Measuring device and measuring method

Abstract

The present invention provides a measuring device and a measuring method that can obtain a signal of interest with high accuracy from a measurement signal containing both the signal of interest and noise. [Solution] A measuring device comprising: a measurement sensor unit having a plurality of measurement sensors for measuring a measurement signal in which a signal of interest and noise are mixed; a signal separation unit that performs signal separation on the data of the measurement signal measured by the plurality of measurement sensors; a component selection unit that compares each of the plurality of components obtained by the signal separation with a template waveform of the signal of interest and selects the component that is considered to contain the signal of interest; and a component reconstruction unit that reconstructs data for each of the plurality of measurement sensors based on the component selected by the component selection unit.

Description

Technical Field

[0001] The present disclosure relates to a measuring device and a measuring method.

Background Art

[0002] In order to obtain good analysis results in various measurements, it is necessary to remove various noises superimposed on a signal of interest, which is a desired signal. Such noises include, for example, environmental noise depending on the environment, sensor noise depending on the sensor, and noise generated by the measurement target itself. Generally, a method of removing noise by filtering processing is used. However, in this method, for example, when there is a bias in the position of the noise in the time series, or when the frequency bands of the signal of interest and the noise overlap, it may be difficult to completely remove the noise.

[0003] In the biological information monitoring system described in Patent Document 1, a template generated based on the peak of the heartbeat component is used (see Patent Document 1). Specifically, in the biological information monitoring system described in Patent Document 1, a signal indicating the physical quantity is acquired from a plurality of sensors installed on a member that supports a subject and detects a mechanical physical quantity, and based on the peak of the signal indicating the acquired physical quantity, biological information regarding the heartbeat of the subject supported by the member is acquired. Then, in the biological information monitoring system, signal separation is performed by performing component analysis processing on the signal indicating the physical quantity, the heartbeat component is specified from the separated signals, and biological information regarding the heartbeat is acquired based on the matching result between the template generated based on the peak of the specified heartbeat component and the specified heartbeat component.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

[0005] However, the conventional technology described in Patent Document 1 sometimes resulted in insufficient accuracy in the acquired biological information. For example, in the conventional technology described in Patent Document 1, a template is generated using the signal separation result obtained from the signal acquired by the sensor. Therefore, if the signal-to-noise ratio (SNR) of the components after signal separation is poor, an unsuitable template is generated.

[0006] This disclosure has been made in consideration of these circumstances, and aims to provide a measuring device and a measuring method that can obtain a signal of interest with high accuracy from a measurement signal in which the signal of interest and noise are mixed. [Means for solving the problem]

[0007] One embodiment is a measuring device comprising: a measurement sensor unit having a plurality of measurement sensors for measuring a measurement signal in which a signal of interest and noise are mixed; a signal separation unit that performs signal separation on the data of the measurement signal measured by the plurality of measurement sensors; a component selection unit that compares each of the plurality of components obtained by the signal separation with a template waveform of the signal of interest and selects the component that is considered to contain the signal of interest; and a component reconstruction unit that reconstructs data for each of the plurality of measurement sensors based on the component selected by the component selection unit.

[0008] One embodiment is a measurement method comprising: measuring a measurement signal using the measurement sensors of a measurement sensor unit having a plurality of measurement sensors that measure a measurement signal containing a signal of interest and noise; performing signal separation on the data of the measurement signal measured by the plurality of measurement sensors using a signal separation unit; comparing each of the plurality of components obtained by the signal separation with a template waveform of the signal of interest using a component selection unit to select the component that is considered to contain the signal of interest; and reconstructing the data for each of the plurality of measurement sensors using a component reconstruction unit based on the component selected by the component selection unit. [Effects of the Invention]

[0009] According to this disclosure, the measuring device and measuring method make it possible to obtain a signal of interest with high accuracy from a measurement signal in which the signal of interest and noise are mixed. [Brief explanation of the drawing]

[0010] [Figure 1] This figure shows a schematic configuration of the measuring device according to the embodiment. [Figure 2] This diagram shows the configuration of the measurement sensor unit according to the embodiment. [Figure 3] This figure shows an example of stored data in the template waveform storage unit according to the embodiment. [Figure 4] This figure shows an example of the processing flow performed by the measuring device according to the embodiment. [Figure 5A] This figure shows an example of the waveform of the 42-channel measurement signal according to the embodiment. [Figure 5B] This figure shows an example of the waveform of the measurement signal for channel 1 (ch.1) according to the embodiment. [Figure 6A] This figure shows an example of the waveforms of the 42 components after signal separation according to the embodiment. [Figure 6B] This figure shows an example of the waveform of the first component signal according to the embodiment. [Figure 7] This figure shows an example of a template waveform according to the embodiment. [Figure 8] It is a diagram showing an example of the cross-correlation coefficient between the waveform of a component signal and a template waveform for some of the 42 components according to the embodiment. [Figure 9] It is a diagram showing an example of some of the components selected from the 42 components according to the embodiment. [Figure 10A] It is a diagram showing an example of the waveform of the reconstructed signal of 42 channels according to the embodiment. [Figure 10B] It is a diagram showing an example of the waveform of the reconstructed signal of channel 27 (ch. 27) according to the embodiment.

Mode for Carrying Out the Invention

[0011] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings.

[0012] [Measuring Device] FIG. 1 is a diagram showing a schematic configuration of a measuring device 1 according to the embodiment. Also, FIG. 1 shows a subject 51, who is a human being to be measured by the measuring device 1.

[0013] The measuring device 1 includes a measurement sensor unit 11 and an information processing unit 21. The information processing unit 21 includes an input unit 111, an output unit 112, a storage unit 113, and a control unit 114. The input unit 111 includes an acquisition unit 131. The output unit 112 includes a display unit 141. The storage unit 11 contains a template waveform storage unit A1. The control unit 114 includes a processing unit 151 and a display control unit 152. The processing unit 151 includes a signal separation unit 161, a component selection unit 162, and a component reconstruction unit 163. Note that the processing unit 151 may be referred to as an arithmetic unit or the like.

[0014] In the example shown in Figure 1, the measurement sensor unit 11 and the information processing unit 21 are configured as a single unit, but in other examples, the measurement sensor unit 11 and the information processing unit 21 may be configured as separate units. For example, the measurement sensor unit 11 and the information processing unit 21 may be separate components, and the information processing unit 21 may be configured using a computer.

[0015] <Measurement Sensor Unit> Figure 2 shows the configuration of the measurement sensor unit 11 according to the embodiment. For the sake of explanation, Figure 2 shows the XY Cartesian coordinate system, which is a two-dimensional Cartesian coordinate system. The measurement sensor unit 11 has multiple measurement sensors. In this embodiment, the measurement sensor unit 11 has 42 measurement sensors B1 to B42. In this embodiment, for the sake of explanation, each of the 42 measurement sensors B1 to B42 is also referred to as a channel (ch). In other words, in this embodiment, there are 42 channels. Multiple measuring sensors may be referred to as, for example, a group of measuring sensors.

[0016] In the example shown in Figure 2, the 42 measurement sensors B1 to B42 are arranged in an array in the XY plane. In this example, 42 measurement sensors B1 to B42 are arranged in an array at equal intervals in one direction (parallel to the X-axis) and also at equal intervals in a direction perpendicular to that direction (parallel to the Y-axis). In this embodiment, for the sake of explanation, arrangements parallel to the X-axis are referred to as rows, and arrangements parallel to the Y-axis are referred to as columns.

[0017] Specifically, six measurement sensors B1 to B6 are arranged at equal intervals parallel to the X-axis, the next six measurement sensors B7 to B12 are arranged at equal intervals parallel to the X-axis, and so on, with six measurement sensors arranged at equal intervals parallel to the X-axis in subsequent rows. In other words, at the end of this sequence, six measurement sensors B37 to B42 are arranged at equal intervals parallel to the X-axis. In this example, the six measurement sensors are arranged sequentially from the negative to the positive direction of the X-axis.

[0018] Furthermore, the rows of six measurement sensors, arranged parallel to the X-axis, are sequentially spaced equally in the direction parallel to the Y-axis. Specifically, seven measurement sensors B1, B7, B13, B19, B25, B31, and B37 are arranged at equal intervals parallel to the Y-axis, followed by seven more measurement sensors B2, B8, B14, B20, B26, B32, and B38, also arranged at equal intervals parallel to the Y-axis, and so on. In other words, at the end of this sequence, seven measurement sensors B6, B12, B18, B24, B30, B36, and B42 are arranged at equal intervals parallel to the Y-axis. In this example, the seven measurement sensors are arranged sequentially from the negative to the positive direction of the Y-axis. In other words, these rows of seven measurement sensors, arranged parallel to the Y-axis, are then arranged at equal intervals parallel to the X-axis.

[0019] In the example in Figure 2, the equal intervals in the direction parallel to the X-axis and the equal intervals in the direction parallel to the Y-axis are the same, but in other examples, these intervals may be different. Here, there are no particular limitations on the arrangement (overall arrangement) of the measurement sensors B1 to B42, and other configurations may be used. Furthermore, there are no particular limitations on the arrangement (e.g., position, direction) of each measurement sensor B1 to B42, and various configurations may be used. For example, in the example in Figure 2, the direction of all measurement sensors B1 to B42 is shown to be the same, but the direction of each measurement sensor B1 to B42 may be arbitrary.

[0020] Furthermore, in this embodiment, a case is shown in which 42 measurement sensors B1 to B42 are used as multiple measurement sensors, but theoretically the total number of multiple measurement sensors can be 2 or more, and may be any number. The term "measurement sensor" is merely a descriptive term; it may also be called by other names, such as "signal sensor" or simply "sensor." Furthermore, the array-like structure may also be called, for example, a matrix-like structure.

[0021] In this embodiment, a case is shown where multiple measurement sensors are configured as an integrated unit (measurement sensor unit 11), but as another example, a configuration in which multiple measurement sensors are provided separately is also possible.

[0022] <Placement of objects to be measured> In the example shown in Figure 2, for example, the body parts of the subject 51 to be measured are positioned at the center (or near the center) of the 42 measurement sensors B1 to B42. In this embodiment, the body part in question is the heart. Here, there are no particular limitations on the arrangement of the object to be measured and the measurement sensors B1 to B42, and other configurations may be used.

[0023] <Measurement target> There are no particular limitations on what can be measured; a variety of things are acceptable. In this embodiment, the object of measurement is the heart of the subject 51, but other body parts, such as the brain of the subject 51, may also be used as the object of measurement. Furthermore, the object of measurement may be a living organism other than a human. Furthermore, the object to be measured may not be a living organism. For example, the object to be measured may be a coil that generates magnetism, in which case it is possible to suppress the influence of noise when measuring the magnetism generated from the coil. As another example, a sound source that emits sound (sound signals) may be used as the object to be measured.

[0024] <Signals of Interest> Each of the measurement sensors B1 to B42 measures a desired signal (referred to as the signal of interest for convenience of explanation) originating from the object being measured. In this process, noise is superimposed on the signal of interest. Therefore, each of the measurement sensors B1 to B42 measures a signal (referred to as the measurement signal for convenience of explanation) that contains both the signal of interest and the noise. Then, each of the measurement sensors B1 to B42 acquires the measurement result. Any physical quantity signal may be used as the signal of interest, such as magnetic fields, electric currents, voltages, light, forces like pressure, sound, ultrasound, etc.

[0025] In this embodiment, since a template waveform of the signal of interest is used, it is preferable, for example, that the template waveform be applied to a valid signal of interest. Specifically, it may be applied to a signal of interest in which the same waveform (template waveform) is repeated. The signal of interest may also be called, for example, the target signal or the desired signal. Furthermore, measurement may also be called, for example, detection, measurement, or sensing.

[0026] <Noise> Examples of noise include environmental noise, which depends on the environment; sensor noise, which depends on the measurement sensors B1 to B42; and noise emitted by the object being measured itself. Environmental noise includes, for example, noise generated from noise sources present in the vicinity of the object being measured. Environmental noise signals may also be called, for example, interference signals. Sensor noise can include, for example, noise generated from the components or connecting cables of each measurement sensor B1 to B42. Noise emitted by the object being measured itself includes, for example, signals generated from areas other than the area of ​​interest.

[0027] <Types of measuring sensors> In this embodiment, all of the multiple measuring sensors B1 to B42 are of the same type, that is, sensors that measure the same physical quantity. However, as another example, the multiple measuring sensors B1 to B42 may include sensors of different types, that is, sensors that measure different physical quantities. Furthermore, even if two sensors measure the same physical quantity, they may be considered different types of sensors if they are different products with different ratings.

[0028] As a specific example, measurement sensors B1 to B42 may include one or more types of sensors such as magnetic sensors, potential sensors, light sensors, pressure sensors, acceleration sensors, vibration sensors, microphones, or ultrasonic sensors. This embodiment shows a case where all 42 measuring sensors B1 to B42 are magnetic sensors.

[0029] <Connection between the information processing unit and each measurement sensor> In this embodiment, the information processing unit 21 and the respective measurement sensors B1 to B42 are connected to each other via wired or wireless means so that they can communicate with each other. The information processing unit 21 is then capable of acquiring the measurement results from each of the measurement sensors B1 to B42. Note that in the examples shown in Figures 1 and 2, the details of the connection between the information processing unit 21 and the respective measurement sensors B1 to B42 are omitted from the illustration.

[0030] In this embodiment, the information processing unit 21 communicates with each of the measurement sensors B1 to B42 to obtain the measurement results from each of the measurement sensors B1 to B42. However, as another example, a configuration may be used in which the measurement results from each of the measurement sensors B1 to B42 are first stored in a portable storage medium, and then the measurement results are output from the storage medium to the information processing unit 21, allowing the information processing unit 21 to obtain the measurement results.

[0031] <Information Processing Department> The information processing unit 21 will now be described. The input unit 111 receives input from outside the information processing unit 21. In this embodiment, the input unit 111 receives signals (measurement result signals) output from each of the measurement sensors B1 to B42. Specifically, the input unit 111 may receive the signals transmitted from each of the measurement sensors B1 to B42, or it may receive signals stored in a portable storage device from that storage device. Furthermore, the input unit 111 may have, for example, an operating unit operated by a user, and information corresponding to the operation performed by the user may be input to the operating unit.

[0032] For the sake of explanation, it has been described here that the multiple measurement sensors B1 to B42 are located outside the information processing unit 21. However, the multiple measurement sensors B1 to B42 and the information processing unit 21 do not necessarily need to be clearly distinguished as being inside or outside the unit.

[0033] The acquisition unit 131 acquires the signal input by the input unit 111. The acquisition unit 131 may store the acquired signal in the storage unit 113. In this case, if the signal input by the input unit 111 is an analog signal, the acquisition unit 131 may, for example, be equipped with an A / D (Analog to Digital) conversion function to convert the signal from an analog signal to a digital signal. Furthermore, when the information processing unit 21 is applied to real-time processing, the acquisition unit 131 acquires signals in real time. Even when the information processing unit 21 is not applied to real-time processing, the acquisition unit 131 may still acquire signals in real time.

[0034] The output unit 112 outputs to the outside. The display unit 141 displays and outputs information related to the signal processing results. The display unit 141 has a screen such as a liquid crystal display (LCD), and displays and outputs information related to the signal processing results on this screen. In another configuration example, the display unit 141 may print and output the information related to the signal processing results onto paper. The output unit 112 may also have a function to output in other ways, such as audio output.

[0035] The memory unit 113 has a storage device such as a memory, and stores data. The memory unit 113 stores data such as the input signal and the processing result of the signal. Furthermore, the memory unit 113 stores, for example, control programs.

[0036] In this embodiment, the storage unit 113 includes a template waveform storage unit A1. The template waveform storage unit A1 stores data for one or more template waveforms. In this embodiment, for the sake of explanation, the storage area in the storage unit 113 that stores the template waveform is distinguished and shown as the template waveform storage unit A1, but the template waveform storage unit A1 does not necessarily have to be clearly distinguished in the storage unit 113. Data may also be referred to as information, for example.

[0037] The control unit 114 performs various processes and controls. In this embodiment, the control unit 114 has a processor such as a CPU (Central Processing Unit), and the processor executes a control program stored in the storage unit 113 to perform various processes and various controls.

[0038] The processing unit 151 performs predetermined processing based on the measurement result signals from the measurement sensors B1 to B42. The display control unit 152 outputs various types of information to the screen of the display unit 141.

[0039] The following describes the predetermined processing performed by the processing unit 151. In this embodiment, the predetermined process is a process of acquiring a signal of interest from a mixed signal of the signal of interest and noise, and corresponds to a process of removing the noise. Here, the degree to which the noise is removed from the mixed signal can be any degree that is practically effective. In other words, it is not necessary to completely remove the noise component from the mixed signal, and a method of removing the noise component from the mixed signal to the necessary extent may be used. The process of removing noise may also be referred to as noise reduction or noise suppression.

[0040] The signal separation unit 161 applies a predetermined signal separation process to the data of the 42 measurement result signals acquired by the 42 measurement sensors B1 to B42. In this embodiment, blind source separation (BSS) is used as the signal separation process. Blind source separation is generally a technique for separating individual components (signal components) from multiple measurement sequences obtained by mixing multiple unknown signal sequences in an unknown linear mixed system.

[0041] Here, one or more of the following methods may be used for blind source separation: Singular Value Decomposition (SVD), Principal Component Analysis (PCA), Independent Component Analysis (ICA), and Nonnegative Matrix Factorization (NMF).

[0042] In this embodiment, the signal separation unit 161 performs signal separation processing on the measurement result data of 42 measurement sensors B1 to B42, thereby separating 42 components, which is the same number as the number of measurement results.

[0043] The component selection unit 162 compares the 42 components obtained by the signal separation unit 161 after signal separation with a pre-prepared template waveform of the signal of interest, and selects from these 42 components the component that is predicted to contain the signal of interest. In this embodiment, the component selection unit 162 obtains a value representing the degree of correlation between each component and the template waveform, and selects components whose values ​​exceed a predetermined value as components that are predicted to contain the signal of interest.

[0044] In this embodiment, the value representing the degree of correlation takes an absolute value between 0 and 1, and it is assumed that the closer the value is to 1, the higher the degree of correlation. Therefore, this predetermined value is a positive value. As another example, the signs of the degree of correlation and the predetermined value may be reversed. The degree of correlation may also be called, for example, correlation score or similarity score. Conversely, the degree of difference may be used as the opposite concept to the degree of correlation; in this case, a lower degree of difference corresponds to a higher degree of correlation. Furthermore, this predetermined value may be called, for example, a threshold value.

[0045] In the presence of noise, typically between one and 42 components are selected; for example, only a few components may be selected. While in an ideal situation without noise, all components may be selected, this embodiment assumes a situation where noise is present.

[0046] Here, we present an example of a method for selecting a component from among several components that is expected to contain the signal of interest. The first component selection method involves calculating the cross-correlation coefficient between each of the multiple components after signal separation and the template waveform, and selecting components whose cross-correlation coefficient exceeds a predetermined value as components that are predicted to contain the signal of interest. This method allows for the effective selection of components containing the signal of interest, or in other words, the effective rejection of components containing significant unwanted noise, resulting in the extraction of high-quality signals of interest.

[0047] The second component selection method involves adjusting the scale of each of the multiple components after signal separation using a template waveform as a reference, calculating a value representing the magnitude of the error based on the difference between the scaled component and the template waveform, and selecting components whose value is less than a predetermined value as components that are predicted to contain the signal of interest.

[0048] Here, scale adjustment does not necessarily have to be performed on all of the multiple components after signal separation. For example, for some components that are determined not to require scale adjustment, the magnitude of the error based on the difference between those components and the template waveform may be determined without scale adjustment.

[0049] Furthermore, various values ​​(evaluation values) may be used to represent the magnitude of the error. For example, the sum of the absolute values ​​of the errors at multiple sample points between the waveform of each component and the template waveform may be used, the sum of the squares of the errors at multiple sample points between the waveform of each component and the template waveform may be used, or the maximum value among the absolute values ​​of the errors at multiple sample points between the waveform of each component and the template waveform may be used.

[0050] The third component selection method involves the user visually selecting, from among multiple components after signal separation, a component that is perceived to contain the signal of interest based on the template waveform, as the component predicted to contain the signal of interest. In this case, for example, the measuring device 1 displays the waveforms of multiple components after signal separation and a template waveform to the user via the display unit 141, and accepts the user's specification of components via the operation unit of the input unit 111. The component selection unit 162 then selects the specified components as components that are expected to contain the signal of interest.

[0051] The component reconstruction unit 163 reconstructs the data of the signal of interest for each of the 42 measurement sensors B1 to B42 based on one or more components selected by the component selection unit 162. In this embodiment, this reconstruction corresponds to dimensionality reduction. In this reconstruction, only the components selected by the component selection unit 162 are used, and the components that were not selected are not used.

[0052] Figure 3 shows an example of data stored in the template waveform storage unit A1 according to the embodiment. The template waveform storage unit A1 stores one or more template waveforms for the signal of interest. In this embodiment, the template waveform storage unit A1 stores m template waveforms, specifically the first template waveform C1 to the mth template waveform Cm (where m is an integer greater than or equal to 1, and in the example in Figure 3, it is an integer greater than or equal to 3).

[0053] Here, the first template waveform C1 to the mth template waveform Cm are all different waveforms. These distinct waveforms may be used, for example, to compensate for individual differences (which may be called individual variations in the case of humans) that may appear in the signal of interest. For example, if the signal of interest is heart rate, then there are individual differences in heart rate.

[0054] For example, the first template waveform C1 to the mth template waveform Cm may each be a waveform acquired by at least one of the measurement sensors B1 to B42 of the same type. For example, all of the measurement sensors B1 to B42 may be of the same type, and all of the first template waveforms C1 to the mth template waveforms Cm may be waveforms acquired by sensors of the same type as the measurement sensors B1 to B42.

[0055] As a specific example, all of the measurement sensors B1 to B42 may be magnetic sensors, and all of the first template waveforms C1 to the mth template waveform Cm may be waveforms acquired by the magnetic sensors. In this case, the average of waveforms measured multiple times by a single magnetic sensor may be used as the template waveform, which can improve the accuracy of the template waveform.

[0056] As another example, the first template waveform C1 to the mth template waveform Cm may each be waveforms acquired by sensors of a different type than any of the measurement sensors B1 to B42. For example, all of the measurement sensors B1 to B42 may be of the same type, and all of the first template waveforms C1 to the mth template waveforms Cm may be waveforms acquired by sensors of a different type than the measurement sensors B1 to B42, and furthermore, all of the first template waveforms C1 to the mth template waveforms Cm may be waveforms acquired by sensors of the same type.

[0057] As a specific example, all of the measurement sensors B1 to B42 may be magnetic sensors, and all of the first template waveforms C1 to the mth template waveform Cm may be waveforms acquired by an electrocardiogram sensor. Generally, electrocardiogram waveforms have a better signal-to-noise ratio (SNR) than measurement results from magnetic sensors. Therefore, using an electrocardiogram waveform as a template waveform can sometimes improve the accuracy of the template waveform.

[0058] Furthermore, the first template waveform C1 to the mth template waveform Cm may each be generated based on signals measured on the same subject as the measurement target (in the example of Figure 1, the subject 51), or they may be generated based on signals measured on a different subject than the measurement target (in the example of Figure 1, the subject 51). As a specific example, when the heart rate of subject 51 is measured, a template waveform may be generated in advance based on the measurement results of subject 51's (i.e., one's own) heart rate, or a template waveform may be generated based on the measurement results of the heart rate of someone other than subject 51 (i.e., another person). For example, the other person's heart rate may be normal, while subject 51's heart rate may be irregular.

[0059] Here, the first template waveform C1 to the mth template waveform Cm may each be generated, for example, by performing a theoretical simulation. Furthermore, while the example in Figure 3 shows a case where multiple template waveforms (first template waveform C1 to the mth template waveform Cm) are stored in the template waveform storage unit A1, for example, in a configuration where only one template waveform is used, only one template waveform may be stored in the template waveform storage unit A1.

[0060] [Example of a processing flow performed by a measuring device] Figure 4 shows an example of the processing flow performed by the measuring device 1 according to the embodiment. This example shows the case where one template waveform b1 is used. In this example, the following processes are performed in order: measurement process T1, signal separation process T2, component selection process T3, and component reconstruction process T4.

[0061] In measurement process T1, the signal to be measured a1 is measured by 42 measurement sensors B1 to B42. This yields the measurement signal a2. In the example shown in Figure 4, the signals to be measured by each of the 42 measurement sensors B1 to B42 are collectively shown as the measured signal a1. However, the measured signals of each measurement sensor B1 to B42 may be different. In other words, the measured signal of interest and the noise conditions may differ for each of the measurement sensors B1 to B42. The measured signals from each of the measurement sensors B1 to B42 may include, for example, signals of interest and noise, but may also not include signals of interest, or may not include noise.

[0062] In the example in Figure 4, the signals measured by each of the 42 measurement sensors B1 to B42 are collectively shown as measurement signal a2, but the measurement signals from each of the measurement sensors B1 to B42 may be different.

[0063] In signal separation processing T2, the signal separation unit 161 performs a predetermined signal separation process on the data of the measurement signal a2 (42 measurement signals from 42 measurement sensors B1 to B42). As a result of the signal separation process, a predetermined number of components (42 in this example) are obtained. In the example in Figure 4, the predetermined number of components are grouped together and shown as component a3.

[0064] Here, the measurement signal a2 obtained by the measurement process T1 may be subjected to predetermined preprocessing. In this case, the signal separation process T2 performs predetermined signal separation processing on the data resulting from the preprocessing (the data of the measurement signal a2 after preprocessing). Such preprocessing may include various processes, for example, a process that performs a predetermined filter on the measurement signals from each of the measurement sensors B1 to B42, or a process that adjusts the value (level) of the measurement signals from each of the measurement sensors B1 to B42 (a scale adjustment process). The filtering process may involve using a high-pass filter or a low-pass filter. Furthermore, as a preprocessing step, for example, adaptive noise cancellation (ANC) may be used. Such preprocessing may, for example, be a process to reduce power supply noise that may be included in the measurement signal. Note that pre-treatment is not always necessary.

[0065] In component selection processing T3, the component selection unit 162 uses the template waveform b1 to select some components from among the 42 components a3 as selected components a4. Specifically, in component selection process T3, each of the 42 components is compared with the template waveform b1, and based on the degree of correlation between each of these 42 components and the template waveform b1, the component that is predicted to contain the signal of interest is selected. Here, for example, it is possible that the degree of correlation with the template waveform b1 is low for all 42 components, and none of the 42 components are selected. However, for the sake of explanation, in this example, we will assume that one or more components are selected. In the example in Figure 4, the selected component a4 contains one or more selected components.

[0066] In the component reconstruction process T4, the component reconstruction unit 163 reconstructs the measurement signal for each of the 42 measurement sensors B1 to B42 based on the selected component a4. This generates 42 reconstructed signals. In the example in Figure 4, the reconstruction signals of the 42 measurement sensors B1 to B42 are collectively shown as reconstruction signal a5, but the reconstruction signals corresponding to each measurement sensor B1 to B42 may be different. Such a reconstructed signal consists of components (selected components) that are predicted to contain the signal of interest, with noise components reduced.

[0067] <Variations of the processing flow performed by the measuring device> Here, we will explain a modified version of the processing flow shown in Figure 4, specifically the case where multiple template waveforms are used. In this case, as shown in the example in Figure 4, a different process is performed, for example, as component selection process T3.

[0068] A modified example of the component selection process T3 when multiple template waveforms are used will be described. Here, we illustrate the case where the first template waveform C1 to the mth template waveform Cm, as shown in Figure 3, are available.

[0069] In the component selection process T3 related to the first modification, the component selection unit 162 attempts to select a component using one template waveform, but if no component is selected, it attempts to select a component using another template waveform. As a concrete example, first, an attempt is made to select components using the first template waveform C1. If one or more components are selected, the component selection process T3 is completed. If no components are selected, an attempt is made to select components using the second template waveform C2. Similarly, thereafter, if one or more components are selected using a template waveform, the component selection process T3 is completed. If no components are selected, an attempt is made to select components using the next template waveform. In this case, when attempting to select components using a template waveform fails to select any components, it is, for example, when the degree of correlation between any of the 42 components and the template waveform is insufficient (for example, the degree of correlation is below a predetermined threshold).

[0070] In the component selection process T3 according to the second modified example, the component selection unit 162 attempts to select components using each of the multiple different template waveforms, and selects all (or some) of the components selected for these multiple different template waveforms as the selected component a4. In other words, selects all (or some) of the components that have a high degree of correlation with at least one of the multiple different template waveforms (for example, the degree of correlation exceeds a predetermined threshold) as the selected component a4.

[0071] [Examples of waveforms in processing performed by measuring devices] Refer to Figures 5A, 5B, 6A, 6B, 7, 8, 9, 10A, and 10B to see examples of waveforms in the processing performed by the measuring device 1. These waveforms are examples for illustrative purposes only and are not necessarily representative of the exact waveforms.

[0072] In this example, SVD is used as the signal separation method. In this example, measurement sensors B1 to B42 are all magnetic sensors. In this example, the measurement target of the measurement sensors B1 to B42 is the cardiac magnetic field. In this example, the signal of interest is the heart rate signal.

[0073] Specifically, in this example, the target measured by measurement sensors B1 to B42 is the cardiac magnetic field signal, which is caused by the magnetic field generated by the myocardium of the subject's heart 51. In this example, heart rate can be detected from the measured signal. Generally, heart rate is the number of times the heart beats per minute. In this example, the measurement sensor unit 11, which has 42 measurement sensors B1 to B42, is positioned near the heart of the subject 51, and measurements are taken using the measurement sensors B1 to B42.

[0074] Figure 5A shows an example of the waveform of the 42-channel measurement signal according to the present invention. Figure 5B shows an example of the waveform of the measurement signal E1 of channel 1 (ch.1) according to the embodiment.

[0075] In the example shown in Figure 5A, graphs for 42 channels corresponding to 42 measurement sensors B1 to B42 are arranged in a (7x6) matrix. In the example in Figure 5A, each of the seven rows is arranged horizontally, and these seven rows are arranged vertically. Each of the six columns is also arranged vertically, and these six columns are arranged horizontally.

[0076] In the example in Figure 5A, the seven rows are arranged as follows: Channel 1 (ch.1) to Channel 6 (ch.6), Channel 7 (ch.7) to Channel 12 (ch.12), Channel 13 (ch.13) to Channel 18 (ch.18), Channel 19 (ch.19) to Channel 24 (ch.24), Channel 25 (ch.25) to Channel 30 (ch.30), Channel 31 (ch.31) to Channel 36 (ch.36), and Channel 37 (ch.37) to Channel 42 (ch.42).

[0077] In the example in Figure 5A, the illustration is simplified by omitting the horizontal and vertical axes of each graph. However, the horizontal and vertical axes of each graph, the range of the horizontal axis direction in each graph, and the range of the vertical axis direction in each graph are the same as those shown in Figure 5B. In other words, each of these 42 graphs represents a measurement signal acquired at the same time by each of the 42 measurement sensors B1 to B42.

[0078] In this example, the measurement signals for channels 1 (ch.1) to 42 (ch.42) are described by assigning the symbols E1 to E42 to each of them. In the example shown in Figure 5A, the illustration has been simplified, and only some of the measurement signals E1 to E42 are labeled with symbols. Here, the conditions of the signal of interest and noise measured and acquired by each of the measurement sensors B1 to B42 may differ depending, for example, on the placement of each of the measurement sensors B1 to B42.

[0079] Figure 5B shows one representative graph out of 42 graphs, representing the waveform of the measurement signal E1 for channel 1 (ch.1). In this graph, the horizontal axis represents time [seconds], and the vertical axis represents magnetic flux density B [T]. In this graph, the range of the horizontal axis (time range) is, for example, about 10 seconds, but is not limited to this. In the measured signals E1 to E42, for example, noise is superimposed on the signal of interest, and the signal of interest and the noise are mixed together, making it impossible to clearly identify the signal of interest.

[0080] Figure 6A shows an example of the waveforms of the 42 components after signal separation according to the embodiment. In this embodiment, for the sake of explanation, the 42 components are also referred to as components 1 through 42. Figure 6B shows an example of the waveform of the first component signal F1 according to the embodiment.

[0081] In the example shown in Figure 6A, the graphs of 42 components are arranged in a (7x6) matrix. In the example in Figure 6A, each of the seven rows is arranged horizontally, and these seven rows are arranged vertically. Each of the six columns is arranged vertically, and these six columns are arranged horizontally.

[0082] In the example in Figure 6A, the seven rows represent the order of components 1 through 6, components 7 through 12, components 13 through 18, components 19 through 24, components 25 through 30, components 31 through 36, and components 37 through 42, respectively.

[0083] In the example in Figure 6A, the illustration is simplified by omitting the horizontal and vertical axes of each graph. However, the horizontal and vertical axes of each graph, the range of the horizontal axis direction in each graph, and the range of the vertical axis direction in each graph are the same as those shown in Figure 6B. In other words, each of these 42 graphs represents the signal of 42 components during the same time period.

[0084] In this example, each component signal from the 1st to the 42nd component is described by assigning the symbols F1 to F42 to each component signal. In the example shown in Figure 6A, the illustration has been simplified, and only a portion of the component signals F1 to F42 are labeled with symbols.

[0085] Figure 6B shows a representative graph of the waveform of the first component signal F1, out of the 42 components. In this graph, the horizontal axis represents time [seconds], and the vertical axis represents magnetic flux density B [T]. In this graph, the range of the horizontal axis (time range) is the same as that of the graph shown in Figure 5B.

[0086] Figure 7 shows an example of a template waveform C according to the embodiment. In the graph shown in Figure 7, the horizontal axis represents time, and the vertical axis represents the waveform value (level). In the example shown in Figure 7, the horizontal axis represents time, indicated by the numbers of the sample points arranged in chronological order. In this graph, the range of the horizontal axis (time range) is, for example, approximately 0.8 seconds, but is not limited to this.

[0087] Here, template waveform C is an example of a template waveform. In this example, we show the case where one template waveform C is used, but in other examples, multiple template waveforms may be used. For example, in the example in Figure 3, the first template waveform C1 to the mth template waveform Cm are each waveforms similar to template waveform C, but are different from each other.

[0088] Figure 8 shows an example of the cross-correlation coefficient between the waveform of the component signal and the template waveform for some of the 42 components according to the embodiment. In the example in Figure 8, graphs showing the cross-correlation coefficients between the waveform of the component signal and the template waveform are shown for each of the eight components, components 1 through 8, which are part of the 42 components.

[0089] In the example shown in Figure 8, the graphs for the eight components are arranged in a (2x4) matrix. In the example in Figure 8, each of the two rows is arranged horizontally, and these two rows are arranged vertically. Each of the four columns is arranged vertically, and these four columns are arranged horizontally. In the example in Figure 8, the two rows represent the sequence of components 1 through 4 and components 5 through 8, respectively.

[0090] In the example in Figure 8, the horizontal axis of the graph for the eight components represents time, and the vertical axis represents the cross-correlation coefficient (CC). In the example shown in Figure 8, the horizontal axis represents time, indicated by the numbers of the sample points arranged in chronological order. These eight graphs represent the cross-correlation coefficients (CC) for eight components during the same time period. In these graphs, the range of the horizontal axis (time range) is, for example, about 5 seconds, but is not limited to this.

[0091] In this example, the cross-correlation coefficients of the first to eighth components are denoted with signs G1 to G8 for explanation. Furthermore, for the sake of clarity, the cross-correlation coefficients of the ninth to forty-second components are denoted with signs G9 to G42 for explanation. The cross-correlation coefficients G1 to G8 in the graphs of the first to eighth components represent the cross-correlation coefficients between the waveforms of the respective component signals F1 to F8 and the template waveform C. In this example, the same template waveform C is used for all eighth components.

[0092] In the example in Figure 8, the graphs for each of the first to eighth components show the threshold Q1 that has the same value with respect to the cross-correlation coefficient (CC) on the vertical axis. In the example in Figure 8, the threshold Q1 is 0.75, but it is not limited to this value.

[0093] In this example, the component selection unit 162 determines the number of times the cross-correlation coefficients G1 to G8 exceed the threshold Q1 within the time range on the horizontal axis for each graph of the first to eighth components, and selects the component for which this number is 3 or more. In the example shown in Figure 8, the number of times that the cross-correlation coefficients G1 to G8 for the first to eighth components exceed the threshold Q1 is 0, 5, 0, 3, 1, 0, 0, and 2, respectively. The component selection unit 162 then selects the second and fourth components from the first to eighth components.

[0094] In the example shown in Figure 8, the first to eighth components were illustrated, but the component selection unit 162 performs similar processing on the other components (ninth to forty-second components). In this example, for the other components (components 9 to 42), the number of times their respective cross-correlation coefficients G9 to G42 exceed the threshold Q1 is two or less. In other words, in this example, the component selection unit 162 does not select any of the other components (components 9 to 42). In this way, the component selection unit 162 selects from among the 42 components the component signals F1 to F42, each of which has a high cross-correlation coefficient between its waveform and the template waveform C.

[0095] Here, various values ​​may be used as the threshold Q1 for the cross-correlation coefficient (CC). Furthermore, the threshold for the number of times the cross-correlation coefficient (CC) exceeds the threshold Q1 is not limited to three times as in this example; for example, it could be one time, or any number of two or more times. Furthermore, various time ranges may be used for the time selection process performed by the component selection unit 162 (in the example in Figure 8, the time range on the horizontal axis of each graph).

[0096] Figure 9 shows an example of some of the selected components from the 42 components according to the embodiment. Figure 9 shows graphs of 42 component signals F1 to F42, similar to those in Figure 6A. In Figure 9, the graph of the second component selected by the component selection unit 162 is enclosed in frame R1, and the graph of the fourth component selected by the component selection unit 162 is enclosed in frame R2.

[0097] Thus, in this example, the component selection unit 162 selects two components: the second component, which is a component of the graph enclosed by frame R1, and the fourth component, which is a component of the graph enclosed by frame R2. In other words, in this example, for each of these two components, the component selection unit 162 determines that the number of times the cross-correlation coefficient between the component signal and the template waveform C exceeds the threshold Q1 is three or more within a predetermined time range.

[0098] Figure 10A shows an example of the waveform of the reconstructed signal for 42 channels according to the embodiment. Figure 10B shows an example of the waveform of the reconstructed signal H27 of channel 27 (ch.27) according to the embodiment. In this example, the component reconstruction unit 163 generates a reconstructed signal using SVD with two selected components, the second component and the fourth component.

[0099] In the example shown in Figure 10A, graphs for 42 channels corresponding to 42 measurement sensors B1 to B42 are arranged in a (7x6) matrix. Note that the (7 rows x 6 columns) arrangement is the same as in Figure 5A.

[0100] In the example shown in Figure 10A, the horizontal and vertical axes of each graph have been omitted for simplification of the illustration. However, the horizontal and vertical axes of each graph, the range of the horizontal axis direction in each graph, and the range of the vertical axis direction in each graph are the same as those shown in Figure 6B. In other words, each of these 42 graphs corresponds to one of the 42 measurement sensors B1 to B42, and represents the reconstructed signal for the same time period generated for the measurement signal from each of the measurement sensors B1 to B42.

[0101] In this example, the reconstructed signals for channels 1 (ch.1) to 42 (ch.42) are described by assigning the codes H1 to H42 to each of them. In the example shown in Figure 10A, the illustration has been simplified, and only a portion of the reconstructed signals H1 to H42 are labeled with symbols.

[0102] Figure 10B shows one representative graph out of 42 graphs, representing the waveform of the reconstructed signal H27 for channel 27 (ch.27). In this graph, the horizontal axis represents time [seconds], and the vertical axis represents magnetic flux density B [T]. Note that, for the sake of explanation, the time range on the horizontal axis of the graph in Figure 10B is different from that shown in the graph in Figure 10A.

[0103] Here, in the waveform of the reconstructed signal H27 shown in Figure 10B, the sign is reversed compared to the template waveform C shown in Figure 7, and the peak on the positive side of the template waveform C shown in Figure 7 appears on the negative side. The reason why the positive and negative signs are reversed in the example in Figure 10B and the example in Figure 7 is that the orientation of the measurement sensor B27 in channel 27 (in this example, the orientation of the magnetic sensor) causes the positive and negative signs of the waveform of the measurement signal E27 to be reversed with respect to the positive and negative signs of the template waveform C.

[0104] In the example shown in Figure 10B, the time positions of the negative peaks of the reconstructed signal H27 are clearly visible at regular (or nearly regular) time intervals with respect to time on the horizontal axis. In this way, in each reconstructed signal H1 to H42, the noise is reduced compared to the noise in the respective measured signals E1 to E42, resulting in a clearer waveform for the signal of interest. Ideally, it would be desirable for the noise contained in each reconstructed signal H1 to H42 to be zero (i.e., noise-free), but this is not the only requirement; it is sufficient if the noise is reduced to a level that is practically acceptable.

[0105] In the example in Figure 10B, for the sake of simplicity, only some of the multiple negative peaks appearing in the reconstructed signal H27 (peak I1, peak I2, peak I3) are labeled. In the example shown in Figure 10B, the positive peak of the reconstructed signal H27 corresponds to the negative peak of the template waveform C shown in Figure 7. However, in this example, the time interval of the negative peak of the reconstructed signal H27 was examined.

[0106] (Summary of the embodiments) As described above, the measurement device 1 and measurement method according to this embodiment can acquire a signal of interest with high accuracy from a measurement signal in which the signal of interest and noise are mixed.

[0107] In the measurement device 1 and measurement method according to this embodiment, based on a measurement signal containing a mixture of signal of interest and noise, signal separation into multiple components, component selection using a template waveform, and reconstruction using the selected components can be performed to accurately extract the signal of interest, thereby obtaining a highly accurate signal of interest. In this embodiment, the template waveform is generated in advance, before the component selection process.

[0108] Thus, in this embodiment, by taking advantage of the fact that measurement signals are acquired by multiple measurement sensors B1 to B42, the accuracy of the acquired signal of interest data can be improved. As a result, in this embodiment, for example, even from measurement signals with a poor signal-to-noise ratio (SNR), the data of the signal of interest can be accurately extracted with fewer processing steps. The measuring device 1 and measuring method according to this embodiment can enhance the added value of the measuring device 1.

[0109] In this embodiment, the template waveform of the signal of interest is used for component selection after signal separation, and then the selected components are reconstructed to extract the signal of interest. In this embodiment, a template waveform is used in the component selection process, but it is not used in the reconstruction process. In this embodiment, for example, it is possible not only to identify the peak of the signal of interest, but also to extract the signal of interest without losing any information about it.

[0110] In this embodiment, the number of measurement sensors is shown to be 42, but the number of measurement sensors may be any number of 2 or more. Furthermore, any configuration can be used for arranging the multiple measurement sensors. For example, not only a two-dimensional planar arrangement as shown in Figure 2, but also a three-dimensional arrangement may be used.

[0111] Furthermore, in this embodiment, we have shown a case where the number of measurement signals and the number of components after signal separation are the same, such as applying signal separation to 42 measurement signals to obtain 42 components. However, as another example, an embodiment in which the number of measurement signals and the number of components after signal separation do not match may also be used. In other words, depending on the signal separation method used, the number of multiple measurement signals (number of multiple measurement sensors) may not match the number of components after signal separation. For example, the number of components after signal separation may be greater than the number of multiple measurement signals (number of multiple measurement sensors), or the number of components after signal separation may be less than the number of multiple measurement signals (number of measurement sensors). Furthermore, when singular value decomposition (SVD), principal component analysis (PCA), independent component analysis (ICA), or non-negative matrix factorization (NMF) is used, the number of components after signal separation is basically the same as the number of multiple measurement signals (number of multiple measurement sensors).

[0112] In this embodiment, we have shown a case where multiple measuring sensors are of the same type, but as another example, a configuration in which multiple measuring sensors are of different types may be used. For example, when measuring heart rate, the multiple measurement sensors may include two or three types of sensors from among magnetic sensors, acceleration sensors, and vibration sensors.

[0113] In this case, if multiple measurement sensors are used, including two or more types of sensors (i.e., different types of sensors), the scale of the measurement signals acquired by different types of measurement sensors may be different. In this case, for example, the measurement signals from all measurement sensors, or from some specific types of measurement sensors, may be scaled up, and then the signal separation process may be performed on the scaled-up measurement signals. This makes it possible to improve the accuracy of the signal separation results compared to a case where such scale adjustment is not performed.

[0114] As a specific example, when multiple types of measuring sensors are used, scaling adjustments may be performed on the measurement signals from each type of measuring sensor using a scaling adjustment coefficient specific to that sensor. In the scaling adjustment, a predetermined calculation may be performed on the measurement signal. This predetermined calculation may be, for example, multiplying the measurement signal by the scaling adjustment coefficient, adding the scaling adjustment coefficient to the measurement signal, subtracting the scaling adjustment coefficient from the measurement signal, or dividing the measurement signal by the scaling adjustment coefficient.

[0115] The scaling factor may be pre-set for each type of measurement sensor, or it may be determined by any method. For example, this method may involve determining the scaling factor so that the range of waveform values ​​(levels) of the measurement signal from one type of measurement sensor matches the range of waveform values ​​(levels) of the measurement signal from another type of measurement sensor. Furthermore, for example, a configuration may be used in which no scale adjustment is performed for measurement signals from one or more types of measurement sensors.

[0116] Furthermore, for example, the measurement signals from two or more measurement sensors composed of the same type may be scaled using different scale adjustment coefficients. It should be noted that the scale adjustment described here is not necessarily required; in other words, the measurement signals from multiple measurement sensors may be used for signal separation processing without scale adjustment.

[0117] An example configuration according to the embodiment is shown. As an example configuration, the measuring device 1 according to this embodiment includes a measuring sensor unit 11, a signal separation unit 161, a component selection unit 162, and a component reconstruction unit 163. The measurement sensor unit 11 has multiple measurement sensors that measure a measurement signal containing both the signal of interest and noise, and measures the measurement signal using each of the measurement sensors. The signal separation unit 161 performs signal separation on the data of measurement signals measured by multiple measurement sensors. The component selection unit 162 compares each of the multiple components obtained by signal separation with the template waveform of the signal of interest and selects the component that is considered to contain the signal of interest. The component reconstruction unit 163 reconstructs data for each of the multiple measurement sensors based on the components selected by the component selection unit 162.

[0118] Therefore, the measuring device 1 according to this embodiment can acquire a highly accurate signal of interest from a measurement signal that contains both the signal of interest and noise. In the measuring device 1 according to this embodiment, for example, compared to using the measurement signals from each measurement sensor as they are, signal separation makes it easier to obtain components for each cause of noise generation, and noise can be removed more effectively by selecting and discarding components.

[0119] As an example configuration, the measuring device 1 according to this embodiment has the following configuration. The component selection unit 162 selects a component whose correlation with the template waveform is higher than a predetermined value, based on the degree of correlation between each of the multiple components and the template waveform. Therefore, in the measuring device 1 according to this embodiment, it is possible to select components with a high degree of correlation (degree of correlation) with the template waveform.

[0120] As an example configuration, the measuring device 1 according to this embodiment has the following configuration. Multiple template waveforms, each different from the others, are provided. The component selection unit 162 selects components using multiple template waveforms. Therefore, in the measuring device 1 according to this embodiment, by using multiple template waveforms, it is possible to select components more reliably compared to the case where one template waveform is used.

[0121] As an example configuration, the measuring device 1 according to this embodiment has the following configuration. The template waveform is a waveform acquired using the same type of sensor as the measurement sensor. Therefore, in the measuring device 1 according to this embodiment, for example, a suitable template waveform can be used when it is appropriate to use a waveform acquired using a sensor of the same type as the measuring sensor as a template waveform.

[0122] As an example configuration, the measuring device 1 according to this embodiment has the following configuration. The template waveform is a waveform acquired using a sensor different from the measurement sensor. Therefore, in the measuring device 1 according to this embodiment, for example, a suitable template waveform can be used when it is appropriate to use a waveform acquired using a sensor of a different type than the measurement sensor as a template waveform.

[0123] As an example configuration, the measuring device 1 according to this embodiment has the following configuration. A template waveform is an artificially created waveform. Therefore, in the measuring device 1 according to this embodiment, for example, an artificially created waveform can be used as a template waveform as appropriate.

[0124] As an example configuration, the measuring device 1 according to this embodiment has the following configuration. A template waveform is a waveform obtained by applying signal processing to the original template waveform. Therefore, in the measuring device 1 according to this embodiment, for example, a waveform obtained by applying signal processing to the original template waveform can be used as the template waveform. Here, any waveform may be used as the template waveform. Furthermore, various signal processing techniques may be used for this signal processing.

[0125] As an example configuration, the measuring device 1 according to this embodiment has the following configuration. The signal separation unit 161 performs signal separation after scaling the data of the measurement signals measured by multiple measurement sensors. Therefore, in the measuring device 1 according to this embodiment, for example, the accuracy of signal separation can be improved by performing scale adjustment compared to a case where scale adjustment is not performed.

[0126] Furthermore, in particular, in configurations where multiple measurement sensors include different types of sensors, performing scale adjustment may improve the accuracy of signal separation compared to when no scale adjustment is performed. For example, the configuration may be such that the same type of scaling adjustment is performed on measurement signals from the same type of measuring sensor, and different types of scaling adjustment are performed on measurement signals from different types of measuring sensors.

[0127] As an example configuration, the measuring device 1 according to this embodiment has the following configuration. The signal separation is blind signal separation. Therefore, in the measuring device 1 according to this embodiment, an effective method can be used as the signal separation method.

[0128] As an example configuration, the measuring device 1 according to this embodiment has the following configuration. Each of the multiple measuring sensors is either a magnetic sensor or a potential sensor. Therefore, in the measuring device 1 according to this embodiment, when a magnetic sensor or a potential sensor is used as the measuring sensor, it is possible to obtain a signal of interest with high accuracy from a measurement signal that contains a mixture of signal of interest and noise.

[0129] As an example configuration, the measuring device 1 according to this embodiment has the following configuration. The measurement target of the measurement sensor is either the cardiac magnetic field or cardiac potential. Therefore, in the measuring device 1 according to this embodiment, when the target of measurement is the cardiac magnetic field or cardiac potential, it is possible to obtain a signal of interest with high accuracy from a measurement signal that contains a mixture of the signal of interest and noise.

[0130] As an example configuration, the measuring device 1 according to this embodiment has the following configuration. The signal of interest is the heart rate signal. Therefore, in the measuring device 1 according to this embodiment, when the signal of interest is a heart rate signal, it is possible to obtain a signal of interest with high accuracy from a measurement signal that contains a mixture of the signal of interest and noise.

[0131] As an example configuration, the measurement method according to this embodiment performs the following processing. The measurement signal is measured by the measurement sensors of the measurement sensor unit 11, which has multiple measurement sensors that measure the measurement signal containing both the signal of interest and noise. The signal separation unit 161 performs signal separation on the data of measurement signals measured by multiple measurement sensors. The component selection unit 162 compares each of the multiple components obtained by signal separation with the template waveform of the signal of interest and selects the component that is considered to contain the signal of interest. The component reconstruction unit 163 reconstructs data for each of the multiple measurement sensors based on the components selected by the component selection unit 162. Therefore, the measurement method according to this embodiment makes it possible to obtain a signal of interest with high accuracy from a measurement signal that contains both the signal of interest and noise.

[0132] Furthermore, a program to realize the function of any component in any of the devices described above may be recorded on a computer-readable recording medium, and that program may be loaded into a computer system and executed. Here, "computer system" includes hardware such as operating systems and peripheral devices. "Computer-readable recording medium" refers to portable media such as flexible disks, magneto-optical disks, ROMs, CD (Compact Disc)-ROMs (Read Only Memory), and storage devices such as hard disks built into computer systems. Moreover, "computer-readable recording medium" also includes volatile memory within a computer system that retains a program for a certain period of time, such as volatile memory within a computer system that acts as a server or client when a program is transmitted via a network such as the Internet or a communication line such as a telephone line. Such volatile memory may be, for example, RAM (Random Access Memory). The recording medium may be, for example, a non-temporary recording medium.

[0133] Furthermore, the above program may be transmitted from a computer system that stores the program in a memory device or the like to another computer system via a transmission medium or by transmission waves within the transmission medium. Here, the "transmission medium" used to transmit the program refers to a medium that has the function of transmitting information, such as a network like the Internet or a communication line like a telephone line. Furthermore, the above program may be intended to implement only a portion of the functions described above. Moreover, the above program may be a so-called differential file, capable of implementing the aforementioned functions in combination with programs already recorded in the computer system. A differential file may also be called a differential program.

[0134] Furthermore, the functions of any component in any device described above may be implemented by a processor. For example, each process in the embodiment may be implemented by a processor that operates based on information such as a program, and a computer-readable recording medium that stores information such as a program. Here, the processor may be implemented by having the functions of each part implemented by separate hardware, or by having the functions of each part implemented by integrated hardware. For example, the processor includes hardware, and the hardware may include at least one of a circuit that processes digital signals and a circuit that processes analog signals. For example, the processor may be configured using one or more circuit devices or one or both of one or more circuit elements mounted on a circuit board. An IC (Integrated Circuit) may be used as the circuit device, and a resistor or capacitor may be used as the circuit element.

[0135] Here, the processor may be, for example, a CPU. However, the processor is not limited to a CPU; various types of processors may be used, such as a GPU (Graphics Processing Unit) or a DSP (Digital Signal Processor). The processor may also be, for example, a hardware circuit using an ASIC (Application Specific Integrated Circuit). Furthermore, the processor may be composed of, for example, multiple CPUs, or multiple hardware circuits using ASICs. The processor may also be composed of, for example, a combination of multiple CPUs and multiple hardware circuits using ASICs. The processor may also include, for example, one or more amplifier circuits or filter circuits that process analog signals.

[0136] While embodiments of this disclosure have been described in detail above with reference to the drawings, the specific configuration is not limited to these embodiments and may include designs and other elements that do not depart from the gist of this disclosure.

[0137] [Note] (Configuration Examples 1) to (Configuration Examples 13) are shown.

[0138] (Configuration Example 1) A measurement sensor unit having multiple measurement sensors that measure a measurement signal containing a mixture of signals of interest and noise, A signal separation unit that performs signal separation on the data of the measurement signals measured by the multiple measurement sensors, A component selection unit compares each of the multiple components obtained by the signal separation with the template waveform of the signal of interest and selects the component that is considered to contain the signal of interest. A component reconstruction unit reconstructs data for each of the multiple measurement sensors based on the components selected by the component selection unit, A measuring device equipped with the following features.

[0139] (Configuration example 2) The component selection unit selects a component whose correlation with each of the plurality of components is higher than a predetermined value, based on the degree of correlation between each of the plurality of components and the template waveform. The measuring device described in (Configuration Example 1).

[0140] (Configuration Example 3) Multiple template waveforms that are different from each other are provided, The component selection unit selects the component using a plurality of template waveforms. The measuring device described in (Configuration Example 1) or (Configuration Example 2).

[0141] (Configuration example 4) The template waveform is a waveform acquired using a sensor of the same type as the measurement sensor. A measuring device as described in any one of (Configuration Example 1) to (Configuration Example 3).

[0142] (Configuration example 5) The template waveform is a waveform acquired using a sensor of a different type than the measurement sensor. A measuring device as described in any one of (Configuration Example 1) to (Configuration Example 4).

[0143] (Configuration example 6) The aforementioned template waveform is an artificially created waveform. A measuring device as described in any one of (Configuration Example 1) to (Configuration Example 5).

[0144] (Configuration example 7) The template waveform is a waveform obtained by applying signal processing to the original template waveform. A measuring device as described in any one of (Configuration Example 1) to (Configuration Example 6).

[0145] (Configuration example 8) The signal separation unit performs signal separation after scaling the data of the measurement signals measured by the multiple measurement sensors. A measuring device as described in any one of (Configuration Example 1) to (Configuration Example 7).

[0146] (Configuration example 9) The aforementioned signal separation is blind signal separation. A measuring device as described in any one of (Configuration Example 1) to (Configuration Example 8).

[0147] (Configuration example 10) Each of the multiple measurement sensors is either a magnetic sensor or a potential sensor. A measuring device as described in any one of (Configuration Example 1) to (Configuration Example 9).

[0148] (Configuration Example 11) The object to be measured by the aforementioned measurement sensor is the cardiac magnetic field or cardiac potential. The measuring device described in (Configuration Example 10).

[0149] (Configuration Example 12) The aforementioned signal of interest is the heart rate signal. A measuring device as described in any one of (Configuration Example 1) to (Configuration Example 11).

[0150] This provides a measurement method that performs the same processing as that performed by a measuring device. (Configuration Example 13) The measurement sensor unit, which has multiple measurement sensors that measure a measurement signal containing a mixture of signals of interest and noise, measures the measurement signal using the measurement sensors. The signal separation unit performs signal separation on the data of the measurement signals measured by the multiple measurement sensors. The component selection unit compares each of the multiple components obtained by the signal separation with the template waveform of the signal of interest, and selects the component that is considered to contain the signal of interest. The component reconstruction unit reconstructs data for each of the multiple measurement sensors based on the components selected by the component selection unit. Measurement method. [Explanation of Symbols]

[0151] 1... Measuring device, 11... Measurement sensor unit, 21... Information processing unit, 111... Input unit, 112... Output unit, 113... Storage unit, 114... Control unit, 131... Acquisition unit, 141... Display unit, 151... Processing unit, 152... Display control unit, 161... Signal separation unit, 162... Component selection unit, 163... Component reconstruction unit, A1... Template waveform storage unit, B1~B42... Measurement sensor, C1... First template waveform, C2... Second template waveform, C, Cm... mth template waveform, E1~E42... Measurement signal, F1~F42... Component signal, G1~G42... Cross-correlation coefficient, H1~H42... Reconstructed signal, I1~I3... Peak, Q1... Threshold, R1~R2... Frame

Claims

1. A measurement sensor unit having multiple measurement sensors that measure a measurement signal containing a mixture of signals of interest and noise, A signal separation unit that performs signal separation on the data of the measurement signals measured by the multiple measurement sensors, A component selection unit compares each of the multiple components obtained by the signal separation with the template waveform of the signal of interest and selects the component that is considered to contain the signal of interest. A component reconstruction unit reconstructs data for each of the multiple measurement sensors based on the components selected by the component selection unit, A measuring device equipped with the following features.

2. The component selection unit selects a component whose correlation with each of the plurality of components is higher than a predetermined value, based on the degree of correlation between each of the plurality of components and the template waveform. The measuring device according to claim 1.

3. Multiple template waveforms that are different from each other are provided, The component selection unit selects the component using a plurality of template waveforms. The measuring device according to claim 1 or claim 2.

4. The template waveform is a waveform acquired using a sensor of the same type as the measurement sensor. The measuring device according to claim 1 or claim 2.

5. The template waveform is a waveform acquired using a sensor of a different type than the measurement sensor. The measuring device according to claim 1 or claim 2.

6. The aforementioned template waveform is an artificially created waveform. The measuring device according to claim 1 or claim 2.

7. The template waveform is a waveform obtained by applying signal processing to the original template waveform. The measuring device according to claim 1 or claim 2.

8. The signal separation unit performs signal separation after scaling the data of the measurement signals measured by the multiple measurement sensors. The measuring device according to claim 1 or claim 2.

9. The aforementioned signal separation is blind signal separation. The measuring device according to claim 1 or claim 2.

10. Each of the multiple measurement sensors is either a magnetic sensor or a potential sensor. The measuring device according to claim 1 or claim 2.

11. The object to be measured by the aforementioned measurement sensor is the cardiac magnetic field or cardiac potential. The measuring device according to claim 10.

12. The aforementioned signal of interest is the heart rate signal. The measuring device according to claim 1 or claim 2.

13. The measurement sensor unit, which has multiple measurement sensors that measure a measurement signal containing a mixture of signals of interest and noise, measures the measurement signal using the measurement sensors. The signal separation unit performs signal separation on the data of the measurement signals measured by the multiple measurement sensors. The component selection unit compares each of the multiple components obtained by the signal separation with the template waveform of the signal of interest, and selects the component that is considered to contain the signal of interest. The component reconstruction unit reconstructs data for each of the multiple measurement sensors based on the components selected by the component selection unit. Measurement method.

Images