: acquiring a biological signal; extracting one or more first eigenvalues ​​from the acquired biological signal; determining a scaling factor based on the extracted first eigenvalues; and estimating the biological information based on the determined scaling factor and the first eigenvalues.

[0024] The step of determining the proportional factor may include: calculating a second eigenvalue by combining the first eigenvalues; calculating a proportional control ratio based on the second eigenvalues; and determining the proportional factor by adjusting a reference proportional factor based on the calculated proportional control ratio.

[0025] Furthermore, the step of calculating the second eigenvalue may include combining at least one of individual variations and combined variations of the first eigenvalues.

[0026] In this case, the step of calculating the proportional control ratio may include calculating the proportional control ratio according to the size of the second eigenvalue by applying the calculated second eigenvalue to a proportional control ratio decision function.

[0027] In addition, the proportional control ratio decision function may have a valley shape, wherein in the valley shape, the proportional control ratio has a minimum value at a point with reference to the second eigenvalue, and increases from the reference to the second eigenvalue as the second eigenvalue changes, and when the second eigenvalue falls outside a threshold range, the proportional control ratio is saturated to a predetermined proportional control ratio.

[0028] In this case, the step of determining the proportional factor may include: calculating individual proportional control ratios for the first characteristic value; calculating the proportional control ratio based on statistical values ​​of the individual proportional control ratios; and determining the proportional factor by adjusting the reference proportional factor based on the calculated proportional control ratio.

[0029] In addition, the step of estimating the biometric information may include: calculating a third eigenvalue by combining the first eigenvalues; estimating the biometric information by multiplying the difference between the calculated third eigenvalue and the reference third eigenvalue by a determined proportional factor, and adding an offset value for the biometric information to the multiplied value.

[0030] The step of determining the proportional factor may include: determining whether the first eigenvalue exceeds a predetermined threshold; and determining a reference proportional factor as the proportional factor when it is determined that the first eigenvalue exceeds the predetermined threshold. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 is a block diagram illustrating an example of an apparatus for estimating biological information.

[0032] Figure 2 is an exemplary diagram illustrating an example of a biosignal.

[0033] Figure 3 is an exemplary diagram for explaining an example of adjusting a scale factor.

[0034] Figure 4 is an explanatory diagram for explaining an example of calculating the second eigenvalue.

[0035] Figure 5 is an exemplary diagram showing an example of calculating a scale control ratio.

[0036] Figure 6 is an explanatory diagram for explaining another example of calculating the proportional control ratio.

[0037] Figure 7 is a block diagram illustrating another example of an apparatus for estimating biological information.

[0038] Figure 8 is a flowchart illustrating an example of a method of estimating biological information.

[0039] Fig. 9 is a flowchart illustrating another example of a method of estimating biological information.

[0040] Throughout the drawings and detailed description, unless otherwise described, the same drawing reference numerals will be understood to refer to the same elements, features, and structures. The relative sizes and descriptions of these elements may be exaggerated for clarity, illustration, and convenience. DETAILED DESCRIPTION

[0041] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. It should be noted that the same reference symbols represent the same parts wherever they are used, even in different drawings. In the following description, when known functions and configurations contained herein obscure the subject matter of the present invention, their detailed description may be omitted.

[0042] Unless a specific order is clearly stated in the context of the present disclosure, the process steps described herein may be performed differently from the specific order. That is, each step may be performed in a specific order, at substantially the same time, or in a reverse order.

[0043] Furthermore, the terms used throughout the present specification are defined in view of functions according to the exemplary embodiments, and may be changed according to the purpose of a user or an administrator or precedents, etc. Therefore, the definition of the terms should be made based on the overall context.

[0044] It will be understood that although the terms first, second, etc. can be used to describe various elements here, these elements should not be limited by these terms. These terms are only used to distinguish one element from another element. Unless otherwise explicitly stated, any reference to the singular may include the plural. In this specification, it should be understood that terms such as "including" or "having" are intended to indicate the presence of features, quantities, steps, actions, components, parts or combinations thereof disclosed in this specification, but are not intended to exclude the possibility that one or more other features, quantities, steps, actions, components, parts or combinations thereof may be present or may be increased.

[0045] Hereinafter, embodiments of an apparatus and method for estimating bio-information will be described in detail with reference to the accompanying drawings.

[0046] Figure 1 is a block diagram showing an example of an apparatus for estimating biological information, Figure 2 is an exemplary diagram illustrating an example of a biosignal.

[0047] The bio-information estimating apparatus 100 may acquire a bio-signal, may extract features from the acquired bio-signal, and may estimate bio-information based on the extracted features.

[0048] For example, the biological information estimation device 100 may obtain Figure 2 The first eigenvalue is extracted from the vascular volume graph (PPG) signal composed of the superposition of the propagation wave and the reflection wave shown in FIG. The blood pressure can be estimated by multiplying the first eigenvalue with an appropriate scale factor and adding an offset value (such as a blood pressure value in a steady state) to the multiplied first eigenvalue. Here, the first eigenvalue may include a characteristic f associated with the cardiac output (CO). 1_co , characteristics associated with total peripheral resistance (TPR) 1_TPR and combinations thereof, wherein the combination may include a feature f associated with cardiac output (CO) 1_co and characteristics associated with total peripheral resistance (TPR) 1_TPRThe addition, subtraction, multiplication, division, etc. of the value after addition, subtraction, multiplication, division, etc. may also include adding a real number to the added value, subtraction, multiplication, division, etc., or subtracting a real number from the added value, subtraction, multiplication, division, etc. Cardiac output (CO) is the amount of blood pumped by the heart per unit time. Total peripheral resistance (TPR) is the total resistance provided by systemic arteries to blood flow through the systemic arteries. For example, the feature f can be determined by extracting a first peak and a second peak in the PPG signal and determining the ratio of the first peak to the second peak as 1_co value to obtain the characteristic f associated with cardiac output (CO) 1_co , but example embodiments are not limited thereto. Doppler ultrasound, chest bioimpedance, pulse profile analysis, or multi-linear regression analysis based on the pulse width of the PPG signal may be used to extract features associated with cardiac output (CO) 1_co and characteristics associated with total peripheral resistance (TPR) 1_TPR .

[0049] In this case, the bio-information estimating apparatus 100 may stably estimate the bio-information by calculating an adaptive scale factor based on a change in the extracted first eigenvalue.

[0050] For example, the bio-information estimation apparatus 100 may calculate a proportional control ratio for adjusting a proportional factor based on the extracted first eigenvalue, and may determine a proportional factor adapted to a change in the first eigenvalue by adjusting a reference proportional factor based on the calculated proportional control ratio. For example, the reference proportional factor may have a predetermined value, and the proportional control ratio may have a value greater than 0 and less than or equal to 1. If the extracted first eigenvalue has a value greater than a predetermined threshold value, the proportional control ratio may have a value of 1, thereby estimating the bio-information using the reference proportional factor without adjusting the reference proportional factor. If the extracted first eigenvalue has a value less than or equal to a predetermined threshold value, the proportional control ratio may have a value between 0 and 1, and the reference proportional factor may be scaled down according to the value of the proportional control ratio.

[0051] As described above, by adaptively determining the scale factor, the bio-information estimation apparatus 100 can stably estimate bio-information even when the bio-signal is estimated in a harsh environment such as when an unstable bio-signal is acquired due to motion noise or the like.

[0052] Figure 3 is an exemplary diagram explaining an example of adjusting a scale factor.

[0053] Reference Figure 3, depending on whether the change of the feature extracted from the biological signal belongs to the homoeostasis maintaining region, the linear change region, or the nonlinear saturation region, the change shape of the biological information may be changed with respect to the change of the feature extracted from the biological signal.

[0054] For example, in the body balance maintenance area, the change of biological information is smaller than the change of the characteristics of the biological signal according to the characteristics of the human body to maintain the body balance; in the linear change area, the change of biological information has a predetermined correlation with the change of the characteristics of the biological signal; in the nonlinear saturation area, the biological information changes irregularly (has a nonlinear correlation or no specific correlation with the change of the characteristics of the biological signal).

[0055] Therefore, the bioinformation estimation device 100 can determine whether the first eigenvalue extracted from the biosignal or the change in the first eigenvalue belongs to the body balance maintenance region, the linear change region or the nonlinear saturation region, and can stably and accurately estimate the bioinformation by adjusting the proportional factor based on the determination.

[0056] More specifically, the bioinformation estimation device 100 can determine that the first eigenvalue belongs to the body balance maintenance area when the change in the first eigenvalue is greater than the change in the biosignal; can determine that the first eigenvalue belongs to the linear change area when the correlation between the change in the first eigenvalue and the change in the biosignal is greater than or equal to a predetermined correlation value; and can determine that the first eigenvalue belongs to the nonlinear saturation area when the correlation between the change in the first eigenvalue and the change in the biosignal is less than a predetermined correlation value.

[0057] For example, in a case where the first eigenvalue belongs to the homeostasis maintaining region, the bio-information estimating apparatus 100 may reduce the scale factor by considering the homeostasis maintaining characteristic, thereby reducing the influence of the change in the first eigenvalue on the change in the bio-information.

[0058] In another example, in a case where the first eigenvalue belongs to the linear change region, the bio-information estimating apparatus 100 may determine a predetermined reference scale factor as the scale factor so that the change in the first eigenvalue may be reflected in the bio-information as it is.

[0059] In another example, when the first eigenvalue belongs to the nonlinear saturation region, the bioinformation estimation device 100 may appropriately adjust the scaling factor by applying the first eigenvalue to a bioinformation estimation model pre-generated by a nonlinear function or machine learning, so that changes in the first eigenvalue can be reflected in changes in the bioinformation.

[0060] As described above, by appropriately adjusting the scale factor according to the change of the feature extracted from the bio-signal, the bio-information estimation apparatus 100 can accept the feature change that can affect the estimation of the bio-information as it is, and can reflect the feature change in the estimation of the bio-information. In addition, the bio-information estimation apparatus 100 can adaptively reduce the feature change that does not affect the estimation of the bio-information, and can reduce the error caused by unnecessary motion noise. Therefore, even in an environment where an unstable bio-signal is measured, the bio-information estimation apparatus 100 can stably estimate the bio-information.

[0061] For ease of explanation, if necessary, the following description will be made using an example of estimating blood pressure based on a biological signal. However, the biological information is not limited thereto and may include at least one of blood pressure, pulse, oxygen saturation, pressure index, cardiac output, blood sugar, triglyceride, keratin, etc.

[0062] Return to reference Figure 1 , an example of estimating biological information by the biological information estimating apparatus 100 will be described in detail below.

[0063] like Figure 1 As shown in FIG. 1 , the bio-information estimating apparatus 100 includes a bio-signal acquirer 110 and a processor 120. Here, the processor 120 may be composed of one or more processors, memories, and combinations thereof.

[0064] The biosignal acquirer 110 may acquire a biosignal of a user.

[0065] Here, the biosignal may include an electrocardiogram (ECG) signal, a plethysmogram (PPG) signal, an electromyogram (EMG) signal, a ballistocardiogram (BCG) signal, a cardiac output (CO) signal, a total peripheral resistance (TRP) signal, heart sounds, and the like.

[0066] For example, the biosignal acquirer 110 may include a sensor including at least one of the following items: one or more electrodes for measuring biosignals, a pressure sensor, and a photodetector module including a light source and a detector. The biosignal acquirer 110 may be directly interfaced with a user through the sensor to acquire biosignals.

[0067] In addition, the bio-signal acquirer 110 may communicate with an external device to receive the user's bio-signal data from the external device. For example, the bio-signal acquirer 110 may receive the user's bio-signal data from the external device using the following methods: Bluetooth communication, low-power Bluetooth (BLE) communication, near field communication (NFC), WLAN communication, Zigbee communication, infrared data association (IrDA) communication, Wi-Fi direct (WFD) communication, ultra-wideband (UWB) communication, Ant + communication, WIFI communication, radio frequency identification (RFID) communication, etc. In addition, examples of external devices may include cellular phones, smart phones, tablet PCs, laptop computers, personal digital assistants (PDAs), portable multimedia players (PMPs), navigators, MP3 players, digital cameras, wearable devices, etc. However, the external device is not limited to the above examples and may include various devices that store the user's bio-signal data.

[0068] The processor 120 may extract one or more first feature values ​​f from the acquired biological signal. 1 .

[0069] The first feature value extracted from the biosignal may indicate a feature having a predetermined correlation with the bioinformation desired to be estimated. One or more first feature values ​​may be extracted; in the case of estimating blood pressure by using, for example, the bioinformation estimating apparatus 100, the first feature value may include a feature f associated with cardiac output (CO). 1_co , characteristics associated with total peripheral resistance (TPR) 1_TPR and combinations thereof, etc., wherein cardiac output (CO) indicates the amount of blood pumped out by the heart in one minute. The first feature value may be changed according to the type of biological information desired to be estimated.

[0070] When the first feature value is extracted from the biosignal, the processor 120 may convert the first feature value.

[0071] For example, the processor 120 may calculate the first feature value extracted from the biosignal by dividing the first feature value extracted in the reference state (eg, f 1_co_ref 、f 1_TPR_ref etc.) (e.g., f 1_co_norm =f 1_co / f 1_co_ref 、f 1_TPR_norm =f 1_TPR / f 1_TPR_ref etc.), to normalize the extracted first eigenvalue.

[0072] In this case, the reference state is a state other than a motion state or a sleep state, and may represent, for example, a state in which the pulse and respiratory rate are stable, or a state in which the blood pressure measured by an external device for measuring blood pressure is maintained without substantial change.

[0073] In addition, a reference second eigenvalue and a reference third eigenvalue to be described later may refer to second eigenvalues ​​and third eigenvalues ​​calculated based on a reference first eigenvalue extracted in a reference state. Hereinafter, a first eigenvalue may refer to a first eigenvalue normalized using a reference first eigenvalue.

[0074] The processor 120 may determine a scale factor based on the extracted first eigenvalue.

[0075] The scaling factor may be a coefficient for adjusting the scale of the first eigenvalue extracted for estimating the biometric information, but is not limited thereto, and as will be described later, the scaling factor may be a coefficient for adjusting the scale of the third eigenvalue calculated based on the first eigenvalue for estimating the biometric information.

[0076] For example, the processor 120 may calculate the second eigenvalue f by combining multiple first eigenvalues. sc , and the proportional control ratio can be calculated based on the second eigenvalue.

[0077] Here, the second characteristic value may indicate a characteristic value for determining the proportional control ratio, and the processor 120 may calculate the second characteristic value using individual changes or combined changes of the first characteristic values.

[0078] Figure 4 is an explanatory diagram for explaining an example of calculating the second eigenvalue.

[0079] Reference Figure 1 and Figure 4 , Figure 4 The first eigenvalue f according to the elapsed time is shown 1a and the first eigenvalue f 1b In this case, the processor 120 may be based on the reference second eigenvalue f sc, ref to calculate the change of the first eigenvalue.

[0080] When calculating one or more first eigenvalues, the processor 120 may calculate a second eigenvalue by using individual changes and / or combined changes of the first eigenvalues.

[0081] In this case, the combined variation can be represented by the first eigenvalue f 1a and the first eigenvalue f 1b A linear combination of 1a and f 1b The value obtained by adding, subtracting, multiplying, or a combination of adding, subtracting, and multiplying, etc., is the same as the reference second eigenvalue f sc, ref The difference between the individual changes can be expressed as the first eigenvalue f 1a and the first eigenvalue f 1bEach of the reference second eigenvalues ​​f sc, ref The difference between.

[0082] For example, the following Equation 1 may represent an example of calculating the second eigenvalue by using a combined variation of the first eigenvalue.

[0083] [Equation 1]

[0084] f sc = f 1a + f 1b – f sc,ref

[0085] Furthermore, the following Equation 2 may represent another example of calculating the second eigenvalue by using the individual variation of the first eigenvalue.

[0086] [Equation 2]

[0087] f sc = |f 1a – f sc,ref | + |f 1b – f sc,ref | + f sc,ref

[0088] Furthermore, the following Equation 3 may represent another example of calculating the second eigenvalue by using a combination of the combined variation and the individual variation of the first eigenvalue.

[0089] [Equation 3]

[0090] f sc = (f 1a + f 1b + |f 1a – f sc,ref | + |f 1b – f sc,ref |) / 2

[0091] As described above, the processor 120 may calculate the second feature value for determining the proportional control ratio by using the individual first feature values ​​and / or a combination thereof.

[0092] In another example, the processor 120 may calculate a second eigenvalue by applying a weighted value to a first eigenvalue having a higher correlation with the biometric information to be estimated than other first eigenvalues ​​among the extracted first eigenvalues, as represented by the following Equation 4.

[0093] [Equation 4]

[0094] f sc =α×f 1a +β×f 1b – fsc,ref

[0095] α and β can be expressed as respectively applied to the first eigenvalue f 1a and the first eigenvalue f 1b The weight of the first eigenvalue f 1a The correlation between the biological information is greater than the first eigenvalue f 1b When the correlation between α and biological information is large, α may have a value greater than β.

[0096] Equations 1 to 4 may represent examples of calculating the second characteristic value, but the calculation of the second characteristic value is not limited thereto, and the second characteristic value for determining the proportional control ratio may be determined by various combinations of the first characteristic values.

[0097] When the second characteristic value is calculated, the processor 120 may calculate a proportional control ratio based on the second characteristic value, and may determine the proportional factor by adjusting a reference proportional factor based on the calculated proportional control ratio.

[0098] Based on the magnitude of the calculated second eigenvalue, the processor 120 may determine whether the change in the first eigenvalue belongs to a homeostasis maintenance region, a linear change region, or a nonlinear saturation region, and may calculate a proportional control ratio based on the determination.

[0099] For example, in the case where the calculated second eigenvalue belongs to the body balance maintenance area, the processor 120 can reduce the reference proportional factor by calculating the proportional control ratio to 1 or less, and can determine the reduced reference proportional factor as the proportional factor, thereby reducing the impact of the change in the first eigenvalue on the change in the biological information.

[0100] In another example, in the case where the calculated second eigenvalue belongs to the linear change region, the processor 120 may determine a predetermined reference scale factor as the scale factor by determining the scale control ratio to be 1. In this manner, by using the predetermined reference scale factor as the scale factor, the processor 120 may reflect the change in the first eigenvalue in the change in the biometric information according to a predetermined ratio.

[0101] In another example, when the second eigenvalue belongs to the nonlinear saturation region, the processor 120 can determine the proportional control ratio for adjusting the reference proportional factor by applying the second eigenvalue to a predetermined proportional control estimation model pre-generated by a nonlinear function or machine learning, so that changes in the first eigenvalue can be reflected in changes in the biological information.

[0102] As described above, the processor 120 may adjust the reference scale factor based on the calculated scale control ratio, and may estimate the bio-information by adaptively adjusting the scale factor for estimating the bio-information.

[0103] The processor 120 may apply the second eigenvalue to a proportional control ratio decision function ρ sc The proportional control ratio is calculated according to the size of the second eigenvalue.

[0104] Figure 5 is an explanatory diagram showing an example of calculating a proportional control ratio.

[0105] Reference Figure 5 , proportional control ratio decision function ρ sc (f sc ) may have a valley shape, wherein in the valley shape, the proportional control ratio is within the reference second characteristic value f sc,ref The point where the minimum value ρ min , and increases from the reference second eigenvalue as the second eigenvalue changes, and when the second eigenvalue falls outside the threshold range, the proportional control ratio is saturated to the predetermined proportional control ratio (that is, the proportional control ratio is determined to be the predetermined proportional control ratio).

[0106] Here, the threshold range may indicate the low point μ of the second eigenvalue low With high point μ high In other words, the threshold range may indicate the region where the first eigenvalue changes in the body balance maintenance region and the region where the proportional control ratio changes adaptively. That is, at the second eigenvalue f sc In the case of a change within a predetermined threshold range, the proportional control ratio increases in both directions starting from the reference second characteristic value.

[0107] As described above, since the proportional control ratio has a minimum value at a point referring to the second eigenvalue, a change in the biosignal feature in a stable state (such as a reference state) may have a small effect on the estimation of the bioinformation; and as the second eigenvalue deviates from the reference state, the proportional control ratio for estimating the bioinformation increases. Therefore, when the amplitude of the biosignal increases, the change in the biosignal feature may have a greater effect on the estimation of the bioinformation.

[0108] Then, as the second characteristic value continuously changes to fall outside the threshold range, the proportional control ratio is saturated to a predetermined proportional control ratio (eg, 1), so that the processor 120 may determine the predetermined reference proportional factor itself as the proportional factor.

[0109] Return to reference Figure 5 , the proportional control ratio decision function (a) and the proportional control ratio decision function (b) can be linearly changed within the threshold range starting from the reference second eigenvalue. However, as can be seen from the proportional control ratio decision function (b), the slope of the proportional control ratio decision function (a) can be changed according to the low point μ of the second eigenvalue. low and high point μ highTo be determined differently.

[0110] In addition, the proportional control ratio decision function (c) can be given in the form of an nth-order power function (e.g., a quadratic function, etc.) within a threshold range; the proportional control ratio decision function (d) can be given in the form of a trigonometric function (e.g., a cosine function, etc.) within a threshold range.

[0111] The shape of the curve within the threshold range of the proportional control ratio decision function may be changed according to the biosignal and the bioinformation to be estimated, and may be pre-generated based on an estimation model, wherein the estimation model is pre-generated by machine learning or based on the correlation between the biosignal and the bioinformation. In addition, the shape of the curve is not limited thereto, and the processor 120 may periodically acquire the biosignal of the user, and may directly generate the proportional control ratio decision function from the learning model used to generate the proportional control ratio decision function.

[0112] In addition, the processor 120 may calculate individual proportional control ratios for the first feature value, and may calculate the proportional control ratio based on statistical values ​​of the individual proportional control ratios.

[0113] Figure 6 is an explanatory diagram for explaining another example of calculating the proportional control ratio.

[0114] Reference Figure 1 and Figure 6 , the processor 120 may calculate the individual proportional control ratio for each of the first eigenvalues ​​based on the proportional control ratio decision function for each of the first eigenvalues.

[0115] For example, after extracting the first eigenvalue f 1a 、f 1b and f 1c When the second eigenvalue is not generated by combining the first eigenvalues, the processor 120 may use a proportional control ratio function ρ for each of the first eigenvalues. 1 (f 1a ),ρ 2 (f 1b ) and ρ 3 (f 1c ) to calculate the individual proportional control ratio ρ 1 , 2 and ρ 3 , and the statistical value of the individual proportional control ratio can be used as the proportional control ratio.

[0116] For example, the processor 120 may calculate an average of the individual proportional control ratios (e.g., ρ = (ρ 1 +ρ 2 +ρ 3) / 3) as the proportional control ratio; the processor 120 may apply a weighted value to a feature having a higher correlation with the biological information to be estimated than other feature values ​​in the extracted first feature value, and may calculate an average value of the feature (e.g., ρ = (α×ρ 1 +β×ρ 2 +γ×ρ 3 ) / 3) as the proportional control ratio. However, the proportional control ratio is not limited thereto, and the processor 120 may calculate statistical values ​​(such as a maximum value, a minimum value, and a median value) of individual proportional control ratios as the proportional control ratio.

[0117] Furthermore, in response to the first feature value exceeding a predetermined threshold, the processor 120 may determine a reference scale factor as the scale factor.

[0118] For example, the processor 120 may determine whether the first feature value exceeds a predetermined threshold.

[0119] In one embodiment, the processor 120 may compare the extracted first feature value with a predetermined low point μ of the proportional control ratio decision function. low (For example, Figure 6 μ 1a _ low , μ 1b _ low , μ 1c _ low ) or the predetermined high point μ high (For example, Figure 6 μ 1a _ high , μ 1b _ high , μ 1c _ high ) and determine whether the first eigenvalue is lower than the predetermined low point μ low Or determine whether the first characteristic value is higher than the predetermined high point μ high .

[0120] That is, the processor 120 may compare the first feature value with a predetermined threshold; in response to the first feature value exceeding the threshold (eg, below a predetermined low point μ low Or higher than the predetermined high point μ high ), the processor 120 may determine that a change in the first eigenvalue deviates from the body balance maintaining region and enters the linear change region, and may determine a reference proportional factor as the proportional factor.

[0121] In addition, in response to the first characteristic value not exceeding a predetermined threshold (eg, higher than a predetermined low point μ low And below the predetermined high point μ high), the processor 120 may determine that the change in the first eigenvalue belongs to the body balance maintaining region, and may calculate the proportional control ratio to adjust the reference proportional factor based on the calculated proportional control ratio.

[0122] As described above, the processor 120 may determine whether the first eigenvalue exceeds a predetermined threshold; in response to the first eigenvalue exceeding the threshold, the processor 120 may omit calculation of the proportional control ratio, thereby accurately and quickly estimating biometric information.

[0123] The processor 120 may calculate the third eigenvalue f by combining the first eigenvalues est , and the biological information can be estimated based on the calculated third eigenvalue and the determined scale factor.

[0124] For example, when calculating the third eigenvalue, the processor 120 may estimate the bio-information using a bio-information estimation model as shown in Equation 5 below.

[0125] [Equation 5]

[0126] BI est =SF×(f est -f est_ref )+BI offset

[0127] Here, BI est represents the biological information to be estimated, SF represents the scale factor, and f est_ref Indicates the reference third eigenvalue, BI offset Indicates the offset value for the biometric information to be estimated. offset may denote bio-information estimated in a reference state, and may be a reference value measured by an external device for estimating bio-information, which value may vary according to the type of bio-information to be estimated.

[0128] That is, when calculating the third eigenvalue, the processor 120 may estimate the biometric information by multiplying the difference between the third eigenvalue and the reference third eigenvalue by the determined scale factor and adding the offset value for the biometric information to the multiplied value.

[0129] In this case, the second eigenvalue and the third eigenvalue may be calculated by different combinations of the first eigenvalues.

[0130] For example, the second eigenvalue and the third eigenvalue are calculated by combining the first eigenvalues, but the methods of combining the first eigenvalues ​​to calculate the second eigenvalue and the third eigenvalue may be different from each other. However, the calculation is not limited to this, and the second eigenvalue and the third eigenvalue may be calculated by the same combination of the first eigenvalues.

[0131] As described above, the bio-information estimating apparatus 100 can stably estimate bio-information by adaptively changing the reference scale factor using an individual scale control ratio for each of a plurality of first eigenvalues ​​and / or a scale control ratio calculated based on a second eigenvalue.

[0132] Figure 7 is a block diagram illustrating another example of an apparatus for estimating biological information.

[0133] Reference Figure 7 , the biological information estimation device 700 includes a biological signal acquirer 710, a processor 720, an input unit 730, a memory 740, a communicator 750, and an output unit 760. Here, the biological signal acquirer 710 and the processor 720 basically perform the same as the reference Figure 1 The biosignal acquirer 110 and the processor 120 described herein have the same functions, and thus the following description will be made based on the non-overlapping portions.

[0134] The input unit 730 may receive input of various operation signals and data required for estimating biometric information from the user.

[0135] For example, the input part 730 may include a keyboard, a dome switch, a touch pad (static pressure / capacitance), a scroll wheel, a jog switch, a hardware (H / W) button, etc. In particular, a touch pad forming a layer structure with a display may be referred to as a touch screen.

[0136] For example, the input unit 730 may receive user characteristic information or measurement points of biological signals, types of biological signals and biological information, etc., wherein the user characteristic information includes one or more of the following items: age, gender, weight, body mass index (BMI), and medical history of the user.

[0137] The memory 740 may store a program or command for the operation of the bio-information estimation apparatus 700, and may store data input to and output from the bio-information estimation apparatus 700. For example, the memory 740 may store user information input through the input unit 730, bio-signal data acquired by the bio-signal acquirer 710, an extracted first eigenvalue, a calculated second eigenvalue, and a calculated third eigenvalue, a scale factor, a reference scale factor, a scale control ratio, an individual scale control ratio, a scale control ratio decision function, a reference first eigenvalue, a reference second eigenvalue, a reference third eigenvalue, and a bio-information estimation model.

[0138] The memory 740 may include at least one of the following storage media: a flash memory, a hard disk memory, a multimedia card micro memory, a card type memory (e.g., an SD memory, an XD memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, etc. In addition, the biometric information estimation apparatus 700 may operate an external storage medium (e.g., a web storage, etc.) that performs a storage function of the memory 740 on the Internet.

[0139] The communicator 750 can perform communication with an external device. For example, the communicator 750 can send user feature information input through the input unit 730, the bio-signal acquired through the bio-signal acquirer 710, the estimation result of the bio-information of the processor 720, etc. to the external device; or can receive various data (such as user feature information, bio-signal, proportional control ratio decision function, bio-information estimation model, etc.) from the external device.

[0140] In this case, the external device may be a medical device that uses a bio-information database (DB) and / or an estimation result of the bio-information, a printer for printing out the result, or a display for displaying the estimation result of the bio-information. In addition, the external device may be a digital TV, a desktop computer, a cellular phone, a smart phone, a tablet PC, a laptop computer, a personal digital assistant (PDA), a portable multimedia player (PMP), a navigator, an MP3 player, a digital camera, a wearable device, etc., but the external device is not limited thereto.

[0141] The communicator 750 may communicate with an external device by using Bluetooth communication, Bluetooth low energy (BLE) communication, near field communication (NFC), WLAN communication, Zigbee communication, infrared data association (IrDA) communication, Wi-Fi direct (WFD) communication, ultra-wideband (UWB) communication, Ant+ communication, WIFI communication, radio frequency identification (RFID) communication, 3G communication, 4G communication, 5G communication, etc. However, this is exemplary only and is not intended to be limiting.

[0142] The output part 760 may output at least one of the following items: a biosignal, a feature value of the biosignal, a multiplication coefficient ratio control factor, a multiplication coefficient ratio, and estimated bioinformation.

[0143] For example, the output unit 760 may output at least one or more of the following items: an estimation result of the biological information, warning information about the state of the acquired biological signal, reliability of the biological information estimated by using at least one of an acoustic method, a visual method, and a tactile method. To this end, the output unit 760 may include a display, a speaker, a vibrator, etc.

[0144] For example, the processor 720 may measure the biosignal in a reference state, and may output biosignal measurement guide information for calculating a reference first eigenvalue, a reference second eigenvalue, a reference third eigenvalue, and an offset value of the bioinformation.

[0145] In addition, the processor 720 may receive a new biosignal from an external biosignal database (DB) through the communicator 750 .

[0146] Figure 8 is a flowchart illustrating an example of a method of estimating biological information. Figure 8 The bioinformatics estimation method can be Figure 1 The biological information estimation device 100 shown in Figure 7 The biological information estimation device 700 shown in FIG.

[0147] In 810 , the bio-information estimating apparatus 700 may acquire a bio-signal.

[0148] The bio-information estimation device 700 may include a sensor including at least one of the following items: one or more electrodes for measuring bio-signals, a pressure sensor, and a photodetector module including a light source and a detector. The bio-information estimation device 700 may be directly interfaced with a user to obtain the bio-signal. In addition, the bio-information estimation device 700 is not limited thereto and may communicate with an external device to receive the user's bio-signal data from the external device.

[0149] In 820, when the biological signal is acquired, the biological information estimation apparatus 700 may extract one or more first feature values ​​f from the acquired biological signal. 1 .

[0150] The first feature value extracted from the bio-signal may indicate a feature having a predetermined correlation with the bio-information desired to be estimated, and may vary according to the type of the bio-information.

[0151] When extracting the first feature value from the biosignal, the bioinformation estimation apparatus 700 may convert the first feature value. For example, the bioinformation estimation apparatus 700 may normalize the first feature value by dividing the extracted first feature value by a reference first feature value extracted in a reference state.

[0152] In this case, the reference state is a state other than a motion state or a sleep state, and may represent, for example, a state in which the pulse and respiratory rate are stable, or a state in which the blood pressure measured by an external device for measuring blood pressure is maintained without substantial change.

[0153] In addition, a reference second eigenvalue and a reference third eigenvalue to be described later may mean a second eigenvalue and a third eigenvalue calculated based on a reference first eigenvalue extracted in a reference state.

[0154] Then, in 830, the bio-information estimation apparatus 700 may determine a scale factor based on the extracted first eigenvalue. Operation 830 may include a first step of calculating a scale control ratio and a second step of multiplying a default scale factor by the scale control ratio. For example, the bio-information estimation apparatus 700 may store the default scale factor in a memory, and may adjust the default scale factor by multiplying the default scale factor by the scale control ratio. The term "default scale factor" may also be referred to as a "reference scale factor."

[0155] The bio-information estimating apparatus 700 may calculate the second eigenvalue f by combining the first eigenvalues. sc , and the proportional control ratio can be calculated based on the second eigenvalue.

[0156] For example, when calculating one or more first feature values, the bio-information estimating apparatus 700 may calculate a second feature value by using individual variations and / or combined variations of the first feature values.

[0157] In this case, the combined variation can be represented by the first eigenvalue f 1a and the first eigenvalue f 1b A linear combination of 1a and f 1b The value obtained by adding, subtracting, multiplying, or a combination of adding, subtracting, and multiplying (e.g., the second eigenvalue f sc, ref The difference between the individual changes can be expressed as the first eigenvalue f 1a and the first eigenvalue f 1b Each of the reference second eigenvalues ​​f sc, ref The difference between.

[0158] The bio-information estimating apparatus 700 may apply a weighted value to a first eigenvalue having a higher correlation with the bio-information to be estimated than other first eigenvalues ​​among the extracted first eigenvalues.

[0159] As described above, when the second eigenvalue is calculated, the bio-information estimating apparatus 700 may calculate a scale control ratio based on the second eigenvalue, and may determine the scale factor by adjusting a reference scale factor based on the calculated scale control ratio.

[0160] For example, the bio-information estimating apparatus 700 may calculate the proportional control ratio according to the size of the second eigenvalue by applying the calculated second eigenvalue to the proportional control ratio decision function.

[0161] Here, the proportional control ratio decision function may have a valley shape, wherein in the valley shape, the proportional control ratio is sc,ref The point where the minimum value ρ min , and increases as the second eigenvalue changes from the reference second eigenvalue, and in the case where the second eigenvalue falls outside the threshold range, the proportional control ratio is saturated to a predetermined proportional control ratio.

[0162] Here, the threshold range may indicate the low point μ of the second eigenvalue low With high point μ high In other words, the threshold range may indicate the region where the first eigenvalue changes in the body balance maintenance region and the region where the proportional control ratio changes adaptively. That is, at the second eigenvalue f sc In the case of a change within a predetermined threshold range, the proportional control ratio increases in both directions starting from the reference second characteristic value.

[0163] As described above, since the proportional control ratio has a minimum value at a point referring to the second eigenvalue, a change in the biosignal feature in a stable state (such as a reference state) may have a small effect on the estimation of the bioinformation; and as the second eigenvalue deviates from the reference state, the proportional control ratio for estimating the bioinformation increases. Therefore, as the amplitude of the biosignal increases, a change in the biosignal feature may have a greater effect on the estimation of the bioinformation.

[0164] Then, as the second feature value continuously changes to fall outside the threshold range, the scale control ratio is saturated to a predetermined scale control ratio (eg, 1), so that the bio-information estimation apparatus 700 can determine the predetermined reference scale factor itself as the scale factor.

[0165] In another example, the bio-information estimating apparatus 700 may calculate the individual proportional control ratio for each of the first eigenvalues ​​based on the proportional control ratio decision function for each of the first eigenvalues.

[0166] For example, after extracting the first eigenvalue f 1a 、f 1b and f 1c When the second eigenvalue is not generated by combining the first eigenvalues, the biological information estimation device 700 may use a proportional control ratio decision function ρ for each first eigenvalue 1 (f 1a ),ρ 2 (f1b ) and ρ 3 (f 1c ) to calculate the individual proportional control ratio ρ 1 , 2 and ρ 3 , and the statistical value of the individual proportional control ratio can be used as the proportional control ratio.

[0167] As described above, when the scale control ratio is calculated, the bio-information estimating apparatus 700 may determine the scale factor by adjusting the reference scale factor using the calculated scale control ratio.

[0168] Then, in operation 840, when the scale factor is determined, the bio-information estimating apparatus 700 may estimate the bio-information based on the determined scale factor and the first eigenvalue. Operation 840 may include extracting the feature f from the extracted bio-signal. est Subtract the reference feature f est_ref A first step of obtaining a first value, a second step of multiplying the first value by the adjusted scale factor SF to obtain a second value, and a second step of multiplying the offset value BI offset Add to the second value to get the estimated result BI est The third step.

[0169] For example, the bio-information estimating apparatus 700 may calculate the third eigenvalue f by combining the first eigenvalues. est , and the biological information can be estimated based on the calculated third eigenvalue and the determined scale factor.

[0170] When the third eigenvalue is calculated, the biometric information estimating apparatus 700 may estimate the biometric information by multiplying a difference between the third eigenvalue and the reference third eigenvalue by a determined scale factor and adding an offset value for the biometric information to the multiplied value.

[0171] Fig. 9 is a flowchart illustrating another example of a method of estimating biological information.

[0172] The acquiring of biological signals in 910, the extracting of one or more first eigenvalues ​​from the acquired biological signals in 920, the determining of scaling factors based on the extracted first eigenvalues ​​in 950, and the estimating of biological information based on the determined scaling factors and the first eigenvalues ​​in 960 are respectively substantially the same as the acquiring of biological signals in 810, the extracting of one or more first eigenvalues ​​from the acquired biological signals in 820, the determining of scaling factors based on the extracted first eigenvalues ​​in 830, and the estimating of biological information based on the determined scaling factors and the first eigenvalues ​​in 840, so that the following description will be based on the non-overlapping parts.

[0173] In 910 , the bio-information estimating apparatus 700 may acquire a bio-signal.

[0174] In 920 , when the biosignal is acquired, the bioinformation estimating apparatus 700 may extract one or more first feature values ​​from the acquired biosignal.

[0175] Then, in 930 , the bio-information estimating apparatus 700 may determine whether the first feature value exceeds a predetermined threshold.

[0176] For example, the biological information estimation apparatus 700 may compare the extracted first feature value with a predetermined low point μ of the proportional control ratio decision function. low The value at or the predetermined high point μ high The first characteristic value is compared with the value at the predetermined low point μ low Or it can be determined whether the first characteristic value is higher than a predetermined high point μ high .

[0177] In 940 , upon determining that the first feature value exceeds the threshold, the bio-information estimating apparatus 700 may determine a reference scale factor as the scale factor.

[0178] For example, the bioinformation estimation device 700 may compare the first eigenvalue with a predetermined threshold; in response to the first eigenvalue exceeding the threshold based on the comparison, the bioinformation estimation device 700 may determine that the change in the first eigenvalue deviates from the body balance maintenance area and enters the linear change area, and may determine the reference proportional factor as the proportional factor.

[0179] As described above, the bio-information estimating apparatus 700 may determine whether the first eigenvalue exceeds a predetermined threshold; in response to the first eigenvalue exceeding the threshold, the bio-information estimating apparatus 700 may omit calculation of the proportional control ratio, thereby accurately and quickly estimating bio-information.

[0180] Furthermore, in 950 , in response to the first eigenvalue not exceeding the predetermined threshold, the bio-information estimating apparatus 700 may determine that the change in the bio-signal belongs to a homeostasis maintaining region, and may determine a scale factor based on the extracted first eigenvalue.

[0181] In 960 , when the scale factor is determined, the bio-information estimating apparatus 700 may estimate the bio-information based on the determined scale factor and the first eigenvalue.

[0182] The present disclosure may be implemented as a computer-readable code written on a computer-readable recording medium. Ordinary computer programmers in the art may easily derive the code and code segments required to implement the present disclosure. The computer-readable recording medium may be any type of recording device that stores data in a computer-readable manner. The computer-readable recording medium may be any type of recording device that stores data in a computer-readable manner. Examples of computer-readable recording media include: ROM, RAM, CD-ROM, magnetic disk, floppy disk, optical disk, etc. In addition, the computer-readable recording medium may be distributed on a plurality of networked computer systems so that the computer-readable recording medium is written therein and executed therefrom in a distributed manner.

[0183] The present invention has been described herein with respect to preferred embodiments. However, it will be apparent to those skilled in the art that various modifications may be made without departing from the gist of the present invention. Therefore, it will be understood that the scope of the present invention is not limited to the above-mentioned embodiments, but is intended to include various modifications and equivalents included within the spirit and scope of the appended claims.

Claims

1. A device for estimating biological information, the device comprising: A biological signal acquirer configured to acquire a biological signal; a processor configured to extract a plurality of first eigenvalues ​​from the acquired biological signal, determine a scale factor based on the extracted plurality of first eigenvalues, and estimate biological information based on the determined scale factor and the plurality of first eigenvalues, The processor is further configured to: calculate a second eigenvalue by combining the plurality of first eigenvalues, and determine the proportional factor by adjusting the reference proportional factor using a proportional control ratio calculated based on the second eigenvalue, The plurality of first characteristic values ​​include: a characteristic associated with cardiac output, a characteristic associated with total peripheral resistance, and a combination of a characteristic associated with cardiac output and a characteristic associated with total peripheral resistance. The processor is further configured to: calculate the second eigenvalue by at least one of the individual changes and the combined changes of the plurality of first eigenvalues, wherein the combined variation represents the difference between a value obtained by a linear combination of the plurality of first eigenvalues ​​and a reference second eigenvalue, wherein the individual variation represents the difference between each of the plurality of first eigenvalues ​​and a reference second eigenvalue, Herein, the reference second eigenvalue refers to a second eigenvalue calculated based on a reference first eigenvalue extracted under a reference state.

2. The device according to claim 1, wherein: The processor is further configured to calculate the proportional control ratio according to the magnitude of the second eigenvalue by applying the calculated second eigenvalue to the proportional control ratio decision function.

3. The device according to claim 2, wherein: The proportional control ratio decision function has a valley shape, wherein in the valley shape, the proportional control ratio has a minimum value at a point referenced to the second eigenvalue, and increases from the referenced second eigenvalue as the second eigenvalue changes, and when the second eigenvalue falls outside a threshold range, the proportional control ratio is saturated to a predetermined proportional control ratio.

4. The device according to claim 1, wherein: The processor is further configured to calculate a plurality of individual proportional control ratios for the plurality of first feature values, and determine the proportional factor by adjusting a reference proportional factor using a proportional control ratio calculated based on statistical values ​​of the plurality of individual proportional control ratios.

5. The device according to claim 1, wherein: The processor is further configured to calculate a third eigenvalue by combining the plurality of first eigenvalues, and estimate biometric information based on the calculated third eigenvalue and the determined scale factor.

6. The device according to claim 5, wherein: The processor is further configured to estimate the biometric information by multiplying a difference between the third eigenvalue and the reference third eigenvalue by a determined scale factor and adding an offset value for the biometric information to the multiplied value. Herein, the reference third eigenvalue refers to a third eigenvalue calculated based on the reference first eigenvalue extracted under the reference state.

7. The device according to claim 1, wherein: The processor is further configured to: in response to the plurality of first feature values ​​exceeding a predetermined threshold, determine a reference scale factor as the scale factor.

8. The device according to claim 1, wherein: The processor is further configured to normalize the plurality of first eigenvalues ​​based on a reference first eigenvalue.

9. The apparatus according to claim 1, further comprising: The output unit is configured to output at least one of a biological signal, a characteristic value of the biological signal, a multiplication coefficient ratio control factor, a multiplication coefficient ratio, and biological information.

10. The device according to claim 1, wherein: The biological information includes at least one of blood pressure, pulse, cardiac output, blood sugar, triglyceride and keratin.

11. A method for estimating biological information, the method comprising: Acquiring biological signals; extracting a plurality of first feature values ​​from the acquired biological signal; determining a scaling factor based on the extracted plurality of first eigenvalues; estimating biological information based on the determined scale factor and the plurality of first eigenvalues, The step of determining the proportional factor includes: calculating a second eigenvalue by combining the plurality of first eigenvalues; calculating a proportional control ratio based on the second eigenvalue; and determining the proportional factor by adjusting a reference proportional factor based on the calculated proportional control ratio. The plurality of first characteristic values ​​include: a characteristic associated with cardiac output, a characteristic associated with total peripheral resistance, and a combination of a characteristic associated with cardiac output and a characteristic associated with total peripheral resistance. The step of calculating the second eigenvalue comprises: calculating the second eigenvalue by at least one of individual changes and combined changes of the plurality of first eigenvalues, wherein the combined variation represents the difference between a value obtained by a linear combination of the plurality of first eigenvalues ​​and a reference second eigenvalue, wherein the individual variation represents the difference between each of the plurality of first eigenvalues ​​and a reference second eigenvalue, Herein, the reference second eigenvalue refers to a second eigenvalue calculated based on a reference first eigenvalue extracted under a reference state.

12. The method according to claim 11, wherein: The step of calculating the proportional control ratio includes calculating the proportional control ratio according to the size of the second eigenvalue by applying the calculated second eigenvalue to a proportional control ratio decision function.

13. The method according to claim 12, wherein: The proportional control ratio decision function has a valley shape, wherein in the valley shape, the proportional control ratio has a minimum value at a point referenced to the second eigenvalue, and increases from the referenced second eigenvalue as the second eigenvalue changes, and when the second eigenvalue falls outside a threshold range, the proportional control ratio is saturated to a predetermined proportional control ratio.

14. The method according to claim 11, wherein: The steps to determine the scale factor are: calculating a plurality of individual proportional control ratios for the plurality of first eigenvalues; calculating a proportional control ratio based on the statistical values ​​of the plurality of individual proportional control ratios; The proportional factor is determined by adjusting a reference proportional factor based on the calculated proportional control ratio.

15. The method according to claim 11, wherein: The steps to estimate bioinformatics include: calculating a third eigenvalue by combining the plurality of first eigenvalues; estimating the biometric information by multiplying the difference between the calculated third eigenvalue and the reference third eigenvalue by the determined scale factor and adding the offset value for the biometric information to the multiplied value, Herein, the reference third eigenvalue refers to a third eigenvalue calculated based on the reference first eigenvalue extracted under the reference state.

16. The method according to claim 11, wherein: The steps to determine the scale factor are: determining whether the plurality of first characteristic values ​​exceeds a predetermined threshold; When it is determined that the plurality of first characteristic values ​​exceed a predetermined threshold, a reference scale factor is determined as the scale factor.

Images