Program, information processing method, and information processing device

Abstract

To provide a program, etc. capable of extracting a part in which a pulse wave is appropriately measured from a photoelectric pulse wave signal.SOLUTION: An accuracy score calculation system 1 for a photoelectric pulse wave signal acquires a photoelectric pulse wave signal detected by a photoelectric pulse wave meter, calculates a score on a photoelectric pulse wave waveform on the basis of the photoelectric pulse wave signal, and causes a computer to execute processing for extracting data of a part that the score satisfies a condition for a predetermined threshold from the photoelectric pulse wave signal. Preferably, the accuracy score calculation system 1 for a photoelectric pulse wave signal calculates a first score indicating a collapsing degree of a waveform according to whether or not the photoelectric pulse wave signal exceeds an intensity threshold on light reception intensity, and extracts data of a part that satisfies a condition for the threshold on the basis of the first score.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present disclosure relates to a program, an information processing method, and an information processing device. [Background technology]

[0002] Heart failure and other heart diseases are collectively called one of the three major diseases along with malignant neoplasms and stroke. Because the prognosis for patients with heart failure is poor, medical institutions need to continuously monitor the condition of the patient. Patients with heart failure have a faster pulse rate than healthy people in order to deliver sufficient blood to the entire body.

[0003] Photoplethysmography signals are measured by shining light from an LED or other source onto a fingertip, earlobe, etc., and comparing the irradiated light with the transmitted or reflected light. Using the photoplethysmography signal, medical personnel can measure a patient's blood volume and pulse rate.

[0004] Patent Document 1 discloses a technology that uses a pulse wave sensor that utilizes a photoelectric pulse wave signal and an acceleration sensor to detect a user's pulse wave signal and body movement signal, performs noise removal processing based on the body movement signal from the pulse wave signal, and extracts the pulse wave. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] JP2010-172645A Summary of the Invention [Problem to be solved by the invention]

[0006] An object of the present invention is to provide a program or the like capable of extracting an appropriately measured portion of a pulse wave from a photoelectric pulse wave signal. [Means for solving the problem]

[0007] The present invention provides (1) a program that causes a computer to execute a process of acquiring a photoplethysmographic signal detected by a photoplethysmograph, calculating a score related to the photoplethysmographic waveform based on the photoplethysmographic signal, and extracting data from the photoplethysmographic signal for a portion of the score that satisfies a condition relative to a predetermined threshold value. Here, in an embodiment of the present invention, (2) It is preferable that the program in (1) above calculates a first score indicating the degree of waveform distortion depending on whether the photoplethysmographic signal exceeds an intensity threshold of the received light intensity, and extracts data that satisfies a condition for the threshold based on the first score. (3) It is preferable that the program in (2) above calculates the first score by generating an image of a signal waveform from the photoelectric pulse wave signal and determining, based on the image, the degree of waveform distortion depending on whether the photoelectric pulse wave signal exceeds an intensity threshold of the received light intensity. (4) It is preferable that the program described in any one of (1) to (3) above calculates a second score for evaluating the periodicity of the photoelectric pulse waveform based on an autocorrelation analysis of the photoelectric pulse wave signal, and extracts data of a portion that satisfies a condition for the threshold value based on the second score. (5) It is preferable that the program in (4) above derives an envelope of the photoplethysmographic signal, and calculates the second score based on autocorrelation analysis of waveform data obtained by dividing the photoplethysmographic signal by the envelope. (6) It is preferable that the program described in any one of (1) to (5) above extracts waveform portions from the photoplethysmogram signal that match a predetermined photoplethysmogram waveform template, calculates a third score representing the degree of match with the template depending on the number of extracted waveform portions, and extracts data of portions that satisfy a condition for the threshold value based on the third score.

[0008] The present invention provides an information processing method (7) in which a computer executes a process of acquiring a photoplethysmographic signal detected by a photoplethysmograph, calculating a score related to the photoplethysmographic waveform based on the photoplethysmographic signal, and extracting data from the photoplethysmographic signal where the score satisfies a condition relative to a predetermined threshold value.

[0009] The present invention provides (8) an information processing device that includes a control unit, which acquires a photoplethysmographic signal detected by a photoplethysmograph, calculates a score related to the photoplethysmographic waveform based on the photoplethysmographic signal, and extracts data from the photoplethysmographic signal for a portion of the score that satisfies a condition relative to a predetermined threshold value.

[0010] The present invention (9) provides a program that causes a computer to execute a process of acquiring a photoplethysmographic signal detected by a photoplethysmograph, matching a predetermined photoplethysmographic waveform template with a signal waveform indicated by the photoplethysmographic signal, and extracting data of a waveform portion that matches the template from the photoplethysmographic signal. Effect of the Invention

[0011] According to the present invention, it is possible to extract a portion where the pulse wave is appropriately measured from a photoelectric pulse wave signal. [Brief description of the drawings]

[0012] [Figure 1] FIG. 1 is an explanatory diagram showing an example of the configuration of a photoplethysmographic signal accuracy score calculation system; [Diagram 2] FIG. 10 is an explanatory diagram showing an example of a photoplethysmographic signal acquired by a wearable device. [Diagram 3] FIG. 11 is an explanatory diagram showing an example of a method for calculating a first score. [Figure 4] FIG. 11 is an explanatory diagram showing an example of a method for calculating a second score. [Diagram 5] FIG. 11 is an explanatory diagram showing an example of a method for creating a template. [Figure 6] FIG. 11 is an explanatory diagram showing an example of a method for calculating a third score. [Figure 7] 10 is a flowchart showing information processing of a photoplethysmographic signal accuracy score calculation system. [Figure 8] 13 is a flowchart showing a calculation process of a first score. [Figure 9] 13 is a flowchart showing a calculation process of a second score. [Figure 10] 13 is a flowchart showing a calculation process of a third score. [Figure 11] 1 is a flowchart showing a method for creating a template. [Figure 12] 13 is a flowchart showing a method for evaluating each score. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] (Embodiment) The present invention will be described in detail below with reference to the drawings showing embodiments thereof. FIG. 1 is an explanatory diagram showing an example of the configuration of a photoelectric pulse wave signal accuracy score calculation system. Photoelectric pulse wave signal accuracy score calculation system 1 acquires a user's photoelectric pulse wave signal, calculates a score related to the acquired photoelectric pulse wave signal, and extracts photoelectric pulse wave signals for which the calculated score satisfies a condition relative to a predetermined threshold. Photoelectric pulse wave signal accuracy score calculation system 1 of this embodiment includes a wearable device 150 and a server 100. Wearable device 150 and server 100 are connected via a communication network.

[0014] Wearable device 150 includes a control unit 151, a storage unit 152, a light emitting unit 153, a light receiving unit 154, a data processing unit 155, and a communication unit 156, which are connected via a bus. Wearable device 150 measures the photoelectric pulse wave signal of the user and transmits the measured photoelectric pulse wave signal to server 100.

[0015] In this embodiment, a wearable device 150 will be described as an example of a photoplethysmograph, but the present invention is not limited to this.

[0016] There are two types of photoplethysmography signal measuring devices: a transmission type and a reflection type. The transmission type has a configuration in which the light-emitting unit 153 and the light-receiving unit 154 are arranged to sandwich the user's fingertip, earlobe, etc. The transmission type irradiates the user's fingertip, earlobe, etc. with red light from the light-emitting unit 153, and detects the red light that has passed through the user's fingertip, earlobe, etc. with the light-receiving unit 154. The reflection type has a configuration in which the light-emitting unit 153 and the light-receiving unit 154 are arranged adjacent to each other. The reflection type irradiates the user's body surface with red light and green light, etc. with the light-receiving unit 154, and detects the red light and green light, etc. reflected inside the user's body.

[0017] In this embodiment, only an example using a reflective type will be described, but a transmissive type may also be used.

[0018] The control unit 151 includes one or more processors such as a CPU (Central Processing Unit), an MPU (Micro Processing Unit), or a GPU (Graphics Processing Unit). The control unit 151 may be configured by combining DSPs (Digital Signal Processors), FPGAs (Field Programmable Gate Arrays), or the like. The control unit 151 reads out and executes a control program 152P (program product) stored in the storage unit 152, thereby performing various information processing and control processing related to the wearable device 150, and realizing the functions of the storage unit 152, the light emitting unit 153, the light receiving unit 154, the data processing unit 155, and the communication unit 156.

[0019] The storage unit 152 includes a static random access memory (SRAM), a dynamic random access memory (DRAM), a flash memory, etc. The storage unit 152 temporarily stores a control program 152P and data required for the control unit 151 to execute arithmetic processing.

[0020] The light-emitting unit 153 includes, for example, an LED light source having a wavelength range of red light and green light, etc. The light-emitting unit 153 may also include an LED light source having a different wavelength range. The control unit 151 uses the light-emitting unit 153 to irradiate light onto the user's body surface, etc.

[0021] The light receiving unit 154 includes a photodiode, an infrared cut filter, etc. The photodiode detects reflected light of the LED light irradiated from the light emitting unit 153 to the user, and converts the detected reflected light into an electrical signal. The infrared cut filter reduces the influence of disturbance light such as infrared light detected by the photodiode.

[0022] Data processing unit 155 includes a signal amplifier, an AD converter, a microcomputer, etc. The signal amplifier amplifies the electrical signal converted by light receiving unit 154. The AD converter performs A / D conversion on the amplified electrical signal. The microcomputer performs various arithmetic processing and acquires a photoplethysmographic signal from the A / D converted electrical signal.

[0023] The control unit 151 may transmit the electrical signal A / D converted by the data processing unit 155 to the server 100. In this case, the server 100 converts the electrical signal A / D converted by the wearable device 150 into a photoplethysmographic signal.

[0024] The communication unit 156 transmits the photoplethysmographic signal to the server 100. Although only one wearable device 150 is illustrated in Fig. 1, multiple wearable devices 150 may be provided for each target user.

[0025] Communication unit 156 may assign an identifier (ID) to the photoelectric pulse wave signal and transmit it to server 100. In this case, the identifier is composed of a plurality of letters and numbers including information such as the user name and the date and time when the photoelectric pulse wave signal was acquired.

[0026] The server 100 includes a communication unit 10, a storage unit 20, and a control unit 30.

[0027] In this embodiment, the server 100 will be described as an example of the information processing device, but the information processing device is not limited to this.

[0028] The communication unit 10 realizes a communication function with the wearable device 150. The communication unit 10 receives a photoelectric pulse wave signal of the user transmitted by the wearable device 150 via a communication network. The communication unit 10 may be an interface for transmitting and receiving information to and from the wearable device 150 by wired communication via a cable.

[0029] The storage unit 20 includes a static random access memory (SRAM), a dynamic random access memory (DRAM), a flash memory, etc. The storage unit 20 acquires in advance various data necessary for the execution of the control unit 30. The storage unit 20 stores data etc. that are generated when the control unit 30 executes various processes.

[0030] The control unit 30 includes one or more processors such as a central processing unit (CPU), a micro processing unit (MPU), or a graphics processing unit (GPU). The control unit 30 may also be configured by combining digital signal processors (DSPs) or field programmable gate arrays (FPGAs). The control unit 30 includes a semiconductor memory such as a flash memory, and stores a computer program (program product). The control unit 30 can execute a computer program deployed in the semiconductor memory. The computer program may be downloaded from an external device via the communication unit 10 and stored in the semiconductor memory. The control unit 30 may also read a computer program stored in a storage medium 1a (e.g., an optically readable disk storage medium such as a CD-ROM) and store the read computer program in the semiconductor memory. The computer program may also be stored in the storage unit 20 of a hard disk. In this case, the computer program may be deployed so as to be executable on a single computer or on multiple computers located at one site, or distributed across multiple sites and interconnected by a communication network.

[0031] The control unit 30 removes specific frequency band components contained in the photoplethysmographic signal using existing techniques such as a low-pass filter, a high-pass filter, and a band-pass filter, etc. A detailed description of the above-mentioned techniques will be omitted.

[0032] The control unit 30 includes a first score calculation function 31, a second score calculation function 32, a third score calculation function 33, and a score evaluation function 34. Details will be described later.

[0033] Fig. 2 is an explanatory diagram showing an example of a photoelectric pulse wave signal acquired by a wearable device. Fig. 2 shows an example of a photoelectric pulse wave signal, which includes a signal waveform portion (area A in Fig. 2) that is less affected by external factors such as body movement and disturbance light, and a signal waveform portion (area B) that is more affected by body movement, etc. The vertical axis of the graph of the photoelectric pulse wave signal indicates voltage (received light intensity of LED light), and the horizontal axis indicates measurement time.

[0034] In this embodiment, server 100 acquires the user's photoelectric pulse wave signal from wearable device 150, and extracts feature quantities such as pulse rate from the acquired photoelectric pulse wave signal. Here, as shown in area B of Fig. 2, when the influence of external factors such as body movement is large, noise is introduced into the photoelectric pulse wave signal, making it difficult to appropriately extract the pulse rate, etc.

[0035] In such cases, in the past, additional sensors such as acceleration sensors were used to identify and exclude areas affected by body movement, etc., but this increases the number of sensors, leading to increased costs and power consumption.

[0036] Therefore, in this embodiment, a score indicating whether the photoelectric pulse wave has been properly measured is calculated from the photoelectric pulse wave signal itself, and the signal waveform portion where the calculated score satisfies a condition relative to a preset threshold value is extracted as the portion where the photoelectric pulse wave has been properly measured.

[0037] Specifically, as described below, server 100 calculates three types of scores (first to third scores) from the photoplethysmographic signal, and extracts data of signal waveform portions where each score satisfies a condition relative to a preset threshold value.

[0038] First, the procedure for calculating the first score will be described.

[0039] First score calculation function 31 (control unit 30) measures the degree of corruption of the acquired photoelectric pulse wave signal, and calculates the rate at which the photoelectric pulse wave signal is corrupted as a first score.

[0040] Fig. 3 is an explanatory diagram showing an example of a method for calculating the first score. Fig. 3 shows a photoplethysmographic signal acquired by wearable device 150 and a preset upper limit of voltage (threshold of received light intensity). The vertical axis of the graph indicates voltage, and the horizontal axis indicates measurement time.

[0041] The photoplethysmographic signal shown in Fig. 3 is composed of reflected light of the LED light irradiated to the user, disturbance light, etc. When the output of the LED light source irradiated from the light-emitting unit 153 is increased, the reflected light of the LED light irradiated to the user increases. Therefore, the light-receiving unit 154 can detect more reflected light of the LED light irradiated from the light-emitting unit 153 compared to disturbance light, etc.

[0042] However, when the output of the reflected light of the LED light irradiated on the user exceeds a preset upper voltage limit, a breakdown in the photoplethysmographic signal is detected. The breakdown in the photoplethysmographic signal is, for example, an area where part of the signal waveform becomes rectangular (C, D, and E in FIG. 3). In other words, the breakdown in the photoplethysmographic signal occurs when the voltage value of the photoplethysmographic signal exceeds the upper limit of the measurable voltage, i.e., the intensity threshold of the received light intensity.

[0043] The intensity threshold value of the received light intensity does not have to be set to the same value as the upper limit value, but may be set to a value lower than the upper limit value.

[0044] First score calculation function 31 calculates the degree of corruption of the photoplethysmographic signal as a first score.

[0045] In FIG. 3, the first score calculation function 31 identifies the portion (C, D, and E in FIG. 3) of the photoelectric pulse wave signal that is higher than a preset upper voltage limit. The first score calculation function 31 calculates the time (c, d, and e) during which the voltage value of the photoelectric pulse wave signal exceeded the upper voltage limit based on the time when the photoelectric pulse wave signal was acquired. The first score calculation function 31 divides the total time during which the voltage value of the photoelectric pulse wave signal exceeded the upper voltage limit by the total time (T1) during which the photoelectric pulse wave signal was measured. The control unit 30 calculates the value obtained by the division as the first score. Specifically, in FIG. 3, the first score is calculated as (c+d+e) / T1. That is, the larger the first score, the greater the degradation of the photoelectric pulse wave signal, and the smaller the first score, the smaller the degradation of the photoelectric pulse wave signal. Therefore, the condition for the predetermined threshold in the first score is satisfied when the first score is smaller than the preset threshold.

[0046] First score calculation function 31 may acquire the photoelectric pulse wave signal as a signal waveform image and calculate the first score using the acquired signal waveform image. In this case, first score calculation function 31 acquires the total time during which the photoelectric pulse wave signal was measured from the signal waveform image and the time during which the voltage value of the photoelectric pulse wave signal exceeded a preset upper voltage limit value as the image length (pixels). First score calculation function 31 executes the above-mentioned process based on the acquired image length and calculates the first score.

[0047] With the above-described configuration, first score calculation function 31 can calculate the degree of corruption of the photoplethysmographic signal as a first score.

[0048] Next, a procedure for calculating the second score will be described.

[0049] Second score calculation function 32 (controller 30) performs autocorrelation analysis based on the photoelectric pulse wave signal and the envelope derived from the photoelectric pulse wave signal. Second score calculation function 32 uses the autocorrelation coefficient calculated by the autocorrelation analysis to determine a score (second score) that evaluates the periodicity of the photoelectric pulse wave signal synchronized with the pulsation.

[0050] Fig. 4 is an explanatory diagram showing an example of a method for calculating the second score. Fig. 4A includes a photoelectric pulse wave signal and an envelope based on the photoelectric pulse wave signal. Fig. 4A shows a signal waveform portion (thick line area in Fig. 4A) that is less affected by external factors such as body movement and disturbance light, and a signal waveform portion (area other than the thick line) that is significantly affected by body movement, etc. In Fig. 4A, the photoelectric pulse wave signal is shown by a solid line, and the envelope based on the photoelectric pulse wave signal is shown by a dotted line. The vertical axis of the graph in Fig. 4A indicates the voltage value of the photoelectric pulse wave signal and the envelope based on the photoelectric pulse wave signal, and the horizontal axis of the graph indicates the measurement time.

[0051] The photoplethysmographic signal shows, as a signal waveform, the change in blood vessel volume that occurs when the heart pumps blood into the body. Therefore, in signal waveform portions that are less affected by external factors such as body movement and disturbance light, the photoplethysmographic signal shows periodicity synchronized with the pulsation. On the other hand, in signal waveform portions that are significantly affected by body movement and the like, the photoplethysmographic signal shows little periodicity synchronized with the pulsation because the signal waveform is affected by external factors and noise is added to the signal waveform, making it difficult for the photoplethysmographic signal to show periodicity synchronized with the pulsation.

[0052] Autocorrelation analysis is used as a method to evaluate whether the photoplethysmographic signal has a periodicity synchronized with the heartbeat. Autocorrelation analysis calculates the autocorrelation coefficient corresponding to a certain amount of time shift (hereinafter referred to as lag) from the original data.

[0053] The second score calculation function 32 calculates the sum of the autocorrelation coefficients for each lag. If the photoplethysmographic signal has periodicity synchronized with the pulsation, the sum of the autocorrelation coefficients will be large. Specifically, if the photoplethysmographic signal has periodicity synchronized with the pulsation, the autocorrelation coefficients for each lag will be large at a constant rate, and the sum of the autocorrelation coefficients will be large.

[0054] If the photoplethysmographic signal does not have a periodicity synchronized with the pulsation, the sum of the autocorrelation coefficients will be low. Specifically, if the photoplethysmographic signal does not have a periodicity synchronized with the pulsation, the autocorrelation coefficients for each lag will be generally low, and the sum of the autocorrelation coefficients will be small.

[0055] However, in a portion of the signal waveform that is significantly affected by body movement, the amplitude of the photoelectric pulse wave signal may change periodically in addition to the periodicity of the signal waveform itself. In such a case, the autocorrelation coefficient for each lag becomes large even though the photoelectric pulse wave signal does not have a periodicity synchronized with the pulsation. It becomes difficult for the control unit 30 to determine whether the photoelectric pulse wave signal has a periodicity synchronized with the pulsation by simply evaluating the sum of the calculated autocorrelation coefficients.

[0056] In order to address the above-mentioned problem, the second score calculation function 32 creates an envelope based on the photoelectric pulse wave signal shown in FIG. 4A. The envelope is generated, for example, by rectifying and smoothing the photoelectric pulse wave signal or by performing a Hilbert transform. The method of creating the envelope is not limited to this. The second score calculation function 32 extracts the voltage values ​​(f and g) of the photoelectric pulse wave signal and the envelope at a specific time (point t2 in FIG. 4A). The second score calculation function 32 divides the voltage value (f) of the photoelectric pulse wave signal at the specific time (point t2) by the voltage value (g) of the envelope. Through the above-mentioned process, the second score calculation function 32 calculates a value (hereinafter, referred to as a correction value) obtained by dividing the voltage value of the photoelectric pulse wave signal by the voltage value of the envelope. Specifically, at t2 in FIG. 4A, the correction value is calculated as f / g. The second score calculation function 32 calculates the correction value for each time during the period (T2) in which the photoelectric pulse wave signal was measured.

[0057] Fig. 4B is an example of a graph showing the correction value for each time when the photoelectric pulse wave signal is detected. The vertical axis of the graph in Fig. 4B shows the correction value, and the horizontal axis of the graph shows the measurement time. The envelope based on the photoelectric pulse wave signal is tangent to a part of the positive amplitude of the photoelectric pulse wave signal. Therefore, the voltage value of the envelope is always equal to or greater than the voltage value of the photoelectric pulse wave signal. The correction value calculated by dividing the voltage value of the envelope from the voltage value of the photoelectric pulse wave signal is always in the range of -1 to +1.

[0058] The second score calculation function 32 performs an autocorrelation analysis using the correction value for each time shown in Fig. 4B. Specifically, the second score calculation function 32 calculates the autocorrelation coefficient for each lag using the correction value for each time shown in Fig. 4B. The second score calculation function 32 calculates the sum of the autocorrelation coefficients for each lag (the second score). In this way, the second score calculation function 32 calculates the autocorrelation coefficient from the correction value obtained using the created envelope, thereby reducing the influence of external factors such as ambient light and body movement, making it easier to evaluate the periodicity of the photoplethysmographic signal synchronized with the pulsation.

[0059] When the photoplethysmographic signal has periodicity synchronized with the pulsation, the sum of the autocorrelation coefficients (second score) becomes large. On the other hand, when the photoplethysmographic signal does not have periodicity synchronized with the pulsation, the sum of the autocorrelation coefficients (second score) becomes small. Therefore, the second score satisfies the condition for a predetermined threshold value when the second score is larger than a preset threshold value.

[0060] With the above-described configuration, second score calculation function 32 calculates a correction value based on the photoplethysmographic signal and an envelope based on the photoplethysmographic signal. Second score calculation function 32 calculates an autocorrelation coefficient for each lag by performing an autocorrelation analysis based on the calculated correction value. Second score calculation function 32 calculates the sum of the autocorrelation coefficients for each lag (second score).

[0061] Next, a procedure for calculating the third score will be described.

[0062] 2, the photoelectric pulse wave signal acquired from the user includes signal waveform portions that are less affected by external factors such as body movement and disturbance light, and signal waveform portions that are significantly affected by body movement, etc. When a photoelectric pulse wave signal is acquired that includes many signal waveform portions that are less affected by external factors, control unit 30 can more easily extract the pulse rate, etc. from the photoelectric pulse wave signal.

[0063] The third score calculation function 33 (control unit 30) creates a pulse wave template used to extract a signal waveform portion that is less affected by external factors from the user's photoplethysmogram signal based on a photoplethysmogram signal with less noise (for example, a photoplethysmogram signal previously measured from another user) prepared in advance. The third score calculation function 33 separates the user's photoplethysmogram signal for each pulse. One pulse is, for example, a region sandwiched between two adjacent valleys in the photoplethysmogram signal. The third score calculation function 33 compares the template with the separated pulses. The third score calculation function 33 extracts pulses (hereinafter referred to as high-precision pulses) that have a high correlation coefficient with the template from the separated pulses through the above-mentioned comparison. The third score calculation function 33 calculates the high-precision pulse rate of all the separated pulses as the third score.

[0064] Explain how to create a template.

[0065] FIG. 5 is an explanatory diagram showing an example of a method for creating a template. FIG. 5A is an explanatory diagram showing a photoelectric pulse wave signal with less noise prepared in advance. The photoelectric pulse wave signal with less noise is extracted based on a certain criterion from a photoelectric pulse wave signal acquired in advance through a network. The certain criterion is set based on domain knowledge, medical viewpoints, and the like. The vertical and horizontal axes of the graph in FIG. 5A are the same as those in FIG. 2. The photoelectric pulse wave signal shown in FIG. 5A shows multiple peaks (F1, F2, and F3 in FIG. 5A) and valleys (G1, G2, and G3). A peak is, for example, a portion where the voltage (displacement) is the highest in the wavelength of the photoelectric pulse wave signal. A valley is, for example, a portion where the voltage (displacement) is the lowest between two adjacent peaks in the photoelectric pulse wave signal.

[0066] Third score calculation function 33 processes the photoelectric pulse wave signal shown in Fig. 5A using a bandpass filter. Third score calculation function 33 separates the photoelectric pulse wave signal for each pulse, with a region between two adjacent valleys (H1 and H2 in Fig. 5A) being one pulse. Third score calculation function 33 normalizes the separated pulses. For example, by normalization, the voltage value of the pulse is converted to 0 to 1, and the time at which the pulse was acquired is converted to 0 to 100.

[0067] FIG. 5B is an explanatory diagram showing a normalized pulse and a template. The vertical axis of the graph in FIG. 5B indicates the voltage value of the normalized pulse, and the horizontal axis of the graph indicates the time at which the normalized pulse is acquired. In FIG. 5B, the normalized pulses are indicated by dotted lines. In FIG. 5B, the template is indicated by a solid line. The third score calculation function 33 calculates the average value or median value of the pulses from the normalized pulses. The third score calculation function 33 creates a template using the average value or median value of the pulses.

[0068] Before creating a template, the third score calculation function 33 may exclude pulses that deviate from the average value, median value, etc., from among the normalized pulses. In that case, the third score calculation function 33 uses the interquartile range, etc., to identify pulses that deviate from the average value, median value, etc., from the normalized pulses, and excludes the identified pulses.

[0069] The third score calculation function 33 may create multiple templates according to time periods (e.g., daytime and nighttime). In this case, the third score calculation function 33 uses multiple templates to compare photoplethysmographic signals, making it easier to identify a photoplethysmographic signal that is less affected by external factors.

[0070] Fig. 6 is an explanatory diagram showing an example of a method for calculating the third score. Fig. 6 shows an example of a photoplethysmographic signal, which includes a signal waveform identified as a high-precision pulse (area indicated by a thick line in Fig. 6) and a signal waveform other than the high-precision pulse (area indicated by a non-thick line). The vertical and horizontal axes of the graph in Fig. 6 are the same as those in Fig. 2.

[0071] The third score calculation function 33 separates the photoelectric pulse wave signal of the user for each pulse in the same procedure as the creation of the template, and performs normalization processing on the separated pulses. The third score calculation function 33 calculates the degree of conformity between the template and the normalized pulse. Specifically, the third score calculation function 33 calculates the correlation coefficient between the template and the normalized pulse. The correlation coefficient is calculated using, for example, template matching, Euclidean distance, image matching, etc. The third score calculation function 33 extracts pulses having a correlation coefficient of a certain value or more (for example, 0.95) from all the normalized pulses as high-precision pulses. The third score calculation function 33 calculates the high-precision pulse rate of all the normalized pulses as the third score. Specifically, when the number of pulses identified as high-precision pulses is H and the number of all pulses per unit time is I, the third score is calculated as H / I. That is, the larger the third score is, the larger the number of high-precision pulses is, and the smaller the third score is, the smaller the number of high-precision pulses is. Therefore, the third score satisfies the condition for a predetermined threshold value when it is greater than a preset threshold value.

[0072] The third score calculation function 33 may extract an area identified as a high-precision pulse from the user's photoelectric pulse wave signal per unit time (T3 in FIG. 6) and measure the pulse rate, etc. from the extracted area. In this case, the third score calculation function 33 matches the user's photoelectric pulse wave signal with a template to extract an area (bold area) including a signal waveform that matches the template. The third score calculation function 33 stores the data area of ​​the extracted high-precision pulse in the storage unit 20. The control unit 30 measures the pulse rate, etc. from the area including the high-precision pulse stored in the storage unit 20. The measurement of the pulse rate, etc. may be performed by an external device instead of the server 100.

[0073] With the above-mentioned configuration, the third score calculation function 33 compares the pulses contained in the acquired photoelectric pulse wave with a template created in advance, and calculates a high-precision pulse rate (third score).

[0074] It is not necessary to calculate all of the first score, the second score, and the third score. In that case, the control unit 30 calculates only an arbitrary score from the first score, the second score, and the third score.

[0075] Next, a method for evaluating whether or not the pulse rate and the like can be extracted from the photoplethysmographic signal based on the first score, the second score, and the third score will be described.

[0076] The score evaluation function 34 determines whether or not the pulse rate and the like can be extracted from the acquired photoplethysmographic signal based on the first score, the second score, the third score, and a threshold value previously set for each score.

[0077] The thresholds preset for each score are set by a first score calculation function 31, a second score calculation function 32, and a third score calculation function 33. In addition, the control unit 30 may adjust the thresholds based on a medical viewpoint such as domain knowledge.

[0078] The first score is an index showing the degree to which the acquired photoplethysmographic signal is corrupted. The larger the first score, the more corrupted the photoplethysmographic signal is, and therefore the more difficult it is to extract the pulse rate and the like from the photoplethysmographic signal. Therefore, the score evaluation function 34 determines whether the calculated first score is equal to or less than a threshold value that is preset for the first score. In other words, when the first score is smaller than the preset threshold value, the score evaluation function 34 determines that the condition for the threshold value of the first score is satisfied.

[0079] The second score is an index showing whether or not the acquired photoelectric pulse wave signal has periodicity synchronized with the pulse. The smaller the second score, the less periodicity synchronized with the pulse in the photoelectric pulse wave signal, making it more difficult to extract the pulse rate and the like from the photoelectric pulse wave signal. Therefore, the score evaluation function 34 determines whether or not the calculated second score is equal to or greater than a threshold value that is preset for the second score. In other words, when the second score is greater than the preset threshold value, the score evaluation function 34 determines that the condition for the threshold value of the second score is satisfied.

[0080] The third score is an index showing the proportion of high-precision pulses contained in the acquired photoplethysmographic signal. The smaller the third score, the fewer the number of high-precision pulses contained in the photoplethysmographic signal, making it more difficult to extract the pulse rate and the like from the photoplethysmographic signal. Therefore, the score evaluation function 34 determines whether the calculated third score is equal to or greater than a threshold value that is preset for the third score. In other words, when the third score is greater than the preset threshold value, the score evaluation function 34 determines that the condition for the threshold value of the third score is satisfied.

[0081] When the first score, the second score, and the third score satisfy the conditions for the thresholds for each score, the score evaluation function 34 determines that the pulse rate and the like can be extracted from the acquired photoplethysmographic signal.

[0082] If at least one of the first score, second score, and third score does not satisfy the condition for the threshold value for each score, the score evaluation function 34 determines that the pulse rate, etc. cannot be extracted from the acquired photoplethysmographic signal.

[0083] The score evaluation function 34 may use only any one of the first score, the second score, and the third score. In that case, the score evaluation function 34 determines whether or not the pulse rate and the like can be extracted from the acquired photoplethysmographic signal based on the arbitrary score and a threshold value preset for each score.

[0084] 7 is a flowchart showing information processing of the photoelectric pulse wave signal accuracy score calculation system. Control unit 151 of wearable device 150 uses light-emitting unit 153 to irradiate LED light onto the user's arm or the like (step S101). Control unit 151 detects reflected light of the LED light irradiated onto the user via light-receiving unit 154 (step S102). Control unit 151 acquires a photoelectric pulse wave signal from the detected reflected light using data processing unit 155 (step S103). Control unit 151 transmits the acquired photoelectric pulse wave signal to server 100 via communication unit 156 (step S104).

[0085] The control unit 30 of the server 100 receives the photoplethysmographic signal transmitted by the wearable device 150 (step S105).

[0086] Control unit 30 starts a subroutine for calculating the first score using first score calculation function 31 (step S106). Specifically, first score calculation function 31 calculates the rate at which the acquired photoplethysmographic signal is corrupted.

[0087] The control unit 30 removes specific frequency band components contained in the received photoplethysmographic signal using existing techniques such as a low-pass filter, a high-pass filter, and a band-pass filter (step S107).

[0088] Control unit 30 uses second score calculation function 32 to start a subroutine for calculating the second score (step S108). Specifically, second score calculation function 32 calculates a correction value based on the photoplethysmographic signal and an envelope based on the photoplethysmographic signal, and performs autocorrelation analysis using the calculated correction value.

[0089] Control unit 30 uses third score calculation function 33 to start a subroutine for calculating the third score (step S109). Specifically, third score calculation function 33 compares a template created from a photoelectric pulse wave signal prepared in advance with the acquired photoelectric pulse wave signal to calculate a high-precision pulse rate.

[0090] Control unit 30 uses score calculation function 35 to start a subroutine for evaluating each score calculated by first score calculation function 31, second score calculation function 32, and third score calculation function 33 (step S110). Specifically, score calculation function 35 determines whether or not the pulse rate and the like can be extracted from the acquired photoplethysmographic signal based on the first score, second score, third score, and a threshold value set in advance for each score.

[0091] Based on the determination result of step S110, control unit 30 determines whether or not the pulse rate, etc. can be extracted from the photoelectric pulse wave signal (step S111). If it is determined that the pulse rate, etc. can be extracted from the acquired photoelectric pulse wave signal (step S111: YES), control unit 30 extracts the pulse rate, etc. from the photoelectric pulse wave signal (step S112). Control unit 30 stores the extracted pulse rate, etc. in memory unit 20 (step S113) and ends the process.

[0092] If it is determined that the pulse rate and the like cannot be extracted from the acquired photoelectric pulse wave signal (step S111: NO), control unit 30 returns the process to S101.

[0093] The order of the steps from calculating the first score (step S106) to calculating the third score (step S109) may be changed, and a plurality of steps may be performed in parallel.

[0094] FIG. 8 is a flow chart showing the calculation process of the first score. The subroutine for S106 will be described in detail. The first score calculation function 31 collates the acquired photoelectric pulse wave signal with a preset upper limit of voltage, and identifies the portion where the signal waveform becomes rectangular (C, D, and E in FIG. 3) (step S201). Specifically, the first score calculation function 31 identifies the region where a voltage value exceeding the preset upper limit of voltage is measured. The first score calculation function 31 calculates the time (c, d, and e) where the voltage value of the photoelectric pulse wave signal exceeds the upper limit of voltage based on the time when the photoelectric pulse wave signal is acquired (step S202). The first score calculation function 31 calculates the ratio of the portion where the signal waveform becomes rectangular to the measured photoelectric pulse wave signal as the first score (step S203). Specifically, the first score calculation function 31 divides the time (c, d, and e) where a value exceeding the preset upper limit of voltage is detected by the total time (T1) where the photoelectric pulse wave signal is acquired. First score calculation function 31 stores the calculated first score in storage unit 20 (step S204). After storing the first score in storage unit 20, control unit 30 returns the process to S107.

[0095] If the calculated first score is equal to or greater than a preset value, first score calculation function 31 may send an instruction to wearable device 150 to reduce the output of the LED light source.

[0096] If the calculated first score is equal to or greater than a preset value, first score calculation function 31 may determine that the acquired photoplethysmographic signal is unusable, and discontinue subsequent processing.

[0097] 9 is a flowchart showing the calculation process of the second score. The subroutine of S108 will be described in detail. The second score calculation function 32 generates an envelope based on the photoelectric pulse wave signal processed by the bandpass filter (step S301). The envelope is generated, for example, by rectifying and smoothing the photoelectric pulse wave signal or by performing a Hilbert transform.

[0098] The second score calculation function 32 extracts voltage values ​​at a specific time for the photoelectric pulse wave signal and the envelope (step S302). Specifically, the second score calculation function 32 extracts voltage values ​​(f and g) of the photoelectric pulse wave signal and the envelope at a specific time (point t2 in FIG. 4A). The second score calculation function 32 calculates a value obtained by dividing the voltage value of the photoelectric pulse wave signal by the voltage value of the envelope (hereinafter, referred to as a correction value) (step S303). Specifically, the second score calculation function 32 divides the voltage value (f) of the photoelectric pulse wave signal by the voltage value (g) of the envelope at the specific time (point t2). The second score calculation function 32 stores the voltage values ​​of the photoelectric pulse wave signal and the envelope, and the correction value in the storage unit 20 (step S304). The second score calculation function 32 judges whether the process from S302 to S304 is completed during the period (T2) during which the photoplethysmographic signal is measured (step S305). If the process from S302 to S304 is completed during the specific period (step S305: YES), the second score calculation function 32 calculates a correction value for each time using the correction value stored in the storage unit 20 (step S306). The second score calculation function 32 performs an autocorrelation analysis using the correction value for each time to calculate an autocorrelation coefficient for each lag (step S307). The second score calculation function 32 calculates the sum of the autocorrelation coefficients as the second score (step S308). The second score calculation function 32 stores the calculated second score in the storage unit 20 (step S309). After storing the second score in the storage unit 20, the control unit 30 returns the process to S109.

[0099] If the processes from S302 to S304 have not been completed within the specific period (step S305: NO), second score calculation function 32 returns the process to step S302.

[0100] If the calculated second score is equal to or less than a preset value, second score calculation function 32 may determine that the acquired photoplethysmographic signal is unusable, and terminate subsequent operations.

[0101] 10 is a flowchart showing the calculation process of the third score. The subroutine related to step S109 will be described in detail. Third score calculation function 33 separates the photoelectric pulse wave signal into pulses (step S401). Specifically, third score calculation function 33 separates the photoelectric pulse wave signal into pulses, with a region between two adjacent valleys in the photoelectric pulse wave signal acquired from the user being one pulse. Third score calculation function 33 performs normalization processing on the separated pulses (step S402).

[0102] Control unit 30 uses third score calculation function 33 to start a subroutine for creating a template (step S403). Specifically, third score calculation function 33 creates a template using a low-noise photoplethysmographic signal that is prepared in advance.

[0103] The third score calculation function 33 calculates a correlation coefficient between the normalized pulse and the created template (step S404). The correlation coefficient is calculated using, for example, template matching, Euclidean distance, image matching, etc. The third score calculation function 33 specifies, among the normalized pulses, a pulse having a correlation coefficient equal to or greater than a certain value (for example, 0.95) as a high-precision pulse (step S405). The third score calculation function 33 determines whether or not the processes from S404 to S405 are completed for all normalized pulses (step S406). When it is determined that the processes from S404 to S405 are completed for all normalized pulses (step S406: YES), the third score calculation function 33 calculates a high-precision pulse rate among all normalized pulses as a third score (step S407). The control unit 30 stores the third score in the storage unit 20 (step S408). After storing the third score in the storage unit 20, the control unit 30 returns the process to S110.

[0104] When it is determined that the processes from S404 to S405 have not been completed for all pulses that have been normalized (step S406: NO), third score calculation function 33 returns the process to step S404.

[0105] If the calculated third score is equal to or less than a preset value, the third score calculation function 33 may determine that the acquired photoplethysmogram is unusable, and terminate the subsequent operations.

[0106] FIG. 11 is a flowchart showing a method for creating a template. The subroutine for step S403 will be described in detail. The third score calculation function 33 acquires a photoelectric pulse wave signal with less noise from the storage unit 20 (step S501). For example, the control unit 30 acquires a photoelectric pulse wave signal with less noise in advance through a network, processes the acquired photoelectric pulse wave signal with a bandpass filter, and stores the processed photoelectric pulse wave signal in the storage unit 20. The processing with the bandpass filter may be performed immediately before creating the template. The third score calculation function 33 extracts a portion (e.g., three samples) from the acquired multiple photoelectric pulse wave signals based on a certain criterion (step S502). The certain criterion is set based on domain knowledge, medical viewpoints, and the like. The third score calculation function 33 separates the extracted photoelectric pulse wave signal for each pulse (step S503). The method for separating for each pulse is the same as that of S401. The third score calculation function 33 superimposes the separated pulses to calculate an average value, a median value, or the like (step S504). Third score calculation function 33 creates a template from the separated pulses using the calculated average value or median value, etc. (step S505). Third score calculation function 33 performs normalization processing on the created template (step S506). For example, by normalization, the voltage value of the pulse is converted to 0 to 1, and the time at which the pulse is acquired is converted to 0 to 100. Third score calculation function 33 stores the normalized template in storage unit 20 (step S507). After storing the template in storage unit 20, control unit 30 returns the process to step S404.

[0107] The subroutine for creating the template may be executed before calculating the third score.

[0108] FIG. 12 is a flowchart showing the method of evaluating each score. The subroutine related to step S110 will be described in detail. The score evaluation function 34 extracts the first score, the second score, the third score, and the threshold value for each score from the storage unit 20 (step S601). The score evaluation function 34 judges whether the first score is equal to or less than a preset threshold value (step S602). If it is judged that the first score is equal to or less than the threshold value (step S602: YES), the score evaluation function 34 judges whether the second score is equal to or more than a preset threshold value (step S603). If it is judged that the second score is equal to or more than the threshold value (step S603: YES), the score evaluation function 34 judges whether the third score is equal to or more than a preset threshold value (step S604). If it is judged that the third score is equal to or more than the threshold value (step S604: YES), the score evaluation function 34 judges that the pulse rate and the like can be extracted from the acquired photoplethysmographic signal (step S605). The control unit 30 stores the above-mentioned determination result in the storage unit 20 (step S606). After storing the template in the storage unit 20, the control unit 30 returns the process to step S111.

[0109] If it is determined that the first score is equal to or greater than the threshold value (step S602: NO), control unit 30 determines that the pulse rate and the like cannot be extracted from the acquired photoelectric pulse wave signal (step S607). Control unit 30 returns the process to step S111.

[0110] If it is determined that the second score is equal to or less than the threshold value (step S603: NO), control unit 30 determines that the pulse rate and the like cannot be extracted from the acquired photoelectric pulse wave signal (step S607). Control unit 30 returns the process to step S111.

[0111] If it is determined that the third score is equal to or less than the threshold value (step S604: NO), control unit 30 determines that the pulse rate and the like cannot be extracted from the acquired photoelectric pulse wave signal (step S607). Control unit 30 returns the process to step S111.

[0112] It should be noted that the score assessment function 34 can change the order of the processes in steps S602 to S604.

[0113] If it is determined that the pulse rate and the like can be extracted from the photoelectric pulse wave signal, the control unit 30 may link the original data of the photoelectric pulse wave signal used for the first score, the second score, and the third score, and the determination result, and store them in the storage unit 20. In this case, the control unit 30 can share the original data of the photoelectric pulse wave signal and the determination result with an external analysis device via the communication unit 10.

[0114] It should be noted that the score evaluation function 34 may determine whether or not the pulse rate, etc., can be extracted from the photoplethysmographic signal using only an arbitrary score, without using all of the first score, second score, and third score.

[0115] With the above-described configuration, score evaluation function 34 can determine whether or not the pulse rate and the like can be extracted from the photoplethysmographic signal based on the first score, the second score, the third score, and the threshold values ​​preset for each score. Control unit 30 can use score evaluation function 34 to extract a portion of the photoplethysmographic signal where the pulse wave is appropriately measured.

[0116] The embodiments disclosed herein are illustrative in all respects and should not be considered as limiting. The scope of the present invention is defined by the claims, not by the above meaning, and is intended to include all modifications within the scope and meaning equivalent to the claims.

[0117] The matters described in each embodiment may be combined with each other. In addition, the independent claims and dependent claims described in the claims may be combined with each other in any and all combinations regardless of the citation format. Furthermore, the claims use a format in which a claim cites two or more other claims (multiple claim format), but this is not limited to this. The claims may also be written in a format in which a multiple claim cites at least one other multiple claim (multi-multi claim). [Explanation of symbols]

[0118] 1: Photoplethysmography signal accuracy score calculation system 1a:Storage medium 100: Server 10: Communications Department 20: Storage section 30: Control section 31: First score calculation function 32: Second score calculation function 33: Third score calculation function 34: Score evaluation function 150: Wearable devices 151: Control unit 152: Storage section 152P: Control program 153: Light emitting part 154: Light receiving part 155: Data processing section 156: Communications Department

Claims

1. A photoplethysmographic signal is obtained by detecting the photoplethysmographic signal. Calculating a score related to the photoplethysmogram waveform based on the photoplethysmogram signal; Data of a portion of the photoplethysmographic signal where the score satisfies a condition relative to a predetermined threshold is extracted. A program that causes a computer to carry out processing.

2. calculating a first score representing a degree of waveform distortion depending on whether the photoelectric pulse wave signal exceeds an intensity threshold of the received light intensity; Extracting data that satisfies a condition for the threshold based on the first score. The program according to claim 1.

3. generating an image of a signal waveform from the photoplethysmographic signal; The first score is calculated by determining a degree of waveform distortion according to whether or not the photoelectric pulse wave signal exceeds an intensity threshold of the received light intensity based on the image. The program according to claim 2.

4. calculating a second score for evaluating the periodicity of the photoplethysmographic waveform based on an autocorrelation analysis of the photoplethysmographic signal; Extracting data that satisfies a condition for the threshold based on the second score. The program according to claim 1.

5. Deriving an envelope of the photoplethysmographic signal; The second score is calculated based on an autocorrelation analysis of waveform data obtained by dividing the photoplethysmographic signal by the envelope. The program according to claim 4.

6. extracting a waveform portion from the photoplethysmogram signal that matches a predetermined photoplethysmogram waveform template; calculating a third score representing a degree of matching with the template according to the number of the extracted waveform portions; Extracting data that satisfies a condition for the threshold based on the third score. The program according to claim 1.

7. A photoplethysmographic signal is obtained by detecting the photoplethysmographic signal. Calculating a score related to the photoplethysmogram waveform based on the photoplethysmogram signal; Data of a portion of the photoplethysmographic signal where the score satisfies a condition relative to a predetermined threshold is extracted. An information processing method in which processing is performed by a computer.

8. An information processing device including a control unit, The control unit: A photoplethysmographic signal is obtained by detecting the photoplethysmographic signal. Calculating a score related to the photoplethysmogram waveform based on the photoplethysmogram signal; Data of a portion of the photoplethysmographic signal where the score satisfies a condition relative to a predetermined threshold is extracted. Information processing device.

9. A photoplethysmographic signal is obtained by detecting the photoplethysmographic signal. A template of a predetermined photoelectric pulse waveform is matched with the signal waveform indicated by the photoelectric pulse wave signal, and data of a waveform portion that matches the template is extracted from the photoelectric pulse wave signal. A program that causes a computer to carry out processing.

Images