Method and device for extracting characteristics of biological signals and device for detecting biological information

Abstract

A method and device for extracting features of a biological signal and a device for detecting biological information are disclosed. A device configured to extract features from a biological signal is disclosed. The device includes: a biological signal acquirer configured to acquire a biological signal; a processor configured to analyze the concavity and convexity of a waveform of the biological signal in an interval of the biological signal, and extract features from the biological signal based on the analyzed concavity and the analyzed convexity.

Description

[0001] This application claims the priority benefit of Korean Patent Application No. 10-2016-0092166 filed on Jul. 20, 2016, in the Korean Intellectual Property Office, the disclosure of which is incorporated herein by reference in its entirety. Technical Field

[0002] Methods and apparatuses disclosed herein consistent with exemplary embodiments relate to a method and apparatus for extracting features from a biosignal to detect bio-information, and an apparatus for detecting bio-information using the method of extracting features. Background Art

[0003] In recent years, due to the soaring medical expenses, the aging population structure and the shortage of human resources for specialized medical services, research on IT-medical fusion technology that combines IT (information technology) with medical technology is being carried out. In particular, the monitoring of the health status of the human body is not limited to being performed in some fixed places (such as hospitals), but extends to mobile medical places that monitor the health status of users anywhere and at any time in daily life, such as at home or in the office. As representative examples of biological signals that represent the health status of an individual, there are ECG (electrocardiogram) signals, PPG (photoplethysmography) signals, and EMG (electromyography) signals, and various biological signal sensors have been developed for measuring these signals in daily life. In particular, in the case of a PPG sensor, the blood pressure of the human body can be estimated by analyzing the pulse wave that reflects the cardiovascular condition.

[0004] According to the research results related to PPG biosignals, the overall PPG signal is a superposition of the propagation wave from the heart to the distal part of the body and the reflected wave returning from the distal part. In addition, it is known that information that can estimate blood pressure can be obtained by extracting various features related to the propagation wave or the reflected wave. Summary of the invention

[0005] This Summary is provided to introduce a selection of concepts in a simplified form that are further described below in the Detailed Description. This Summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used as an aid in determining the scope of the claimed subject matter.

[0006] According to aspects of an exemplary embodiment, there is provided a device configured to extract features from a biological signal, the device comprising: a biological signal acquirer configured to acquire a biological signal; a processor configured to analyze concavity and convexity of a waveform of the biological signal in an interval of the biological signal, and extract features from the biological signal based on the analyzed concavity and the analyzed convexity.

[0007] The processor may include: a convex-concave difference (CCD) calculator configured to calculate a CCD value based on concavity and convexity of a waveform of the biosignal in the interval; and a feature extractor configured to extract a feature from the biosignal based on the calculated CCD value.

[0008] The CCD calculator may be configured to calculate a difference in upward convexity between a specific time point in one interval and a previous time point or a difference in downward convexity between a specific time point in the one interval and a next time point as the CCD value of the one interval.

[0009] The feature extractor may be configured to select a preset number of intervals in descending order of the calculated CCD values ​​and extract features from each selected interval.

[0010] The processor may further include: an interval divider configured to divide the biosignal into a preset number of intervals.

[0011] The interval divider may be configured to determine a preset number of intervals to be divided in consideration of at least one of complexity of the device, performance of the device, and accuracy of bio-information included in the bio-signal.

[0012] The interval divider may be configured to adaptively adjust the number of intervals to be divided according to the type of waveform of the biosignal.

[0013] The biosignal may include at least one of an ECG (electrocardiogram) signal, a PPG (photoplethysmography) signal, and an EMG (electromyogram) signal.

[0014] According to an aspect of another exemplary embodiment, a method for extracting features from a biological signal is provided, the method comprising: acquiring a biological signal; analyzing concavity and convexity of a waveform of the biological signal in an interval of the biological signal; and extracting features from the biological signal based on the analyzed concavity and the analyzed convexity.

[0015] The analyzing step may include calculating a concave-convex difference (CCD) value based on concavity and convexity of a waveform of the biosignal in the interval, and the extracting step may include extracting a characteristic point from the biosignal based on the calculated CCD value.

[0016] The extracting step may include selecting a preset number of intervals in descending order of the calculated CCD values, and extracting features from the selected intervals.

[0017] The method may further include segmenting the biosignal into a preset number of intervals.

[0018] The segmenting may include adaptively adjusting the number of intervals to be segmented according to the type of waveform of the biosignal.

[0019] According to an aspect of another exemplary embodiment, a device configured to extract features from a biological signal is provided, the device comprising: a biological signal acquirer configured to acquire a biological signal; a processor configured to derive a second-order differential signal of the biological signal, analyze concavity and convexity of a waveform of the second-order differential signal, and extract features from the biological signal based on the analyzed concavity and the analyzed convexity.

[0020] The processor may include: a convex-concave difference (CCD) calculator configured to calculate a difference between a local maximum value and a local minimum value in the interval as a CCD value; and a feature extractor configured to extract a feature from the biosignal based on the calculated CCD value.

[0021] The feature extractor may be configured to select a preset number of intervals in descending order of the calculated CCD values, and extract features based on local minimum points of the selected intervals.

[0022] The feature may include at least one of time information of the local minimum point and amplitude information of the biosignal position corresponding to the time information.

[0023] The processor may further include: an interval divider configured to divide the derived second-order differential signal into a preset number of intervals.

[0024] The interval divider may be configured to divide a portion of the waveform from a local maximum point to a next local minimum point of the second-order differential signal into one interval.

[0025] The interval partitioner may be configured to merge the partitioned two or more intervals based on a difference between a local minimum value and a next local maximum value in the second-order differential signal.

[0026] According to an aspect of another exemplary embodiment, a method for extracting features from a biological signal is provided, the method comprising: acquiring a biological signal; deriving a second-order differential signal of the biological signal; analyzing concavity and convexity of a waveform of the second-order differential signal; and extracting features from the biological signal based on the analyzed concavity and the analyzed convexity.

[0027] The analyzing step may include calculating a difference between a local maximum value and a local minimum value in the interval as a convex-concave difference (CCD) value, and the extracting step may include extracting a feature from the biosignal based on the calculated CCD value.

[0028] The step of extracting the feature may include selecting a preset number of intervals in descending order of the calculated CCD values, and extracting the feature based on a local minimum point of each selected interval.

[0029] The method may further include dividing the derived second order differential signal into a predetermined number of intervals.

[0030] The step of segmenting may include segmenting an interval from a local maximum point to a next local minimum point of the second-order differential signal into a single interval.

[0031] The method may further include merging the divided two or more intervals based on a difference between a local minimum value and a next local maximum value in the second-order differential signal.

[0032] According to an aspect of another exemplary embodiment, there is provided a device configured to detect biological information, the device comprising: a sensor configured to illuminate an object with light and detect light reflected by the object to detect a biological signal; and a processor configured to: when the biological signal is measured, extract features from the biological signal based on concavity and convexity of a waveform of the biological signal in an interval of the biological signal, and detect the biological information based on the extracted features.

[0033] In response to measuring the biosignal, the processor may be configured to derive a second order differential signal of the measured biosignal and divide the derived second order differential signal into intervals based on local maximum points and local minimum points of the derived second order differential signal.

[0034] The processor may be configured to select a predetermined number of intervals from the segmented intervals, extract time information of local minimum points of the selected intervals and amplitude information of the biological signal position corresponding to the time information, and detect biological information by combining at least two extracted amplitude information.

[0035] The processor may be configured to monitor the health status of the user based on the biometric information and generate additional information according to the health status.

[0036] The biological information may include blood pressure information, and the additional information may include warning information.

[0037] The apparatus may further include: a display configured to display the detected biological information and additional information.

[0038] According to another exemplary embodiment, a medical imaging device is provided, including: a biological signal acquirer configured to acquire a biological signal; a processor configured to: divide a waveform of the biological signal into multiple intervals, eliminate one of the multiple intervals based on the waveform, and calculate a feature of the biological signal based on the remaining intervals of the multiple intervals except the eliminated interval.

[0039] Other features and aspects will be apparent from the following detailed description, drawings, and claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 is a block diagram illustrating an apparatus for extracting features of a biosignal according to an exemplary embodiment;

[0041] Figure 2A , Figure 2B and Figure 2C is a diagram describing a general method for extracting features from a biological signal;

[0042] Figure 3 It shows that according to Figure 1 A block diagram of an exemplary embodiment of a processing unit of an apparatus for extracting features of a biosignal of an exemplary embodiment of the present invention;

[0043] Figure 4 is used to describe the Figure 3 A diagram of a method for analyzing a waveform of a biological signal performed by a processing unit of an exemplary embodiment of the present invention;

[0044] Figure 5 It shows that according to Figure 1 A block diagram of another exemplary embodiment of a processing unit of an apparatus for extracting features of a biosignal of an exemplary embodiment of the present invention;

[0045] Fig. 6A , Figure 6B , Figure 6C , Fig.6D and Fig. 6E is used to describe the Figure 5 A diagram of a method for analyzing a waveform of a biological signal performed by a processing unit of an exemplary embodiment of the present invention;

[0046] Fig. 7A and Figure 7B is used to describe the Figure 5 1. A diagram of a method for segmenting an interval of a waveform of a biological signal performed by a processing unit of an exemplary embodiment of the present invention;

[0047] Figure 8 is a flowchart illustrating a method for extracting features of a biosignal according to an exemplary embodiment;

[0048] Fig. 9 is a flowchart illustrating a method for extracting features of a biosignal according to another exemplary embodiment;

[0049] Fig.10 is a block diagram illustrating an apparatus for detecting biological information according to an exemplary embodiment;

[0050] Fig.11 is a detailed block diagram illustrating a processing unit of an apparatus for detecting biological information according to an exemplary embodiment;

[0051] Fig.12 is a flowchart illustrating a method for detecting bio-information according to an exemplary embodiment.

[0052] Throughout the drawings and detailed description, unless otherwise described, the same drawing reference numerals should 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

[0053] In addition to the accompanying drawings, the advantages and features of the present technology disclosure and the methods for achieving them should become clear with reference to the exemplary embodiments described in detail below. However, the scope of the present disclosure is not limited to the exemplary embodiments to be described below, but can be implemented in various forms. Throughout the specification, the same elements represent the same reference numerals.

[0054] 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. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, unless the context clearly indicates otherwise, when a unit "includes" an element, the unit does not exclude another element, but may also include another element. In addition, the terms such as "unit" and "module" used here indicate a unit for performing at least one function or operation, and may be implemented by hardware or software or a combination thereof.

[0055] Hereinafter, exemplary embodiments of a method and apparatus for extracting features of a biological signal and an apparatus for detecting biological information will be described in detail with reference to the accompanying drawings. For the convenience of description, terms or words such as concave, convex, "convex downward", "convex upward", "concave downward", "concave upward", etc. may be used in various ways and in various combinations. However, such terms and words may be defined for the case where the waveform of the biological signal is convex downward. Therefore, convex may be described as the characteristic that the waveform is convex downward, and concave may be described as the characteristic that the waveform is convex upward. In addition, "convex downward" may have the same meaning as a waveform that is concave upward, and "convex upward" may have the same meaning as a waveform that is concave downward.

[0056] Figure 1is a block diagram showing an apparatus for extracting features of a biosignal according to an exemplary embodiment. The apparatus 100 for extracting features of a biosignal may be installed in a terminal (such as a wearable device, a smart phone, a tablet PC, a desktop PC, a notebook PC, etc.) or a medical device (such as a blood pressure measuring device, etc.) in the form of a hardware or software module. Optionally, the apparatus 100 for extracting features of a biosignal may be implemented as an independent hardware device, in which case it may be used for the purpose of research on analyzing the waveform of a biosignal composed of the sum of a plurality of constituent pulses. However, the apparatus 100 for extracting features of a biosignal is not limited thereto, and may be variously modified and implemented according to the purpose of use.

[0057] Reference Figure 1 The device 100 for extracting features of a biological signal includes a biological signal acquisition unit 110 (e.g., a biological signal acquirer) and a processing unit 120 (e.g., a processor). The biological signal acquisition unit 110 and the processing unit 120 may be implemented by one or more circuits, processors, memories, or a combination thereof.

[0058] The biosignal acquisition unit 110 may acquire a biosignal of the object to be used to extract bio-information of the object, and transmit the acquired biosignal to the processing unit 120. The bio-information may be blood pressure, or the bio-signal may be an ECG (electrocardiogram) signal, a PPG (photoplethysmography) signal, an EMG (electromyogram) signal, or the like.

[0059] Hereinafter, for convenience of description, it will be described assuming that a PPG signal (hereinafter, referred to as a "pulse wave signal") is a biosignal and blood pressure is bioinformation. However, the biosignal and bioinformation are not construed as being limited to the PPG signal and blood pressure.

[0060] As an example, the bio-signal acquisition unit 110 may control the sensor according to a predetermined control signal and receive the pulse wave signal of the object from the corresponding sensor. The control signal may be generated by a control module implemented in the device 100 for extracting features. The control module may be implemented as a function of the processing unit 120 and generate a control signal based on the user's input. The sensor may drive the light source to illuminate the object with light upon receiving the control signal, and acquire the pulse wave signal by detecting the light that has been scattered or reflected from the object and then returned. When the pulse wave signal is received, the sensor may send the acquired pulse wave signal to the bio-signal acquisition unit 110.

[0061] The sensor may be installed in the apparatus 100 for extracting features or in other separate hardware devices. When the sensor is installed in different hardware devices, the sensor and the apparatus 100 for extracting features may be equipped with a communication module capable of wired and wireless communication, and connected to each other through the communication module, thereby sending and receiving control signals and pulse wave signals, etc.

[0062] As another example, the biosignal acquisition unit 110 may receive a biosignal, for example, from another external device. The other external device may be a device installed in a medical institution or the like to measure various biosignals for detecting bioinformation and store the measured biosignals. When it is necessary or desired to extract features from the measured biosignals, the other external device may send the biosignal to the biosignal acquisition unit 110.

[0063] The processing unit 120 receives the biosignal from the biosignal acquisition unit 110. The processing unit 120 may perform preprocessing, such as filtering for removing noise from the received biosignal. For example, the processing unit 120 may extract features by analyzing the waveform of the biosignal in each fixed interval unit. To this end, the processing unit 120 may segment the waveform of the biosignal into predetermined intervals, and extract features considering the concavity and convexity of each segmented interval. According to an exemplary embodiment, the feature may be, but is not limited to, information such as time, amplitude, or a combination thereof. The feature may include various information according to the type and characteristics of the biosignal.

[0064] Below, we will refer to Figure 2A Various exemplary embodiments of the processing unit 120 for extracting features from a biosignal are described through FIG. 7 .

[0065] Figure 2A , Figure 2B and Figure 2C is a diagram describing a general method for extracting features from a biological signal.

[0066] Figure 2A The pulse wave signal (solid line) and the component pulses (dashed lines) constituting the pulse wave signal are shown. Figure 2A As shown, it can be seen that a single waveform of the pulse wave signal is formed in a superposition manner of five component pulses. According to an exemplary embodiment, by appropriately combining information such as time, amplitude, etc. corresponding to the component pulses, a feature having a high correlation with blood pressure can be extracted. Usually, pulses up to the third pulse are mainly used as component pulses for estimating blood pressure. Usually, it is difficult to find the pulse after the third pulse due to noise and the pulse after the third pulse having a low correlation with the estimation of blood pressure. In addition, for some people, the pulse after the third pulse cannot even be observed.

[0067] Figure 2BThe general process of extracting features using a biological signal is shown. Generally, an apparatus for extracting features from a biological signal may search for a local minimum point in a second-order differential signal of the biological signal. Next, the apparatus may obtain time information T corresponding to the local minimum point. 1 , time information T 2 and time information T 3 And the biological signal and time information T 1 , time information T 2 and time information T 3 The corresponding amplitude information P 1 , amplitude information P 2 and amplitude information P 3 As characteristic points, and using the acquired characteristic points to extract features for detecting biological information. According to an exemplary embodiment, the local minimum point represents a point in which the corresponding signal for a specific point is reduced and then increased when some intervals of the second-order differential signal are observed. The second-order differential signal can be referred to as a downward convex form.

[0068] Figure 2C The pulse wave signal ( Figure 2C The above curve in the figure) and the signal derived from the second-order differential pulse wave signal ( Figure 2C According to an exemplary embodiment, a pulse wave signal is derived using a sampling frequency of 250 Hz, and the horizontal axis corresponds to a sample index that increases proportionally with time.

[0069] Assuming that the number of characteristic points for extracting features is 3, when the general method for extracting features is applied to such a pulse wave signal, three time positions T corresponding to the local minimum points of the second-order differential signal are derived. 1 、T 2 and T 3 and the time position T 1 、T 2 and T 3 The corresponding pulse wave signal amplitude position P 1 , P 2 and P 3 However, the amplitude position that should actually be obtained from the pulse wave signal is P 1 , P 3 and P 4 The incorrect amplitude position is obtained by the very slightly convex shape due to the waveform at position P. 2 The undesirable change of position P 2 It is generated by the local minimum point in the waveform of the second-order differential signal.

[0070] When the biosignal is generally classified into a front part corresponding to the systolic period of the heart and a rear part corresponding to the diastolic period, the third amplitude position generally corresponds to the top of the upward convex shape in the diastolic period. In addition, there is a downward convex point DN physiologically called a dicrotic notch immediately before the third amplitude position. Figure 2C In the example, the amplitude position P of the top of the upward convex shape at the next time point corresponding to the time point DN where the dicrotic notch appears is 4 The actual position should be the third amplitude position. In addition, the actual position of the second amplitude should be the P before the dicrotic notch. 3 .

[0071] Figure 3 It shows that according to Figure 1 A block diagram of an exemplary embodiment of a processing unit 120 of an exemplary embodiment of an apparatus 100 for extracting a feature of a biosignal, Figure 4 is used to describe the Figure 3 2 is a diagram of a method for analyzing a waveform of a biosignal performed by the processing unit 300 of an exemplary embodiment.

[0072] Reference Figure 3 , the processing unit 300 may include an interval segmentation unit 310 (eg, an interval segmenter), a CCD (convex-concave difference) calculation unit 320 (eg, a CCD calculator), and a feature extraction unit 330 (eg, a feature extractor).

[0073] The interval segmentation unit 310 segments the waveform of the acquired biosignal into predetermined intervals. According to an exemplary embodiment, the interval segmentation unit 310 may segment the waveform of the biosignal to obtain a predetermined number of characteristic points. When blood pressure is measured from a pulse wave signal, the predetermined number may be set to 3. However, the predetermined number is not limited thereto, and the predetermined number may be adjustable according to the types of biosignals and bioinformation to be detected, the characteristics of the waveform of the biosignal, the function of the device 100 or the device to which the device 100 is applied, etc. In addition, the number of intervals to be segmented may be set in consideration of the complexity and performance of the device and the accuracy of the bioinformation to be detected, etc.

[0074] At the same time, the interval segmentation unit 310 can adaptively adjust the predetermined number of intervals to be segmented according to the shape of the waveform of the acquired biological signal. For example, the condition of the last interval to be segmented can be set in advance based on the physiological characteristics of each component pulse, and when the last interval to be segmented does not meet the corresponding condition, the segmentation can be repeatedly performed until the corresponding condition is met.

[0075] Furthermore, the interval dividing unit 310 may allow an interval having a waveform that becomes convex downward and then convex upward to a small extent to be merged into the previous interval.

[0076] When the waveform of the biological signal is divided into constant intervals, the CCD calculation unit 320 can analyze the degree to which the waveform becomes convex downward and then convex upward for each interval, thereby calculating the CCD value. The CCD calculation unit 320 can define the convex downward portion as a positive (+) value in the biological signal waveform, and calculate the difference between the degree of convex downward at a specific time point in each interval and the time point immediately after the specific time point as the CCD value of the corresponding interval. For example, the CCD calculation unit 320 can calculate the degree C1 of the convex downward of the waveform of all biological signals at the starting time point of each interval and the degree C2 of the convex downward of the waveform at the end time point, and obtain the value obtained by subtracting C2 from C1 as the CCD value.

[0077] For example, the CCD calculation unit 320 may calculate the curvature of the start time point and the end time point of each individual interval, and calculate the CCD based on the curvature calculated in each interval. The CCD calculation unit 320 may calculate a value derived by subtracting the curvature at the end time point from the curvature at the start time point of the corresponding interval as the CCD of the corresponding interval. At this time, assuming that the waveform of the biosignal that is convex downward is defined as a positive (+) value, when the start time point or the end time point is convex upward, the negative (-) curvature derived by multiplying the value calculated at the corresponding time point by -1 can be used for the calculation of the CCD. According to an exemplary embodiment, the curvature is the rate of change indicating the degree of curvature of a curve or a surface, and when it is assumed that a very small portion of the curve is a circular arc, the radius of the circular arc may be referred to as the radius of curvature, and the reciprocal of the radius of curvature may be referred to as the curvature.

[0078] Reference Figure 4 , the radius of curvature at the starting time point of the first interval I1 is r 11 , so that the curvature k 11 is 1 / r 11 , the radius of curvature at the end time is r 12 , so that the curvature k 12 is 1 / r 12 Therefore, the waveform of the biological signal at the end time point of the first interval I1 is convex upward, so the curvature k is used to 12 Multiply the value obtained by -1 and set the CCD of the first interval I1 1 Calculated as k 11 -(-k 12 ). The waveform of the biological signal at the start time point of the second interval I2 is convex upward, so the curvature k at the start time point is used. 21 Multiply the negative value obtained by -1 and set the CCD of the second interval I2 2 Calculated as (-k 21 )-(k 22 ). In the same way, the CCD of the third interval I3 is calculated as k31 -(-k 32 ); The CCD of the fourth interval I4 4 Calculated as k 41 -(-k 42 ). The curvature k of the first interval I1 11 and curvature k 12 The curvature is larger than that of other intervals I2, I3 and I4, making CCD 1 has a maximum value; the curvature k at the starting time point of the second interval I2 21 and the curvature k at the end time of the second interval I2 22 The waveform at the start time is convex upward, making the CCD 2 is a negative value of the minimum value. In addition, the fourth interval I4 has a curvature greater than that of the third interval I3, so the CCD of the fourth interval I4 4 The CCD is calculated as the third interval I3 3 big.

[0079] The feature extraction unit 330 may select three intervals I1, I3, and I4 based on the CCD value of each interval, and extract features from the biosignals of the selected intervals. For example, the feature extraction unit 330 may select the same number of intervals as the predetermined number of characteristic points in descending order of the calculated CCD values. In addition, the feature extraction unit 330 may extract the time of a specific point of each selected interval (for example, the last point of each selected interval) and the amplitude P1, amplitude P2, and amplitude P3 of the biosignal corresponding to the time as characteristic points. In addition, the feature extraction unit 330 may use the extracted characteristic points to extract a feature F for detecting bioinformation, as shown in Equation 1 below.

[0080] Reference Figure 4 , since the CCD values ​​are calculated in the order of interval I1, interval I4, interval I3, and interval I2 in this manner, when it is assumed that the predetermined number of characteristic points is 3, the feature extraction unit 330 may select the remaining three intervals I1, interval I3, and interval I4 (i.e., intervals other than the second interval I2) according to the descending order of the CCD values. The feature extraction unit 330 may obtain the amplitudes P1, P2, and P3 from the selected intervals I1, I3, and I4 as characteristic points, and combine the amplitudes P1, P2, and P3 as shown in the following equation 1 and use the combined amplitudes as the feature F for estimating blood pressure.

[0081] [Equation 1]

[0082] F=(P1+P2) / P3

[0083] Figure 5 It shows that according to Figure 1FIG. 6 is a block diagram of another exemplary embodiment of a processing unit 120 of an apparatus 100 for extracting features of a biological signal according to an exemplary embodiment of the present invention. Figure 5 FIG. 7 is a diagram for describing a method for analyzing a waveform of a biological signal by processing unit 500 according to an exemplary embodiment of the present invention. Figure 5 1 and 10. A diagram of a method for segmenting intervals of a waveform of a biosignal performed by a processing unit 500 of an exemplary embodiment.

[0084] Reference Figure 5 , the processing unit 500 may include an interval segmentation unit 510 (e.g., an interval segmenter), a CCD calculation unit 520 (e.g., a CCD calculator), a feature extraction unit 530 (e.g., a feature extractor) and a second-order differential signal derivation unit 540 (e.g., a second-order differential signal deriver).

[0085] When a biosignal is acquired, the second-order differential signal deriving unit 540 derives a second-order differential signal by performing a second-order differential on the acquired biosignal. For example, when it is assumed that the second-order differential signal is a digital sample index whose time value t increases by 1 in the biosignal waveform function g(t), the second-order differential signal deriving unit 540 can use the following equations 2 and 3 to obtain a first-order differential signal g'(t) and a second-order differential signal g"(t).

[0086] [Equation 2]

[0087] g′(t)=g(t)-g(t-1)

[0088] [Equation 3]

[0089] g″(t)=g′(t)-g′(t-1)

[0090] The interval segmentation unit 510 may segment the waveform of the second-order differential signal into predetermined intervals based on the local maximum point and the local minimum point. At this time, the number of intervals to be segmented may be set to be greater than or equal to the number of characteristic points to be obtained. The number of characteristic points may be pre-set based on the type of biological information and biological signal to be detected. The number of intervals to be segmented and the number of characteristic points to be obtained may be pre-set to appropriate numbers in consideration of the complexity and performance of the device, the accuracy of the type of biological information to be detected, and the type of biological signal to be obtained. According to an exemplary embodiment, when the second-order differential signal waveform is defined for the degree of downward convexity, the local maximum point and the local minimum point may represent the point with the maximum convexity and the point with the maximum concavity in the local interval of the second-order differential signal waveform, respectively.

[0091] The CCD calculation unit 520 calculates the CCD value in each interval of the segmented second-order differential signal. For example, the CCD calculation unit 520 can obtain the local maximum value LMax in the segmented interval i by the following equation 4: i and the local minimum LMin i The difference between the two and the result CCD i As the CCD value of the interval i. According to an exemplary embodiment, the local maximum value may be a second-order differential value of a local maximum point, and the local minimum value may be a second-order differential value of a local minimum point.

[0092] [Equation 4]

[0093] CCD i =LMax i -LMin i

[0094] The feature extraction unit 530 extracts features from the biological signal based on the CCD value calculated by the CCD calculation unit 520. When the CCD value is calculated for each interval, the feature extraction unit 530 may arrange the CCD values ​​in descending order, select intervals corresponding to a predetermined number of characteristic points according to the descending order of the CCD values, and extract characteristic points from each selected interval. In addition, the feature extraction unit 530 may extract features for detecting biological information by using the extracted characteristic points.

[0095] For example, the feature extraction unit 530 may obtain a piece of time information and amplitude information of the biosignal corresponding to the time information from each selected interval as a feature point. According to an exemplary embodiment, the time of the local minimum point in each selected interval may be obtained as the time information, and the amplitude of the biosignal corresponding to the time may be obtained as the amplitude information. However, the information of the local maximum point may be considered to obtain the feature point. For example, the average value and the median value between the time at the local maximum point and the time at the local minimum point may be obtained as the time information. The feature extraction unit 530 may be combined such as Figure 1 The amplitude information obtained is used to extract features.

[0096] Reference Fig. 6A , Figure 6B , Figure 6C , Fig.6D and Fig. 6E , the process of extracting features performed by the processing unit 500 will be described. At this time, it is assumed that the number of characteristic points expected to be acquired is 3 and the number of intervals to be divided is 4. First, Fig. 6A The acquired biosignal is shown. When the biosignal is acquired, the second-order differential signal deriving unit 540 may derive a second-order differential signal by differentiating the biosignal. Figure 6B , Figure 6C and Fig.6D Each of the second order differential signals is shown, and the interval segmentation unit 510 can obtain the local maximum points A1, A2, A3 and A4 and the local minimum points B1, B2, B3 and B4 by searching the second order differential signals. In addition, when the local maximum points A1, A2, A3 and A4 and the local minimum points B1, B2, B3 and B4 are obtained, the interval segmentation unit 510 can segment the corresponding interval from the first local maximum point A1 to the subsequent local minimum point B1 into the first interval A1-B1, and then segment the second interval A2-B2, the third interval A3-B3 and the fourth interval A4-B4 in the same manner as the first interval A1-B1 in the described order.

[0097] The CCD calculation unit 520 calculates a CCD value for each divided interval and may calculate a value obtained by subtracting a second-order differential value of a local minimum point from a second-order differential value of a local maximum point of each interval as the CCD value of the corresponding interval.

[0098] As in Fig.6D As shown in , the feature extraction unit 530 may select three intervals corresponding to the number of characteristic points to be acquired based on the calculated CCD value of each interval. At this time, the first interval, the fourth interval, and the second interval may be selected in descending order of CCD values ​​(e.g., in descending order of the difference between the local maximum value and the local minimum value). In addition, as shown in Fig. 6E As shown in FIG. 5 , the feature extraction unit 530 can obtain the time information T corresponding to the local minimum point from the selected three intervals. 1 , time information T 2 and time information T 3 And obtain time information T from biological signals 1 , time information T 2 and time information T 3 The amplitude information P of the corresponding position 1 , amplitude information P 2 and amplitude information P 3 .

[0099] The interval segmentation unit 510 may adaptively adjust the number of intervals to be segmented in consideration of the type of the biosignal waveform (i.e., the physiological characteristics of each constituent pulse). For example, in the case where the number of intervals to be segmented is initially set to 4, when the last fourth interval does not satisfy the preset condition, the interval segmentation unit 510 may continue segmenting until the last fourth interval satisfies the condition. According to an exemplary embodiment, the preset condition may be that the local maximum value and the local minimum value used to calculate the CCD value should have positive and negative values, respectively, but the exemplary embodiment is not limited thereto.

[0100] When the second-order differential signal waveform is divided into fixed intervals, the interval division unit 510 may merge two or more intervals into a single interval in consideration of the characteristics of the waveform, etc. When a high-frequency component that fluctuates irregularly due to undesirable reasons is included in the acquired biological signal, the second-order differential signal may fluctuate unstably. In this case, the interval division unit 510 may merge two or more divided intervals based on the fluctuation of the waveform at two or more points of the divided intervals in the second-order differential signal.

[0101] For example, refer to Fig. 7A , Fig. 7A The severely fluctuating intervals F1, F2 and F3 in the second-order differential signal due to unstable waveforms are shown. When the intervals F1, F2 and F3 of the severely fluctuating second-order differential signal exist in this manner, the interval segmentation unit 510 can merge the severely fluctuating intervals F1, F2 and F3 into a single interval according to a preset standard.

[0102] Figure 7B is an example of describing a method for merging fluctuation intervals in a second-order differential signal. When the first local maximum point A1 and the local minimum point B1, the second local maximum point A2 and the local minimum point B2, and the third local maximum point A3 and the local minimum point B3 are respectively divided into the first interval, the second interval, and the third interval in the order described, the interval division unit 510 can determine whether to merge the currently divided interval into the previous interval.

[0103] For example, when dividing the first interval and then dividing the second interval, the interval dividing unit 510 may calculate the difference between the second order differential value of the local maximum point A2 of the second interval and the second order differential value of the local minimum point B1 of the first interval as the fluctuation D1 of the second interval. If the calculated fluctuation D1 of the second interval is less than the preset reference value, the interval dividing unit 510 may merge the second interval into the first interval.

[0104] Similarly, when dividing the third interval, the interval division unit 510 may calculate the difference between the second-order differential value of the local maximum point A3 of the third interval and the second-order differential value of the local minimum point B2 of the second interval as the fluctuation D2 of the third interval. Next, when the fluctuation D2 of the third interval is less than the preset reference value, the interval division unit 510 may merge the third interval into the second interval.

[0105] When the fluctuation D1 of the second interval and the fluctuation D2 of the third interval are all smaller than the threshold, the first interval, the second interval and the third interval may be merged into a single interval, and the local maximum point and the local minimum point of the merged interval are respectively the local maximum point A1 of the first interval and the local minimum point B3 of the last interval before the corresponding intervals are merged.

[0106] According to an exemplary embodiment, a reference value for determining whether to merge corresponding intervals may be appropriately set in consideration of the measurement conditions of the biosignal and the performance of the device, etc. For example, the reference value may be a value obtained by multiplying the following two values: a first value for considering the case of different measurement values ​​according to the measurement conditions of the biosignal, and a value pre-set in consideration of the performance of the device, etc. According to an exemplary embodiment, the first value may be, but is not limited to, a CCD value calculated for the first component pulse derived in a relatively stable manner.

[0107] According to the disclosed exemplary embodiments, the intervals can be segmented or merged by analyzing the waveform or the second-order differential signal of the biosignal for each fixed interval unit while considering the fluctuation of the waveform while considering the concavity and convexity, thereby reducing the above reference. FIG. 2A to FIG. 2C The above-mentioned error occurs when extracting characteristic points from a biological signal with non-ideal changes by only considering the local minimum point.

[0108] Figure 8 is a flowchart illustrating a method for extracting features of a biosignal according to an exemplary embodiment.

[0109] Figure 8 is an exemplary embodiment of a method performed by the apparatus 100 for extracting features. Figures 1 to 4 The device 100 has been described in detail, so in order to avoid repeated description, it will be briefly described.

[0110] First, in operation 810, the apparatus 100 for extracting features may acquire a biological signal. The apparatus 100 for extracting features may control a sensor (e.g., a PPG sensor) for measuring a biological signal installed in the apparatus 100 or connected to the apparatus 100 by wire or wirelessly according to a predetermined control signal, thereby acquiring a pulse signal. Alternatively, the apparatus 100 for extracting features may receive a biological signal for detecting biological information from other devices storing the biological signal.

[0111] Next, in operation 820, the apparatus 100 for extracting features may segment the acquired biological signal into fixed intervals. At this time, the number of intervals to be segmented may be set to be greater than or equal to the preset number of characteristic points. In addition, the number of intervals to be segmented may be set according to the type of biological signal, the type and accuracy of biological information to be detected, and the performance of the apparatus, and the number of intervals to be segmented may be adaptively adjusted according to the type of waveform of the biological signal.

[0112] Next, the apparatus 100 for extracting features may determine whether some of the segmented intervals should be merged at operation 830, and merge the intervals that should be merged into the previous interval at operation 840. The apparatus 100 for extracting features may determine whether a specific interval should be merged into the previous interval based on the fluctuation of the waveform of each interval. For example, the apparatus 100 for extracting features may calculate the fluctuation of the waveform of the specific interval, compare the calculated fluctuation with a predetermined reference value, and decide to merge the specific interval into the previous interval when the fluctuation is less than the predetermined reference value.

[0113] Operations 820 to 840 may be performed sequentially or simultaneously. For example, operations 830 and 840 need not be performed until all intervals are segmented in operation 820, but may be performed whenever a single interval is segmented in operation 820.

[0114] Next, in operation 850, the apparatus 100 for extracting features may analyze the concavity and convexity of the waveform of the biological signal in the segmented intervals. For example, the apparatus 100 for extracting features may calculate a CCD value, wherein the CCD value indicates the degree to which the waveform of the biological signal in each interval is convex downward and then convex upward. According to an exemplary embodiment, in the case where the waveform of the biological signal convex downward is defined as a positive (+) value, the difference between the downward convexity at the start time point of each interval and the downward convexity at the end time point may be calculated as the CCD value of the corresponding interval.

[0115] Next, in operation 860, the device 100 for extracting features may extract features from the biological signal based on the analysis result. For example, the device 100 for extracting features may select intervals corresponding to the number of characteristic points to be extracted in descending order of the calculated CCD values, and extract characteristic points from each selected interval. Next, the device 100 for extracting features may extract features for detecting biological information using the extracted characteristic points. The characteristic points may include time information and amplitude information, and the features for detecting biological information may be extracted by combining two or more characteristic points (e.g., amplitude information).

[0116] Fig. 9 is a flowchart illustrating a method for extracting features of a biosignal according to another exemplary embodiment.

[0117] Fig. 9 is another exemplary embodiment of a method performed by the apparatus 100 for extracting features. Figure 1 , Figure 5 7 and 8, the device 100 is described in detail, so in order to avoid repeated description, it will be briefly described.

[0118] First, in order to detect biological information such as blood pressure, the apparatus 100 for extracting features may acquire a biological signal in operation 910. At this time, as described above, the biological signal may be a pulse wave signal, or may be various other types of biological signals.

[0119] Next, the apparatus 100 for extracting features may derive a second-order differential signal by performing a second-order differential on the acquired biosignal in operation 920. According to an exemplary embodiment, the second-order differential signal may be derived by Equation 2 and Equation 3 as described above.

[0120] Next, in operation 930, the apparatus 100 for extracting features may divide the derived second-order differential signal into fixed intervals. At this time, the apparatus 100 for extracting features may divide the second-order differential signal into fixed intervals based on the local maximum point and the local minimum point in the waveform of the second-order differential signal. In addition, the apparatus 100 for extracting features may adaptively adjust the preset number of intervals to be divided in consideration of the type of the biological signal waveform (i.e., the physiological characteristics of each constituent pulse).

[0121] Next, in operation 940, the device 100 for extracting features may determine whether to merge some of the segmented fixed intervals. For example, when a high-frequency component that jitters irregularly due to undesirable reasons is included in the acquired biological signal, the second-order differential signal may jitter unstably. In this way, in the case of extracting features from the acquired biological signal using a general method, search errors such as extracting characteristic points at inaccurate points may occur. In order to solve this problem, the device 100 for extracting features may calculate the difference between the local maximum value of the segmented interval and the local minimum value of the previous interval as a fluctuation, and determine whether to merge two or more segmented intervals based on the calculated fluctuation. That is, the device 100 for extracting features may compare the calculated fluctuation with a preset reference value for a specific interval, and decide to merge the specific interval into the previous interval when the fluctuation is less than the reference value.

[0122] Next, the apparatus 100 for extracting features may merge some of the intervals determined to be merged into the previous interval in operation 950. When two or more consecutive intervals are merged by the process as described above, the local maximum point and the local minimum point of the merged interval may be the local maximum point of the first interval and the local minimum point of the last interval before the corresponding intervals are merged, respectively.

[0123] Meanwhile, operations 930 to 950 may be performed sequentially or simultaneously. For example, operations 940 and 950 may be performed sequentially after all intervals are divided in operation 930. Alternatively, when a single interval is divided in operation 930, operations 940 and 950 may be performed immediately, and operations 930 to 950 may be repeatedly performed until the operation reaches the number of intervals to be divided.

[0124] Next, the apparatus 100 for extracting features may calculate a CCD value for each interval to be segmented in operation 960. At this time, the CCD value may be, but is not limited to, a difference between a local maximum value and a local minimum value of each interval.

[0125] Next, in operation 970 , when the CCD values ​​are calculated, the apparatus 100 for extracting features may arrange the CCD values ​​in descending order and select a section corresponding to the number of characteristic points to be acquired according to the descending order of the CCD values.

[0126] Next, the apparatus 100 for extracting features may extract characteristic points from the selected intervals at operation 980, and extract features for detecting biological information using the extracted characteristic points at operation 990. For example, the apparatus 100 for extracting features may acquire a time position corresponding to a local minimum point in each selected interval as time information of the characteristic point, and acquire an amplitude of a position corresponding to the time information in the biological signal as amplitude information of the characteristic point. When characteristic points such as time information and amplitude information are acquired, features for detecting biological information may be extracted based on the acquired characteristic points or a combination thereof.

[0127] Fig.10 is a block diagram illustrating an apparatus for detecting biological information according to an exemplary embodiment, Fig.11 is a detailed block diagram illustrating a processing unit of an apparatus for detecting biological information according to an exemplary embodiment.

[0128] The device 1000 for detecting biological information may be a device for measuring biological information (such as blood pressure) of an object, wherein various exemplary embodiments of the device 100 for extracting features of biological signals as described above are installed in the device 1000 for detecting biological information. For example, the device 1000 for detecting biological information may be a cuffless indirect blood pressure measurement device. The device 1000 for detecting biological information may be implemented in the form of a wearable device that can be worn by an object. For example, the device 1000 for detecting biological information may be implemented in the form of a watch, a bracelet, and a wristband. However, the device 1000 for detecting biological information is not limited thereto, and may be implemented in the form of a ring, glasses, a hairband, and the like.

[0129] Hereinafter, for convenience of description, although the biological information to be detected and the biological signal to be measured will be described as blood pressure and a pulse wave signal, respectively, exemplary embodiments are not limited thereto.

[0130] Reference Fig.10 , the apparatus 1000 for detecting biological information may include a sensor 1010, a processing unit 1020 (eg, a processor), an interface unit 1030 (eg, a display), and a communication unit 1040 (eg, a communicator).

[0131] When receiving a predetermined control signal, the sensor 1010 can illuminate the object with light by driving the light source, and measure the pulse signal by detecting the light reflected or scattered from the object. According to an exemplary embodiment, the object can be a living area that can contact or be adjacent to the sensor 1010 as a biological information detection target, and the object can be a part of the human body that can easily measure the pulse wave by PPG (photoplethysmography). For example, the object can be a surface area of ​​the wrist adjacent to the radial artery. When the pulse wave is measured on the skin surface of the wrist through which the radial artery passes, the technology may be relatively less affected by external factors that cause errors in the measurement (such as the thickness of the skin tissue in the wrist). It is well known that the radial artery corresponds to a blood vessel that can measure accurate blood pressure compared to other blood vessels in the wrist. However, the object is not limited to the surface area of ​​the wrist adjacent to the radial artery, but can also be a distal part of the human body with a high blood vessel density, such as fingers and toes.

[0132] The processing unit 1020 may receive a control command received from a user through an interface module installed in the device 1000 for detecting biological information, and process an operation corresponding to the control command. At this time, the interface module may be, but is not limited to, a display, a microphone, a speaker, a tactile device, etc. installed in the device 1000 for detecting biological information. The processing unit 1020 may perform various functions for performing interaction with the user through the interface module. As an example, a graphical user interface may be displayed on a display so that a user can input a control command by touching the display. As another example, a dialogue management function and a voice recognition function may be installed so that a user can perform interaction (such as input of a control command) using voice through a microphone and a speaker, etc.

[0133] When the user inputs a control command for requesting detection of biological information, the processing unit 1020 may generate a signal for controlling the sensor 1010 and transmit the generated signal to the sensor 1010 .

[0134] When the pulse wave signal is measured by the sensor 1010, the processing unit 1020 may analyze the pulse wave signal in each fixed interval unit and extract features for detecting blood pressure. For example, the processing unit 1020 may analyze the concavity and convexity of the pulse wave signal in each fixed interval unit to extract characteristic points (such as time information or amplitude information) from the pulse wave signal and extract features using the extracted characteristic points.

[0135] Reference Fig.11 , the processing unit 1100 may include a bio-signal analyzing unit 1100 (eg, a bio-signal analyzer) and a bio-information detecting unit 1120 (eg, a bio-information detector).

[0136] The biosignal analysis unit 1100 may obtain a second-order differential signal by performing a second-order differential on the pulse wave signal measured by the sensor 1010. In addition, when the second-order differential signal is obtained, the biosignal analysis unit 1110 may divide the second-order differential signal in each fixed interval unit and analyze the concavity and convexity of the waveform in each divided interval unit. At this time, in order to detect blood pressure, the number of intervals to be divided may be pre-set to be greater than or equal to the number of characteristic points desired to be extracted. In addition, the biosignal analysis unit 1100 may adaptively adjust the number of intervals to be divided according to the type of waveform.

[0137] For example, the biosignal analysis unit 1100 may segment the second-order differential signal into fixed intervals based on a local maximum point having a local convex shape (convex upward) and a local minimum point having a local concave shape (convex downward) in the second-order differential signal. The biosignal analysis unit 1100 may merge two or more segmented intervals in consideration of the fluctuation of the waveform of the second-order differential signal. The biosignal analysis unit 1100 may calculate the difference between the second-order differential value of the local maximum point of the current interval and the second-order differential value of the local minimum point of the previous interval as the fluctuation, and merge the current interval into the previous interval when the fluctuation is less than a preset reference value.

[0138] The biosignal analysis unit 1100 may calculate a CCD value corresponding to a difference between a second-order differential value of a local maximum point and a second-order differential value of a local minimum point of each interval, and select an interval from which characteristic points are to be extracted in descending order of the CCD value. The biosignal analysis unit 1100 may obtain time information of an interval corresponding to a local minimum point in the selected interval and amplitude information of a position corresponding to the time information from the pulse wave signal.

[0139] When one or more characteristic points (such as time information or amplitude information) are derived, the bio-information detection unit 1120 may extract features for detecting blood pressure based on at least one of the acquired characteristic points or a combination thereof, and use the extracted features to detect blood pressure. In addition, the bio-information detection unit 1120 may monitor the health status of the user based on the detected blood pressure. When it is determined based on the monitoring result that the health status is abnormal, the bio-information detection unit 1120 may generate additional information such as a risk level to warn the user.

[0140] Refer again Fig.10 The interface unit 1030 may provide the blood pressure detected by the processing unit 1020, additional information such as risk level, and warning information to the user through an interface module installed in the device 1000 or connected to the device 1000 via wired or wireless communication.

[0141] For example, the interface unit 1030 may display blood pressure information or risk information of a health condition on a display and provide the displayed information to a user. At this time, the interface unit 1030 may display the corresponding information using colors predefined according to risk levels as shown in Table 1 below, so that the user can easily identify the health condition. In addition, the interface unit 1030 may display a warning message predefined according to the risk level on the display, etc.

[0142] In addition, the interface unit 1030 can output blood pressure, risk level, warning information, etc. in an auditory manner using a speaker. In addition, when a tactile device is installed in or connected to the device 1000, the interface unit 1030 can provide the risk level or warning information to the user through touch or vibration or other types of feedback. Blood pressure, risk level, warning information, etc. can be provided to the user by combining two or more visual methods or non-visual methods according to the type of interface module, the performance of the device 1000, the purpose of providing information, etc.

[0143] [Table 1]

[0144] Risk Level color Shock / pressure intensity high red powerful middle yellow middle Low green weak

[0145] The communication unit 1040 can send various information to other external devices and receive various information from other external devices by communicating with other external devices via wireless communication. According to an exemplary embodiment, the wireless communication may include Bluetooth communication, BLE (Bluetooth Low Energy) communication, near field communication, WLAN (Wireless Local Area Network) communication, ZigBee communication, IrDA (Infrared Data Association) communication, WFD (Wi-Fi Direct) communication, UWB (Ultra Wideband) communication, Ant+ communication, Wi-Fi communication and mobile communication, but the exemplary embodiment is not limited thereto.

[0146] The communication unit 1040 can send the results processed by the processing unit 1020 to other external devices with higher performance, so that the other external devices can perform desired operations. For example, the communication unit 1040 can send the detected blood pressure and the like to a smart phone, a tablet PC, a desktop PC, a notebook PC, etc., so that the user can confirm the processed results through a larger display screen.

[0147] In addition, when the user inputs a control command (such as blood pressure detection, etc.) through an external device (such as a smart phone, etc.), the communication unit 1040 can receive the control command from the external device and send the received control command to the processing unit 1020.

[0148] The processing unit 1020 may store the processed results (e.g., detected blood pressure information, generated additional information, blood pressure history information, health status monitoring information, and various static information, etc.) in an external device installed in or connected to the apparatus 1000 for detecting biological information, and manage the stored information. At this time, the storage device may include, but is not limited to, at least one storage medium of a flash memory type, a hard disk type, a multimedia card micro, a card type memory (e.g., an SD or DX memory), a RAM (random access memory), an SRAM (static random access memory), a ROM (read-only memory), an EEPROM (electrically erasable programmable read-only memory), a PROM (programmable read-only memory), a magnetic memory, a magnetic disk, and an optical disk.

[0149] The apparatus 1000 for detecting biological information may include a main body, and the components 1010 , 1020 , 1030 , and 1040 described above may be installed in the main body.

[0150] Fig.12 is a flowchart illustrating a method for detecting bio-information according to an exemplary embodiment.

[0151] Fig.12 An exemplary embodiment of a method for detecting biological information performed by the apparatus 1000 for detecting biological information is shown.

[0152] First, at operation 1210, the apparatus 1000 for detecting biological information may receive a request for detecting biological information from a user. For example, the user may input a control command for requesting detection of biological information by operating an operation unit installed in the apparatus 1000 for detecting biological information. Alternatively, the apparatus 1000 for detecting biological information may receive a control command for requesting detection of biological information from other external devices.

[0153] Next, the apparatus 1000 for detecting bio-information may measure a bio-signal by controlling a sensor according to a control command for requesting detection of the bio-information at operation 1220. At this time, the sensor may illuminate an object with light by driving a light source according to the control command, and detect light reflected or scattered from the object to measure a bio-signal such as a pulse wave.

[0154] Next, the apparatus 1000 for detecting biological information may derive a second-order differential signal by performing a second-order differential on the acquired biological signal at operation 1230, and divide the derived second-order differential signal into fixed intervals in consideration of the waveform of the derived second-order differential signal at operation 1240. At this time, the apparatus 1000 for detecting biological information may adaptively adjust the number of intervals to be divided based on the local maximum point and the local minimum point of the second-order differential signal while dividing the corresponding intervals in consideration of the waveform of the second-order differential signal. In addition, as described above, the apparatus 1000 for detecting biological information may merge the divided intervals in consideration of the fluctuation of the waveform.

[0155] Next, the apparatus 1000 for detecting bio-information may calculate a CCD value indicating concavity and convexity of the segmented intervals in operation 1250. At this time, the CCD value may be obtained by calculating the difference between second-order differential values ​​of a local maximum point and a local minimum point of each interval.

[0156] Next, the apparatus 1000 for detecting biological information may select an interval corresponding to the number of characteristic points based on the CCD value at operation 1260, and extract a feature of the biological signal from the selected interval at operation 1270. According to an exemplary embodiment, the apparatus 1000 for detecting biological information may extract time information corresponding to a local minimum point of the selected interval and amplitude information of the biological signal corresponding to the time information as characteristic points, and extract features for detecting biological information through the extracted characteristic points or a combination thereof.

[0157] Next, the apparatus 1000 for detecting bio-information may detect bio-information by using the extracted features in operation 1280. According to an exemplary embodiment, the apparatus 1000 for detecting bio-information may generate a risk level of a user's health condition by analyzing the detected bio-information (eg, blood pressure information).

[0158] Next, in operation 1290 , the apparatus 1000 for detecting bio-information may provide additional information related to the detected bio-information and a risk level to a user, etc. According to exemplary embodiments, as described above, information may be provided to a user through various visual methods or non-visual methods.

[0159] The exemplary embodiments may be implemented as computer-readable codes in a computer-readable recording medium. The computer-readable recording medium includes various types of recording media storing computer-readable data.

[0160] Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, and optical data storage. In addition, the recording medium may be implemented in the form of a carrier wave (e.g., transmission via the Internet). In addition, the computer-readable recording medium may be distributed to a networked computer system, in which the computer-readable code may be stored and executed in a distributed manner. In addition, the functional program and associated code and code segments for implementing the present disclosure may be easily deduced by a programmer in the field to which the present disclosure belongs.

[0161] Many examples have been described above. However, it will be appreciated that various modifications may be made. For example, if the described methods are performed in a different order, and / or if components in the described systems, architectures, devices, or circuits are combined in a different manner and / or replaced or supplemented by other components or their equivalents, suitable results may be obtained. Therefore, other implementations are also within the scope of the claims.

Claims

1. A device configured to extract features from a biological signal, the device comprising: A biological signal acquirer configured to acquire a biological signal; a processor configured to analyze concavity and convexity of a waveform of the biosignal in a plurality of intervals of the biosignal and extract features from the biosignal based on the analyzed concavity and the analyzed convexity, The processor includes: a convex-concave difference calculator configured to calculate a convex-concave difference value based on the concavity and convexity of the waveform of the biosignal in the plurality of intervals; a feature extractor configured to extract characteristic points from the biosignal based on the calculated convex-concave difference value, and extract features from the biosignal based on the characteristic points, The characteristic points include time information in the interval and amplitude information of the biological signal corresponding to the time information. The feature extractor is configured to select a preset number of intervals from the plurality of intervals in descending order of the calculated convex-concave difference values, and extract features from each selected interval.

2. The device according to claim 1, wherein: The convex-concave difference calculator is configured to calculate a difference in upward convexity between a specific time point in one interval and a previous time point or a difference in downward convexity between a specific time point in the one interval and a next time point as a convex-concave difference value of the one interval.

3. The device according to claim 1, wherein: The processor also includes: The interval divider is configured to divide the biosignal into a predetermined number of intervals.

4. The device according to claim 3, wherein: The interval divider is configured to determine a predetermined number of intervals to be divided in consideration of at least one of complexity of the device, performance of the device, and accuracy of bio-information included in the bio-signal.

5. The device according to claim 3, wherein: The interval divider is configured to adaptively adjust the number of intervals to be divided according to the type of waveform of the biosignal.

6. The device according to claim 1, wherein: The biosignal includes at least one of an electrocardiogram signal, a photoplethysmogram signal, and an electromyogram signal.

7. A method for extracting features from a biological signal, the method comprising: Acquiring biological signals; analyzing concavity and convexity of a waveform of a biosignal in a plurality of intervals of the biosignal; Extracting features from biosignals based on analyzed concavity and analyzed convexity, The step of analyzing includes: calculating a convex-concave difference value based on the concavity and convexity of the waveform of the biological signal in the plurality of intervals, The step of extracting includes: extracting characteristic points from the biological signal based on the calculated convex-concave difference value, and extracting features from the biological signal based on the characteristic points, The characteristic points include time information in the interval and amplitude information of the biological signal corresponding to the time information. The step of extracting includes: selecting a preset number of intervals from the plurality of intervals in descending order of the calculated convex-concave difference values, and extracting features from the selected intervals.

8. The method according to claim 7, further comprising: The biosignal is segmented into a predetermined number of intervals.

9. The method according to claim 8, wherein: The step of segmenting includes adaptively adjusting the number of intervals to be segmented according to the type of waveform of the biosignal.

10. A device configured to extract features from a biological signal, the device comprising: A biological signal acquirer configured to acquire a biological signal; a processor configured to: derive a second order differential signal of the biosignal, analyze concavity and convexity of a waveform of the second order differential signal, and extract features from the biosignal based on the analyzed concavity and the analyzed convexity, The processor includes: a convex-concave difference calculator configured to calculate a difference between a local maximum value and a local minimum value in a plurality of intervals of a second-order differential signal of a biological signal as a convex-concave difference value; a feature extractor configured to extract characteristic points from the biosignal based on the calculated convex-concave difference value, and extract features from the biosignal based on the characteristic points, The characteristic points include the time information of the local minimum point and the amplitude information of the biological signal corresponding to the time information. The processor further comprises: an interval divider configured to divide the derived second-order differential signal into a predetermined number of intervals, The interval divider is configured to merge two or more divided intervals based on the difference between the local minimum value and the next local maximum value in the second-order differential signal. The feature extractor is configured to select a preset number of intervals from the plurality of intervals in descending order of the calculated convex-concave difference values, and extract features based on a local minimum point of each selected interval.

11. The device according to claim 10, wherein: The interval divider is configured to divide a portion of the waveform from a local maximum point to a next local minimum point of the second-order differential signal into one interval.

12. A method for extracting features from a biological signal, the method comprising: Acquiring biological signals; Derive the second-order differential signal of the biological signal; Analyze the concavity and convexity of the waveform of the second-order differential signal; Extracting features from biosignals based on analyzed concavity and analyzed convexity, The analysis step includes: calculating the difference between the local maximum value and the local minimum value in multiple intervals of the second-order differential signal of the biological signal as the convex-concave difference value, The step of extracting includes: extracting characteristic points from the biological signal based on the calculated convex-concave difference, and extracting features from the biological signal based on the characteristic points, The characteristic points include the time information of the local minimum point and the amplitude information of the biological signal corresponding to the time information. Wherein, the method further comprises: Split the derived second-order differential signal into a predetermined number of intervals, Merge two or more partitioned intervals based on the difference between the local minimum and the next local maximum in the second-order differential signal, The step of extracting includes: selecting a preset number of intervals from the plurality of intervals in descending order of the calculated convex-concave difference values, and extracting features based on a local minimum point of each selected interval.

13. The method according to claim 12, wherein: The step of segmenting includes segmenting an interval from a local maximum point to a next local minimum point of the second-order differential signal into a single interval.

14. A device configured to detect biological information, the device comprising: a sensor configured to: illuminate an object with light and detect light reflected by the object to measure a biosignal; a processor configured to, when a biological signal is measured, extract features from the biological signal based on concavity and convexity of a waveform of the biological signal in a plurality of intervals of the biological signal, and detect biological information based on the extracted features, The processor is configured to: in response to measuring the biological signal, derive a second-order differential signal of the measured biological signal, and divide the derived second-order differential signal into the plurality of intervals based on a local maximum point and a local minimum point of the derived second-order differential signal, The processor is configured to calculate the difference between the local maximum value and the local minimum value in the segmented interval of the second-order differential signal of the biological signal as the convex-concave difference value, The processor is configured to: select a predetermined number of intervals from the segmented multiple intervals based on the convex-concave difference value, extract time information of the local minimum point of the selected interval and amplitude information of the biological signal position corresponding to the time information, and detect the biological information by combining at least two extracted amplitude information, The processor is configured to merge the selected two or more intervals based on the difference between a local minimum value and a next local maximum value in the second-order differential signal.

15. The device according to claim 14, wherein: The processor is configured to monitor the health condition of the user based on the biometric information and generate additional information according to the health condition.

16. The device according to claim 15, wherein: The biological information includes blood pressure information, and the additional information includes warning information.

17. The apparatus according to claim 15, further comprising: The display is configured to display the detected biological information and additional information.

18. A medical imaging device comprising: A biological signal acquirer configured to acquire a biological signal; a processor configured to: divide the waveform of the biosignal into a plurality of intervals, calculate a convex-concave difference value based on the concavity and convexity of the waveform of the biosignal in the plurality of intervals, select a preset number of intervals from the divided plurality of intervals in descending order of the calculated convex-concave difference values, eliminate one interval from the plurality of intervals, extract characteristic points from the biosignal based on the remaining intervals of the plurality of intervals excluding the eliminated interval, and calculate a feature of the biosignal based on the characteristic points, The characteristic points include time information in the remaining interval and amplitude information of the biological signal corresponding to the time information.

19. The medical imaging device of claim 18, wherein: The biosignal is a signal acquired from a living subject, and the eliminated interval corresponds to the dicrotic notch of the living subject.

Images