Biological information analysis system, biological information analysis program, device performance evaluation method, attachment state evaluation method, and device configuration selection method

Abstract

A biological information analysis system according to the present invention performs estimation of data quality by using biological information of a subject. The biological information analysis system comprises: an analysis unit determination unit that generates unit biological information obtained by dividing biological information into segments of a preset segment length; an index calculation unit that quantifies the data quality as a plurality of evaluation indexes for the unit biological information; and an estimation unit that estimates a quality index by inputting a plurality of evaluation indexes based on the biological information of the subject to an estimation model which is generated by training using the unit biological information and the plurality of evaluation indexes.

Description

Biological information analysis system, biological information analysis program, method for evaluating device performance, method for evaluating wearing condition, and method for selecting device configuration

[0001] The present invention relates to a series of biometric information analysis systems, biometric information analysis programs, methods for evaluating device performance, methods for evaluating attachment status, and methods for selecting device configurations, which determine arrhythmia based on biometric information acquired from a subject.

[0002] Various pieces of information (biometric information) obtained from a living organism are important for understanding the physical and psychological state of that organism. Examples of biometric information include electrocardiogram, pulse wave, body temperature, and blood pressure. Wearable devices are worn on the user's body (arm, wrist, head, etc.) to obtain biometric information and enable daily health management and optimization of exercise effects. As a result, wearable devices have become increasingly popular in recent years in line with the growing demand for healthcare and fitness.

[0003] However, the quality of biometric data acquired by wearable devices can vary depending on the state of wearing and the configuration. If the data quality deteriorates, the wearable device may not be able to fulfill its original purpose (e.g., diagnosing the user's health condition). Therefore, it is important to consider the configuration and quantify the quality of the acquired data when monitoring the state of wearing.

[0004] Patent Documents 1 and 2 disclose techniques for evaluating data quality of biological information. Specifically, Patent Document 1 quantifies data quality by treating the similarity and variability of shapes of waveforms detected from biological information as evaluation indices. Furthermore, Patent Document 2 quantifies data quality by treating the intensity of a predetermined band focusing on the frequency band of an electrocardiogram signal or the wave height of a baseline component as evaluation indices.

[0005] As in Patent Documents 1 and 2, there are a wide variety of methods for evaluating the data quality of biometric information, each focusing on and quantifying different characteristics. Therefore, in order to more accurately capture changes in data quality, it is desirable to perform evaluation using multiple evaluation indices obtained by these various methods. Furthermore, in general, using more evaluation indices is expected to reduce the risk of overlooking changes in data quality and improve the performance of interpreting data quality.

[0006] JP 2019-098183 A JP 2021-079007 A

[0007] However, the more evaluation indices there are, the more complicated the evaluation becomes, and there is a possibility that the evaluation accuracy may actually decrease. In other words, there is a demand for a simple and highly accurate evaluation of biometric data quality while improving the interpretation performance of data quality using multiple evaluation indices.

[0008] The present invention aims to solve the above-mentioned problems and provide a biometric information analysis system, a biometric information analysis program, a method for evaluating device performance, a method for evaluating the wearing state, and a method for selecting a device configuration, which can easily and accurately evaluate the data quality of biometric information obtained from a living body.

[0009] The biometric information analysis system of the present invention for achieving the above object comprises the following components: (1) A biometric information analysis system for estimating data quality using a target biometric information, comprising: an analysis unit determination unit that generates unit biometric information by dividing the biometric information at preset segment lengths; an index calculation unit that quantifies data quality of the unit biometric information as multiple evaluation indexes; and an estimation unit that estimates a quality index by inputting the multiple evaluation indexes based on the target biometric information into an estimation model that is generated by learning using the unit biometric information and the multiple evaluation indexes.

[0010] (2) The bioinformation analysis system according to (1), wherein the estimation model is obtained by learning using the plurality of evaluation indexes as explanatory variables and the teacher quality index as a target variable.

[0011] (3) The biological information analysis system according to (2), wherein the teacher quality index is a binary label indicating whether or not the analysis is possible.

[0012] (4) The biological information analysis system according to (1), wherein the estimation model is a nonlinear model.

[0013] (5) The bioinformation analysis system according to any one of (1) to (4), wherein the evaluation indexes include one or more of the following: similarity to reference data, spectral intensity in a predetermined band, wave height of the waveform in the bioinformation, wave height of superimposed noise, number of waveforms per predetermined time, and signal-to-noise ratio calculated from features of the waveform in the bioinformation and features of the superimposed noise.

[0014] (6) The biological information analysis system according to (5), wherein the plurality of evaluation indexes include at least a spectral intensity in a predetermined low frequency band and a spectral intensity in a predetermined high frequency band.

[0015] (7) The biological information analysis system according to (5), wherein the plurality of evaluation indices include at least a similarity to reference data, a spectral intensity in a predetermined low frequency band, and a spectral intensity in a predetermined high frequency band.

[0016] (8) A biometric information analysis program for estimating data quality using a target biometric information, the biometric information analysis program causing a computer to execute the following steps: read a preset segment length from a storage unit; generate unit biometric information by dividing the biometric information by the segment length; quantify data quality for the unit biometric information as multiple evaluation indices; input the multiple evaluation indices based on the target biometric information into an estimation model generated by learning using the unit biometric information and the multiple evaluation indices, and estimate a quality index.

[0017] (9) A method for evaluating device performance of a device that acquires biometric information from a subject, comprising: a step of acquiring the biometric information from the subject using the device; a step of estimating the quality index of the acquired biometric information using a biometric information analysis system described in any one of (1) to (7) above; and a step of evaluating device performance based on the quality index.

[0018] (10) A method for evaluating the wearing state of a device that acquires biometric information from a subject, comprising: a step of acquiring the biometric information from the subject using the device; a step of estimating the quality index of the acquired biometric information using a biometric information analysis system described in any one of (1) to (7) above; a step of evaluating the wearing state of the device based on the quality index; and a step of notifying the evaluation result.

[0019] (11) A method for selecting an optimal device configuration for a device that acquires biometric information from a target, the method comprising: acquiring the biometric information from the target using a default device configuration for the device; estimating the quality index of the acquired biometric information using a biometric information analysis system described in any one of (1) to (7) above; and selecting the optimal device configuration for the target based on the quality index.

[0020] According to the present invention, by calculating multiple evaluation indices for biometric information acquired from a living body and estimating a quality index using an estimation model, data quality can be evaluated simply and with high accuracy.

[0021] FIG. 1 is a diagram illustrating an example of a configuration of a bioinformation analysis system according to a first embodiment. FIG. 2 is a diagram illustrating a detailed example of an electrocardiogram signal used for analysis. FIG. 3 is a flowchart illustrating a flow of a data quality evaluation process according to the first embodiment. FIG. 4 is a flowchart illustrating a flow of an estimation model generation process for estimating a quality index according to the first embodiment. FIG. 5 is a flowchart illustrating a learning process in the estimation model generation process according to the first embodiment. FIG. 6A is a diagram (part 1) illustrating an estimation process according to the first embodiment. FIG. 6B is a diagram (part 2) illustrating an estimation process according to the first embodiment. FIG. 7 is a diagram illustrating an example of a configuration of a bioinformation analysis system according to a second embodiment. FIG. 8 is a flowchart illustrating a flow of a performance comparison process according to the second embodiment. FIG. 9 is a flowchart illustrating a flow of a wearing state evaluation process according to the third embodiment. FIG. 10 is a flowchart illustrating a flow of a device configuration determination process according to the fourth embodiment.

[0022] Hereinafter, embodiments of the bioinformation analysis system according to the present invention will be described in detail with reference to the drawings. Note that the present invention is not limited to these embodiments. Furthermore, the individual embodiments of the present invention are not independent and can be implemented in combination with each other as appropriate.

[0023] 1 is a diagram showing an example of the configuration of a biological information analysis system according to embodiment 1 of the present invention. The biological information analysis system includes an electrocardiogram signal measuring device 2, which is a biological information measuring device, an analysis system 3, a receiving terminal 4, and a learning device 5.

[0024] The biological information measuring device is not limited and can be freely determined depending on the biological information to be acquired. The biological information measuring device is also expected to have a biological information output mechanism represented by a connector for electrically connecting to the analysis system 3, a communication device having means for communicating with the biological information analysis system, or an input port for media into which the acquired biological information is written. In the first embodiment, the biological information is an electrocardiogram signal 10, the biological information measuring device is an electrocardiogram signal measuring device 2, the electrocardiogram signal measuring device 2 has an input port for media into which the acquired biological information, the electrocardiogram signal 10, and the subject 1 is a human being.

[0025] 1, an electrocardiogram signal 10 to be analyzed is acquired from a subject 1 via an electrocardiogram signal measuring device 2. The subject 1 is not particularly limited to a human being or an animal. The receiving terminal 4 is, for example, a smartphone carried by the subject 1 or a guard (for example, a surgeon such as a doctor, a factory hygiene manager, or a nearby worker).

[0026] A wearable electrocardiograph is an example of the electrocardiogram signal measuring device 2. Specifically, the electrocardiogram signal measuring device 2 includes a garment body to be worn by the subject 1, a plurality of electrodes, and an electrocardiograph 100 electrically connected to each electrode. The electrocardiograph 100 and the like are fixed to the garment body using a band or the like.

[0027] The electrocardiograph 100 is an example of a measuring device that acquires an electrocardiographic signal 10. The electrocardiograph 100 has a function of continuously acquiring the electrocardiographic signal 10 of the subject 1, a function of storing the acquired electrocardiographic signal 10, and a function of transferring data via communication with the analysis system 3. Note that the electrocardiograph 100 may be configured to transfer data including the electrocardiographic signal 10 to a server device, and the server device may transfer the electrocardiographic signal 10 to the analysis system 3. In the following description, the subject 1 refers to a person wearing an electrocardiographic signal measuring device 2.

[0028] In the first embodiment, an electrocardiogram signal 10 acquired from a subject 1 includes information about the subject 1, the acquisition date and time, the location, and the acquisition result. The acquisition result is waveform data 11 obtained by plotting the horizontal axis as time and the vertical axis as voltage (see FIG. 1 ).

[0029] 2 is a diagram showing a detailed example of an electrocardiographic signal 10 used for analysis. The electrocardiographic signal 10 may include components called P waves 11a, Q waves 11b, R waves 11c, S waves 11d, and T waves 11e, and the shapes of these components are used as criteria for determining arrhythmia. For example, waveform data 11 is composed of repeated patterns of these waveform components. Note that the shape of the waveform data 11 included in the electrocardiographic signal 10 may change depending on the subject and the way the electrocardiograph 100 is worn.

[0030] Returning to FIG. 1, the analysis system 3 includes a communication unit 311 , an analysis unit determination unit 312 , an index calculation unit 313 , an estimation unit 314 , an input / output unit 315 , a control unit 316 , and a storage unit 317 .

[0031] The communication unit 311 can communicate with the electrocardiogram signal measuring device 2 and the receiving terminal 4 via a communication network. The communication network here is configured using, for example, an existing public line network, a local area network (LAN), a wide area network (WAN), etc., and may be wired or wireless. The communication unit 311 is configured, for example, by a connector that electrically connects to a communication target, a communication device that has means for communicating with the electrocardiogram signal measuring device 2, or an input port for a medium on which data is stored.

[0032] The analysis unit determination unit 312 divides the acquired electrocardiogram signal 10 into predetermined segments, which are then used as analysis units (unit biological information).

[0033] The index calculation unit 313 calculates two or more evaluation indexes based on the electrocardiogram signal 10 in the determined analysis unit.

[0034] The estimation unit 314 estimates the quality index of the electrocardiogram signal 10 using the evaluation index calculated by the index calculation unit 313 and the estimation model generated by the learning device 5 .

[0035] The input / output unit 315 is configured by devices having input and output functions, and outputs various information under the control of the control unit 316. The input / output unit 315 includes a user interface such as a keyboard, mouse, microphone, etc., and has an input function of receiving input of the electrocardiogram signal 10 output from the electrocardiogram signal measuring device 2. Furthermore, the input / output unit 315 has output functions such as a display made of, for example, a liquid crystal or organic EL (Electro Luminescence) display, and a speaker that outputs sound.

[0036] Furthermore, it is assumed that the input function of the input / output unit 315 includes an input mechanism for the electrocardiogram signal 10, such as an input port for a medium on which the electrocardiogram signal 10 is written. If the input function of the input / output unit 315 overlaps with the function of the communication unit 311, it is sufficient that either one of them has that function.

[0037] Furthermore, for the biometric information to be input to the analysis system 3, in addition to using previously acquired electrocardiogram signals 10, the acquisition of the electrocardiogram signals 10 by the electrocardiogram signal measuring device 2 and the input to the analysis system 3 may be carried out in parallel. In these cases, it is assumed that the electrocardiogram signal measuring device 2 will constantly transmit values ​​acquired in a continuous manner, or will transmit values ​​acquired in a batch manner in batches at regular intervals, and there is no particular limitation.

[0038] The control unit 316 comprehensively controls the operation of the analysis system 3. The control unit 316 also transmits the estimation result by the estimation unit 314 together with the wearer information to the receiving terminal 4 via the communication unit 311. The wearer information is information for identifying the subject 1, including personal information such as the name and gender of the subject 1, and the management number of the wearable device or peripheral equipment worn by the subject 1. The control unit 316 may also output the estimation result by the estimation unit 314 using an output function of the input / output unit 315.

[0039] The storage unit 317 stores various programs for operating the analysis system 3 and various data including data generated by each unit. The various programs include a program for evaluating data quality executed using a trained model. The storage unit 317 is configured using a ROM (Read Only Memory) in which various programs are pre-installed, a RAM (Random Access Memory), an HDD (Hard Disk Drive), an SSD (Solid State Drive), etc. that store calculation parameters and data for each process.

[0040] The various programs can be widely distributed by recording them on computer-readable recording media such as HDDs, flash memories, CD-ROMs, DVD-ROMs, Blu-ray (registered trademark), etc. The communication unit 311 can also acquire the various programs via a communication network.

[0041] The analysis system 3 having the above-described functional configuration is a computer configured using one or more pieces of hardware such as a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), etc. The analysis system 3 may be configured separately from the electrocardiogram signal measuring device 2, or may be configured integrally with the electrocardiogram signal measuring device 2.

[0042] The receiving terminal 4 includes a communication unit 41 , an output unit 42 , and a control unit 43 .

[0043] The communication unit 41 receives, via the communication network, information output from the analysis system 3. The communication unit 41 receives, for example, the estimation result by the estimation unit 314 and wearer information from the analysis system 3.

[0044] The output unit 42 is configured by an output device such as a display device or a printer, and outputs various information under the control of the control unit 43. The output unit 42 has output functions such as a display made of, for example, a liquid crystal or organic EL display, and a speaker.

[0045] The control unit 43 comprehensively controls the operation of the receiving terminal 4. The control unit 43 also outputs the estimation result by the estimation unit 314 through the output function of the output unit .

[0046] The receiving terminal 4 having the above functional configuration is a computer configured using one or more pieces of hardware such as a CPU, a GPU, an ASIC, an FPGA, etc. The receiving terminal 4 may also have a user interface.

[0047] The learning device 5 is configured using one or more pieces of hardware such as a CPU, a GPU, an ASIC, an FPGA, etc. The learning device 5 is also configured to be able to transmit and receive data to and from the analysis system 3 via a communication network.

[0048] Next, a process for evaluating the data quality of the electrocardiographic signal 10 obtained from the subject 1 will be described. FIG. 3 is a flowchart showing the flow of the process for evaluating the data quality according to the first embodiment. In the process for evaluating the data quality, first, the communication unit 311 or the input / output unit 315 acquires the electrocardiographic signal 10 obtained from the subject 1 as biological information (step S101). The control unit 316 processes the waveform data 11 of the acquired electrocardiographic signal 10 as necessary to generate data to be evaluated. In the first embodiment, the input / output unit 315 has an input port for a medium to which the electrocardiographic signal 10 is written, and after acquisition of the electrocardiographic signal 10 is completed, the electrocardiographic signal 10 stored in the medium is read out as an example.

[0049] After acquiring the electrocardiogram signal 10, the analysis unit determination unit 312 divides the waveform data 11 in the electrocardiogram signal 10 into predetermined segments to determine the analysis units (step 102). The segment units may be time units obtained by dividing the data at a predetermined time length, or waveform units obtained by dividing the data every time a certain number of waveforms (e.g., electrocardiogram waveforms or pulse waveforms) in the biological information are detected. However, the segment length may be the same as the measurement length of the biological information. Furthermore, overlap can be freely set for the analysis units. In this first embodiment, the analysis unit determination unit 312 divides the data into time units of 60 seconds without setting overlap, as an example.

[0050] The index calculation unit 313 then calculates two or more evaluation indexes based on the electrocardiographic signal 10 in the determined analysis unit (step S103). The evaluation index here refers to a quantification of the data quality of the electrocardiographic signal 10 using any method. Examples of evaluation indexes include the similarity of the waveform components of the electrocardiographic signal 10 to a template (reference data), the wave height of each waveform component, the wave height of superimposed noise components, the spectral intensity in a predetermined band, the number of waveforms per specified unit time, and an S / N ratio calculated from the waveform features of the biological information and the features of the superimposed noise. The number and combination of evaluation indexes are not limited, but it is generally desirable to select evaluation indexes that are not multicollinear.

[0051] In this embodiment 1, the index calculation unit 313 calculates four parameters as evaluation indices (hereinafter referred to as "pattern A"): the average value of the combined wave height of R waves and S waves (hereinafter referred to as RS wave height) for each electrocardiogram waveform detected from the analysis unit; the similarity (hereinafter referred to as "DTW similarity") obtained by normalizing each electrocardiogram waveform detected from the analysis unit by RS wave height and applying dynamic time warping to the electrocardiogram waveform detected from the analysis unit using a predetermined template waveform as a comparison; the average value of the spectral intensity in the 0.1 to 0.5 [Hz] band (hereinafter referred to as "low-frequency spectral intensity"); and the average value of the spectral intensity in the 30 to 125 [Hz] band (hereinafter referred to as "high-frequency spectral intensity"); and a pattern (hereinafter referred to as "pattern B") that calculates two of the RS wave height and high-frequency spectral intensity as evaluation indices.

[0052] The estimation unit 314 estimates the quality index using the evaluation index and the estimation model (step S104). The estimation model is an estimation model constructed by the learning process in the learning device 5.

[0053] Here, the generation of an estimation model in the learning device 5 will be described with reference to Fig. 4 and Fig. 5. Fig. 4 is a flowchart showing the flow of an estimation model generation process for estimating a quality index according to the first embodiment.

[0054] The learning device 5 first acquires the electrocardiographic signals 10 (step S201). In the first embodiment, the learning device 5 is provided with an input port for a medium to which the electrocardiographic signals 10 are written, and after acquisition of the electrocardiographic signals 10 is completed, the learning device 5 reads all of the electrocardiographic signals 10 stored in the medium at once. In addition, it is preferable to acquire biological information from various specimens in the estimation model generation process.

[0055] After acquiring the electrocardiographic signal 10, the learning device 5 divides the electrocardiographic signal 10 into predetermined segments to acquire analysis units (step S202). Similar to the analysis unit determination unit 312, the learning device 5 acquires analysis units by dividing the electrocardiographic signal 10 into predetermined units, such as time units that divide the electrocardiographic signal 10 at a predetermined time length or waveform units that divide the electrocardiographic signal 10 every time a certain number of electrocardiographic waveforms are detected.

[0056] The learning device 5 then calculates two or more evaluation indexes based on the acquired electrocardiogram signal 10 in the analysis unit (step S203). The learning device 5 calculates at least two of the above-mentioned parameters as the evaluation indexes. For example, the learning device 5 calculates at least two of the RS wave height, DTW similarity, low-frequency spectrum intensity, and high-frequency spectrum intensity.

[0057] Thereafter, the learning device 5 performs a learning process based on the calculated evaluation index to generate an estimation model (step S204). Fig. 5 is a flowchart showing the flow of the learning process in the estimation model generation process according to the first embodiment.

[0058] First, the learning device 5 selects a kernel function for calculating the similarity between explanatory variables in the estimation model (step S301). Examples of the kernel function include a radial basis function (RBF) kernel, a linear kernel, a polynomial kernel, etc. Since the expressive power for linearity and periodicity and the calculation cost vary, it is preferable to select an appropriate one depending on the number and combination of explanatory variables, but this is not particularly limited. As a method for selecting a kernel function, in addition to artificial selection, it is also possible to compare the performance of multiple kernel functions by cross-validation and select the one with the highest performance, but this is not particularly limited. In the first embodiment, the RBF kernel is selected as the kernel function, but this is not particularly limited.

[0059] Next, the learning device 5 optimizes the model parameters in the estimation model (step S302). Specifically, the learning device 5 optimizes the parameters and hyperparameters of the kernel function. Each parameter differs depending on the selected kernel function. Methods for optimizing each parameter include, but are not limited to, maximum likelihood estimation, Bayesian optimization, cross-validation, and the like. Furthermore, the amount of training data required in the learning process is generally considered to be several hundred to several thousand pieces, but this varies depending on the selected estimation model, the characteristics of the dataset, and the purpose of estimation, and is not particularly limited.

[0060] The estimation model thus optimized and generated outputs a quality index by providing an evaluation index. The estimated quality index is, for example, a single value, a probability value having a value between 0 and 1. The interpretation of the quality index varies depending on the type of estimation model and the method for determining the teacher quality index used as the target variable in the learning process. For example, in the first embodiment, the teacher quality index is assigned one of two values, "1" for electrocardiogram signals of each analysis unit that the technician judges to be analyzable, and "0" for those judged to be unanalyzable, so the quality index is expected to represent the probability value of the technician's judgment that the signal is analyzable.

[0061] The construction of an estimation model is completed by the estimation model generation process and learning process according to the above steps. When the RBF kernel is selected as the kernel function as described above, the estimation model is constructed as a nonlinear model.

[0062] 3 , once the quality index is estimated in step S104, the control unit 316 transmits the estimation result by the estimation unit 314 and wearer information to the receiving terminal 4 (step S105). The receiving terminal 4 displays the quality index for the electrocardiogram signal 10. For example, the guard carrying the receiving terminal 4 checks the estimation result.

[0063] 6A and 6B are diagrams for explaining the estimation process according to the first embodiment. Fig. 6A shows the quality index in pattern A. Fig. 6B shows the quality index in pattern B.

[0064] In the first embodiment, the quality index represents the probability value that the technician determines that the signal is analyzable, and it can be interpreted that the higher the quality index, the higher the quality. As shown in Figures 6A and 6B, there is a clear difference between the quality index estimated from the electrocardiogram signals that the technician determines to be analyzable and the quality index estimated from the electrocardiogram signals that the technician determines to be unanalyzable, and the quality can be expressed as being superior or inferior between the analyzable and unanalyzable signals.

[0065] Furthermore, when comparing pattern A (see FIG. 6A), which uses four evaluation indexes as explanatory variables, with pattern B (see FIG. 6B), which uses two evaluation indexes as explanatory variables, pattern B not only has an extremely narrow interquartile range, but also contains outliers. This can be attributed to the fact that pattern B has fewer explanatory variables and misses changes in data quality, resulting in lower data quality interpretation performance compared to pattern A. In other words, in order to improve data quality interpretation performance, it is important to combine appropriate evaluation indexes to build an estimation model and estimate quality indexes.

[0066] According to the first embodiment described above, multiple evaluation indices are calculated for biometric information acquired from a living body, and the quality index estimated by the estimation model is estimated using multiple evaluation indices obtained by a wide variety of methods, thereby reducing the risk of overlooking changes in data quality.Furthermore, since the output value is a single value, simple evaluation is possible, and therefore a biometric information analysis system can be provided that can evaluate data quality with high accuracy and in a simple manner.

[0067] In the first embodiment, an example has been described in which a learning device 5 is provided separately from the analysis system 3 and an estimation model is constructed, but the analysis system 3 may have the functions of the learning device 5, or a trained estimation model may be directly loaded to skip the estimation model generation process.

[0068] (Embodiment 2) Next, a description will be given of embodiment 2 of the present invention. In embodiment 2, an analysis system 3A that acquires position information is provided instead of the analysis system 3 according to embodiment 1. Below, the same components as in embodiment 1 are given the same reference numerals, and differences will be described.

[0069] 7 is a diagram showing an example of the configuration of a biological information analysis system according to the present embodiment 2. The biological information analysis system according to the present embodiment 2 includes an electrocardiogram signal measuring device 2, an analysis system 3A, a receiving terminal 4, and a learning device 5.

[0070] The analysis system 3A includes a communication unit 311 , an analysis unit determination unit 312 , an index calculation unit 313 , an estimation unit 314 , an evaluation unit 318 , an input / output unit 315 , a control unit 316 , and a storage unit 317 .

[0071] The evaluation unit 318 compares the quality indexes of electrocardiogram signal measuring devices 2 having different configurations to evaluate the performance. The evaluation unit 318 outputs, for example, a comparison result obtained by comparing the performance of the configurations.

[0072] 8 is a flowchart showing the flow of the performance comparison process according to the second embodiment. First, the control unit 316 determines the configurations of the vital information measuring devices to be subjected to performance evaluation / comparison (step S401). In the second embodiment, an example will be described in which performance is compared for one electrocardiograph 100 using two configurations (hereinafter referred to as configuration A and configuration B) in which the physical properties of the electrodes are different, but the number, types, and combinations thereof are not particularly limited.

[0073] Next, the control unit 316 acquires the electrocardiographic signal 10 in each of the determined configurations (step S402). At this time, it is desirable that the source of the electrocardiographic signal 10 in each configuration (configuration A and configuration B) is the same individual, but this is not particularly limited. In the second embodiment, the case where the source of the electrocardiographic signal 10 is the same individual will be described as an example.

[0074] Next, a quality index is estimated for the acquired electrocardiogram signal 10 (step S403). In step S403, the quality index of each component is estimated according to steps S102 to S104 shown in Fig. 3. Note that in the second embodiment, the estimation unit 314 uses a trained estimation model that is common to each component, but this is not particularly limited.

[0075] The evaluation unit 318 then evaluates each configuration by comparing the estimated quality indexes for each configuration (step S404). In the second embodiment, the evaluation unit 318 compares the estimated quality indexes for configurations A and B, and generates a comparison result indicating that configuration A has higher performance (configuration A can acquire higher quality data) if the quality index for configuration A is higher than that of configuration B, and generates a comparison result indicating that configuration B has higher performance (configuration B can acquire higher quality data) if the quality index for configuration B is higher than that of configuration A.

[0076] When the comparison result is generated based on the quality index, the control unit 316 transmits the comparison result and wearer information to the receiving terminal 4 (step S405). The comparison result for configuration A and configuration B is displayed on the receiving terminal 4. For example, the guard carrying the receiving terminal 4 checks the comparison result and selects the configuration to use.

[0077] According to the second embodiment described above, similarly to the first embodiment, multiple evaluation indices are calculated for biometric information acquired from a living body, and the quality index estimated by the estimation model is estimated using multiple evaluation indices obtained by a wide variety of methods, thereby reducing the risk of overlooking changes in data quality. Furthermore, since the output value is a single value, simple evaluation is possible, and therefore a biometric information analysis system capable of evaluating data quality with high accuracy and in a simple manner can be provided.

[0078] Furthermore, in this second embodiment, quality index estimation is performed for each of the different configurations to generate a comparison result, and therefore the configurations are compared based on the quality index estimated with high accuracy as described above, allowing for appropriate selection of the configuration.

[0079] (Embodiment 3) Next, a third embodiment of the present invention will be described. The configuration of the biological information analysis system according to the third embodiment is the same as that of the biological information analysis system according to the second embodiment, and therefore a description thereof will be omitted. Below, processing different from that of the second embodiment will be described. In the third embodiment, the function of the evaluation unit 318 differs from that of the second embodiment, and the state of attachment of the biological information measuring device to the subject 1 is evaluated by evaluating the data quality of the biological information acquired from the subject 1.

[0080] 9 is a flowchart showing the flow of the evaluation process for the wearing state according to the present embodiment 3. First, the communication unit 311 acquires the electrocardiogram signal 10 obtained from the subject 1 via the electrocardiogram signal measuring device 2 (step S501). If necessary, the communication unit 311 processes the waveform data 11 of the acquired electrocardiogram signal 10 to generate data for the determination process.

[0081] Next, a quality index is estimated for the acquired electrocardiographic signal 10 (step S502). In step S502, the quality index is estimated according to the flow of steps S102 to S104 shown in Fig. 3. In the third embodiment, the electrocardiographic signal measuring device 2 transmits values ​​acquired in a batch manner at regular intervals, thereby performing the estimation in parallel with the acquisition of the electrocardiographic signal 10, and the estimation unit 314 uses a trained estimation model, but this is not particularly limited.

[0082] Next, the evaluation unit 318 evaluates the wearing state based on the estimated quality index (step S503). In the third embodiment, for example, when the quality index represents a probability value that the technician determines that the analysis is possible, as in the first embodiment, the evaluation unit 318 generates a determination result that determines the wearing state to be good if the quality index is greater than 0.6, and determines the wearing state to be bad if the quality index is 0.6 or less.

[0083] When the determination result is generated, the control unit 316 transmits the determination result and wearer information to the receiving terminal 4 (step S504). The determination result of the wearing state is displayed on the receiving terminal 4. For example, the guard carrying the receiving terminal 4 checks the determination result and adjusts the position of the electrocardiogram signal measuring device 2 (electrocardiograph 100), replaces it, etc.

[0084] According to the third embodiment described above, similar to the first embodiment, multiple evaluation indices are calculated for biometric information acquired from a living body, and the quality index estimated by the estimation model is estimated using multiple evaluation indices obtained by a wide variety of methods, thereby reducing the risk of overlooking changes in data quality. Furthermore, since the output value is a single value, simple evaluation is possible, and therefore a biometric information analysis system can be provided that can evaluate data quality with high accuracy and in a simple manner.

[0085] Furthermore, in the third embodiment, the state of attachment of the electrocardiogram signal measuring device 2 (electrocardiograph 100) is evaluated based on the estimated results of the quality index, thereby enabling more accurate measurements to be performed.

[0086] In this third embodiment, the biometric information analysis system may notify the subject 1, guards, etc. of the result of the determination of the wearing state by providing a notification function in the biometric information analysis system (e.g., the receiving terminal 4) and making the notification in the biometric information analysis system, or by transmitting the determination result to an external device having a notification function and making the notification in the external device, but this is not particularly limited.

[0087] Furthermore, the method for evaluating the wearing condition of a bioinformation measuring device according to this third embodiment is expected to be utilized, for example, when operating a wearable electrocardiograph using dry electrodes, in which the wearing condition is expected to deteriorate due to body movement, but is not particularly limited thereto.

[0088] (Fourth Embodiment) Next, a fourth embodiment of the present invention will be described. The configuration of the biological information analysis system according to the fourth embodiment is the same as that of the biological information analysis system according to the second embodiment, and therefore the description will be omitted. Below, processing different from that of the second embodiment will be described. In the fourth embodiment, the function of the evaluation unit 318 differs from that of the second embodiment, and the performance of the biological information measuring device is evaluated by evaluating the data quality of the acquired biological information, and an apparatus configuration suitable for the subject 1 is determined.

[0089] 10 is a flowchart showing the flow of the evaluation process for the wearing state according to the fourth embodiment. First, the control unit 316 determines candidate device configurations of the biological information measuring device to be applied to the subject 1 (step S601). In the fourth embodiment, for example, an electrocardiograph from company X is described as having two configurations (hereinafter referred to as configurations A and B) with different electrode physical properties, and an electrocardiograph from company Y (hereinafter referred to as configuration C), for a total of three device configurations as candidates, but the number, types, and combinations thereof are not particularly limited.

[0090] Next, the control unit 316 acquires the electrocardiographic signal 10 in each of the determined device configurations (configurations A to C) (step S602). At this time, in each configuration (configurations A to C), it is desirable that the source of the electrocardiographic signal 10 is the subject 1, but the conditions, such as a body type similar to that of the subject 1, are not particularly limited, and the source of the electrocardiographic signal 10 in each configuration does not have to be the same individual. In the fourth embodiment, an example will be described in which the source of the electrocardiographic signal 10 is the subject 1.

[0091] Next, a quality index is estimated for the acquired electrocardiogram signal 10 (step S603). In step S603, the quality index of each configuration is estimated according to the flow of steps S102 to S104 shown in Fig. 3. Note that in the fourth embodiment, the estimation unit 314 uses a trained estimation model that is common to each configuration, but this is not particularly limited.

[0092] Then, the evaluation unit 318 selects a device configuration of the biological information measuring device suitable for the subject 1 based on the quality index estimated for each configuration (step S604). In the fourth embodiment, for example, among configurations A to C, the configuration with the largest estimated quality index is determined to be the device configuration suitable for the subject 1, and the device configuration is selected. The evaluation unit 318 generates a selection result.

[0093] When the device configuration determination result is generated, the control unit 316 transmits the determination result and wearer information to the receiving terminal 4 (step S605). The comparison result for each configuration is displayed on the receiving terminal 4. For example, the guard carrying the receiving terminal 4 checks the selection result and selects the configuration to use.

[0094] According to the fourth embodiment described above, similarly to the first embodiment, multiple evaluation indices are calculated for biometric information acquired from a living body, and the quality index estimated by the estimation model is estimated using multiple evaluation indices obtained by a wide variety of methods, thereby reducing the risk of overlooking changes in data quality. Furthermore, since the output value is a single value, simple evaluation is possible, and therefore a biometric information analysis system capable of evaluating data quality with high accuracy and in a simple manner can be provided.

[0095] Furthermore, in this fourth embodiment, quality index estimation is performed for each of different device configurations, and the device configuration suitable for the subject 1 is selected, so that the configurations are compared based on the quality index estimated with high accuracy as described above, and therefore the configuration can be appropriately selected.

[0096] In the fourth embodiment, the method for determining the device configuration is expected to be utilized, for example, when a medical institution such as a hospital provides a vital information measuring device to a patient, but is not particularly limited to this.

[0097] In the present embodiments 2 to 4, an example has been described in which the analysis system 3 is equipped with the evaluation unit 318, but various evaluation processes may also be performed in a device separate from the analysis system 3 that has the functions of the evaluation unit 318.

[0098] Other Embodiments While the embodiments for carrying out the present invention have been described above, the present invention should not be limited to the above-described embodiments. For example, although the analysis system has been described as being provided separately from the electrocardiogram signal measuring device 2, the analysis system may be configured integrally with the electrocardiogram signal measuring device 2.

[0099] The biometric information analysis system, biometric information analysis program, method for evaluating device performance, method for evaluating wearing status, and method for selecting device configuration according to the present invention are suitable for easily and accurately evaluating the data quality of biometric information obtained from a living body.

[0100] REFERENCE SIGNS LIST 1 Subject 2 Electrocardiogram signal measuring device 3, 3A Analysis system 4 Receiving terminal 5 Learning device 10 Electrocardiogram signal 11 Waveform data 41, 311 Communication unit 42 Output unit 43, 316 Control unit 312 Analysis unit determination unit 313 Index calculation unit 314 Estimation unit 315 Input / output unit 317 Storage unit 318 Evaluation unit

Claims

1. A biometric information analysis system for estimating data quality using a target biometric information, comprising: an analysis unit determination unit that generates unit biometric information by dividing the biometric information at preset segment lengths; an index calculation unit that quantifies data quality for the unit biometric information as multiple evaluation indexes; and an estimation unit that inputs multiple evaluation indexes based on the target biometric information into an estimation model that is generated by learning using the unit biometric information and the multiple evaluation indexes, and estimates a quality index.

2. The bioinformation analysis system according to claim 1, wherein the estimation model is obtained by learning using the plurality of evaluation indexes as explanatory variables and the teacher quality index as a target variable.

3. The bioinformation analysis system according to claim 2, wherein the teacher quality index is a binary label indicating whether or not the analysis is possible.

4. The bioinformation analysis system according to claim 1, wherein the estimation model is a nonlinear model.

5. The bioinformation analysis system of claim 1, wherein the plurality of evaluation indices include one or more of the following: similarity to reference data, spectral intensity in a predetermined band, wave height of the waveform in the bioinformation, wave height of superimposed noise, number of waveforms per predetermined time, and an S / N ratio calculated from features of the waveform in the bioinformation and features of the superimposed noise.

6. The biological information analysis system according to claim 5, wherein the plurality of evaluation indices include at least a spectral intensity in a predetermined low frequency band and a spectral intensity in a predetermined high frequency band.

7. The biological information analysis system according to claim 5, wherein the plurality of evaluation indices include at least a similarity to reference data, a spectral intensity in a predetermined low frequency band, and a spectral intensity in a predetermined high frequency band.

8. A biometric information analysis program for estimating data quality using a target biometric information, the biometric information analysis program causing a computer to execute the following steps: read a preset segment length from a storage unit; generate unit biometric information by dividing the biometric information at the segment length; quantify data quality for the unit biometric information as multiple evaluation indices; input the multiple evaluation indices based on the target biometric information into an estimation model generated by learning using the unit biometric information and the multiple evaluation indices, and estimate a quality index.

9. A method for evaluating device performance of a device that acquires biometric information from a subject, comprising: a step of acquiring the biometric information from the subject using the device; a step of estimating the quality index of the acquired biometric information using a biometric information analysis system described in any one of claims 1 to 7; and a step of evaluating device performance based on the quality index.

10. A method for evaluating the wearing state of a device that acquires biometric information from a subject, comprising: a step of acquiring the biometric information from the subject using the device; a step of estimating the quality index of the acquired biometric information using a biometric information analysis system described in any one of claims 1 to 7; a step of evaluating the wearing state of the device based on the quality index; and a step of notifying the evaluation result.

11. A method for selecting an optimum device configuration for a device that acquires biometric information from a subject, the method comprising the steps of: acquiring the biometric information from the subject using a default device configuration for the device; estimating the quality index of the acquired biometric information using a biometric information analysis system described in any one of claims 1 to 7; and selecting the optimum device configuration for the subject based on the quality index.

Images