State estimation device, state estimation method, and program

Abstract

One aspect of the present invention is a state estimation device that comprises: a heart state time series acquisition unit that acquires a heart state time series that is a time series of heart state quantities that indicate the state of the heart of an estimation target; and a heart state estimation unit that estimates the state of the heart of the estimation target on the basis of the occurrence time of out-of-range data that, of refractory period samples that are samples of refractory periods from among samples of the heart state time series acquired by the heart state time series acquisition unit, are refractory period samples for which a value is outside the range of a threshold area for processing determined in accordance with the distribution of the refractory period samples.

Description

[Technical field]

[0001] The present invention relates to a state estimation device, a state estimation method, and a program. [Background technology]

[0002] In recent years, biosignals such as cardiac potential and heart rate have become easily accessible through inexpensive and small wearable devices (Non-Patent Document 1). [Prior art documents] [Non-patent literature]

[0003] [Non-Patent Document 1] Kasai, Ogasawara, Nakajima, and Tsukada, "Development and Practical Use of the Functional Material "hitoe" that Enables Bioinformation Measurement by Simply Wearing It," Institute of Electronics, Information and Communication Engineers, Communication Society Magazine, Vol. 41 (June 2017) (Vol. 11 No. 1) [Non-Patent Document 2] "Current Affairs Medical: Ventricular Tachycardia, Ventricular Fibrillation (shinshitsuhinpaku, shinshitsusaido)", [online], [searched on February 10, 2021], Internet<https: / / medical.jiji.com / medical / 012-1027> Summary of the Invention [Problem to be solved by the invention]

[0004] However, reliable means of detecting heart disease are limited to large medical devices such as Holter monitors, and this has not yet been realized with wearable devices.

[0005] One of the reasons is the limited specifications of wearable devices. If a small and inexpensive device is used, only limited calculation processing can be performed, so it is not possible to perform the heart disease detection processing implemented in a Holter electrocardiograph. In other words, a wearable device capable of estimating heart abnormalities has not been realized because the amount of calculation required to estimate the state of the heart, such as whether or not there is an abnormality in the heart, is large. When using existing methods that require a large amount of calculation, a solution can be imagined in which the amount of data is reduced to a capacity that can be implemented in a wearable device. However, since heart diseases tend to occur suddenly and in a short period of time, if the amount of data is reduced, the possibility of missing such rare trends increases. As a result, there is a risk that serious situations that may be life-threatening will not be detected in an emergency. Therefore, for the benefit of users, a technology is required that reduces the amount of calculation required for estimation while not reducing the large amount of data sampled at high frequencies.

[0006] In view of the above circumstances, an object of the present invention is to provide a technique for reducing the amount of calculation required to estimate the state of the heart. [Means for solving the problem]

[0007] One aspect of the present invention is a state estimation device comprising: a cardiac state time series acquisition unit that acquires a cardiac state time series, which is a time series of cardiac state quantities, which are quantities that indicate the cardiac state of a subject to be estimated; and a cardiac state estimation unit that estimates the cardiac state of the subject to be estimated based on the occurrence time of out-of-range data, for refractory period samples, which are refractory period samples among the cardiac state time series samples acquired by the cardiac state time series acquisition unit, and which estimates the cardiac state of the subject to be estimated based on the occurrence time of the out-of-range data, for refractory period samples whose values ​​are outside a processing threshold region that is determined in accordance with the distribution of the refractory period samples.

[0008] One aspect of the present invention is a state estimation device comprising: a cardiac state time series acquisition unit that acquires a cardiac state time series, which is a time series of cardiac state quantities, which are quantities that indicate the cardiac state of a subject to be estimated; and a cardiac state estimation unit that estimates the cardiac state of the subject to be estimated based on a time interval RRI (RR-Interval) of R waves in the cardiac state time series.

[0009] One aspect of the present invention is a state estimation method comprising: a cardiac state time series acquisition step of acquiring a cardiac state time series, which is a time series of cardiac state quantities, which are quantities indicative of the cardiac state of a subject to be estimated; and a cardiac state estimation step of estimating the cardiac state of the subject to be estimated based on the occurrence time of out-of-range data, for refractory period samples, which are refractory period samples among the cardiac state time series samples acquired in the cardiac state time series acquisition step, and which estimate the cardiac state of the subject to be estimated based on the occurrence time of the out-of-range data.

[0010] One aspect of the present invention is a program for causing the above-described state estimation device to function as a computer. Effect of the Invention

[0011] According to the present invention, the amount of calculation required to estimate the state of the heart can be reduced. [Brief description of the drawings]

[0012] [Figure 1] FIG. 1 is an explanatory diagram illustrating an overview of an abnormal state estimation system 100 according to an embodiment. [Diagram 2] FIG. 4 is a diagram showing an example of a cardiac state quantity time series obtained from a heart in a normal state in the embodiment. [Diagram 3] FIG. 4 is a diagram showing an example of a cardiac state quantity time series obtained from a heart in an abnormal state in the embodiment. [Figure 4] 5A and 5B are diagrams showing upper and lower thresholds, threshold regions, and out-of-range data in the embodiment. [Diagram 5] FIG. 2 is a diagram showing an example of a hardware configuration of a monitoring device 4 in the embodiment. [Figure 6]FIG. 2 is a diagram showing an example of the functional configuration of a control unit 41 in the embodiment. [Figure 7] FIG. 2 is a diagram showing an example of a hardware configuration of a control device 5 in the embodiment. [Figure 8] FIG. 2 is a diagram showing an example of the functional configuration of a control unit 51 in the embodiment. [Figure 9] 3 is a flowchart showing an example of a flow of processing executed by the abnormal state estimation system 100 of the embodiment. [Figure 10] FIG. 11 is an explanatory diagram illustrating an effect achieved by a fourth type mental state estimation process in the modified example. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] (Embodiment) FIG. 1 is an explanatory diagram illustrating an overview of an abnormal condition estimation system 100 according to an embodiment. The abnormal condition estimation system 100 estimates an abnormality in the heart of an estimation target 9. The estimation target 9 may be any living organism having a heart, such as a human. The estimation target 9 may also be an animal other than a human. The estimation target 9 includes a biosignal acquisition device 1.

[0014] The biosignal acquiring device 1 acquires information (hereinafter referred to as "cardiac state signal") on a time series (hereinafter referred to as "cardiac state time series") of quantities (hereinafter referred to as "cardiac state quantities") indicating the cardiac state of the subject 9 to be estimated, such as a time series of cardiac potential and a time series of heart rate. The biosignal acquiring device 1 is, for example, a wearable device capable of acquiring a cardiac state signal and worn by the subject 9 to be estimated. The biosignal acquiring device 1 is, for example, a device equipped with a cardiac potential sensor that detects the cardiac potential from the subject 9 to be estimated via conductive electrodes. The biosignal acquiring device 1 repeatedly acquires the cardiac state quantities of the subject 9 to be estimated at a predetermined time interval shorter than a unit processing period described later.

[0015] The abnormal condition estimation system 100 estimates whether or not an abnormality has occurred in the heart of the estimation target 9 based at least on the cardiac state signal acquired by the biological signal acquisition device 1. The abnormal condition estimation system 100 estimates whether or not an abnormality has occurred in the heart of the estimation target 9, for example, while riding in an automobile 90. For ease of explanation, the abnormal condition estimation system 100 will be described below using as an example a case in which a cardiac abnormality is estimated for the estimation target 9 while riding in the automobile 90.

[0016] The abnormal condition estimation system 100 includes a biosignal acquiring device 1, a relay terminal 2, an environmental sensor 3, a monitoring device 4, and a control device 5. The biosignal acquiring device 1 outputs the acquired cardiac state signal to the relay terminal 2.

[0017] The relay terminal 2 is a device that transmits the cardiac state signal acquired by the biological signal acquiring device 1 to the monitoring device 4. The relay terminal 2 is, for example, a device equipped with an antenna that transmits the cardiac state signal. The relay terminal 2 may be, for example, a mobile terminal such as a smartphone or tablet that acquires the cardiac state signal from the biological signal acquiring device 1 and transmits it.

[0018] The relay terminal 2 converts the cardiac state signal, for example, from an analog signal to a digital signal. The cardiac state signal does not necessarily have to be transmitted from the relay terminal 2 in the form of a digital signal, and may be transmitted in the form of an analog signal. The conversion of the cardiac state signal from an analog signal to a digital signal does not necessarily have to be performed by the relay terminal 2, and may be performed by the biosignal acquiring device 1. The conversion of the cardiac state signal from an analog signal to a digital signal may be performed by the monitoring device 4. For simplicity of explanation, the abnormal state estimation system 100 will be explained below taking as an example a case in which the cardiac state signal is transmitted from the relay terminal 2 in the form of a digital signal.

[0019] The environmental sensor 3 is a sensor that acquires information (hereinafter referred to as "environmental information") regarding either or both of the state of motion of the estimation target 9 and the environment in which the estimation target 9 exists. The environmental sensor 3 is, for example, a speedometer that measures the speed of movement of the estimation target 9. In such a case, the environmental information indicates the speed of movement of the estimation target 9. The environmental sensor 3 may be, for example, a temperature sensor that measures the temperature of the space in which the estimation target 9 exists. In such a case, the environmental information indicates the temperature of the space in which the estimation target 9 exists.

[0020] The environmental sensor 3 may be, for example, an acceleration sensor that measures the acceleration of the movement of the estimation target 9. In such a case, the environmental information indicates the acceleration of the movement of the estimation target 9. Note that the environmental sensor 3 does not necessarily need to indicate only one type of information, and may indicate multiple types of information. For example, the environmental sensor 3 may indicate the speed of the movement of the estimation target 9 and the temperature of the space in which the estimation target 9 exists.

[0021] The environmental sensor 3 may be, for example, a sensor mounted on the automobile 90, and may be a sensor that acquires information indicating the state of the automobile 90 (hereinafter referred to as "on-vehicle information"), such as an acceleration sensor, a temperature sensor, or a speedometer mounted on the automobile 90. The on-vehicle information is an example of environmental information.

[0022] The environmental sensor 3 may be, for example, an acceleration sensor. The environmental sensor 3 does not necessarily need to be implemented as a device different from the biosignal acquisition device 1, and may be included in the biosignal acquisition device 1. The environmental sensor 3 may be implemented as a device worn by the estimation target 9, or may be included in the automobile 90 in which the estimation target 9 is riding.

[0023] The environmental sensor 3 transmits the acquired environmental information to the monitoring device 4.

[0024] The monitoring device 4 acquires the cardiac state signal and environmental information. The monitoring device 4 estimates a cardiac abnormality of the estimation target 9 based on at least the cardiac state signal. Hereinafter, the process in which the monitoring device 4 estimates a cardiac abnormality of the estimation target 9 based on at least the cardiac state signal is referred to as a cardiac state estimation process. The cardiac state estimation process is, for example, a first type cardiac state estimation process described later.

[0025] Based on the estimation result of the monitoring device 4, the control device 5 judges whether the estimation result of the monitoring device 4 satisfies a notification criterion, which is a predetermined criterion. Specifically, the notification criterion is a predetermined criterion for judging whether the estimation result of the monitoring device 4 regarding the heart state of the estimation target 9 should be notified to a predetermined notification destination. When the estimation result satisfies the warning notification criterion, the control device 5 notifies the predetermined notification destination that the heart of the estimation target 9 is abnormal. Hereinafter, the process of judging whether the estimation result of the monitoring device 4 satisfies the notification criterion is referred to as a notification judgment process.

[0026] <Explanation of the first type of mental state estimation process> The first type of mental state estimation process will be described. The first type of mental state estimation process includes a statistics calculation process and an abnormality estimation process. The statistics calculation process is repeatedly executed in a predetermined cycle. Hereinafter, the length of one cycle in which the statistics calculation process is executed is referred to as a unit processing period. The length of the unit processing period is, for example, 2 seconds.

[0027] The statistics calculation process is a process for calculating statistics (hereinafter referred to as "mental state statistics") related to the mental state time series indicated by the mental state signal. The mental state statistics are, for example, the time average of the mental state quantity. The statistics of the mental state time series are, for example, the deviation of the distribution of the mental state quantity. The deviation may be any quantity that indicates a difference from the average value. Therefore, the deviation may be, for example, a variance. The deviation may be, for example, a standard deviation.

[0028] In the statistics calculation process, statistics regarding the cardiac state time series are obtained using samples that satisfy a predetermined condition (hereinafter referred to as the "sample condition"). The sample condition is, for example, a condition that all samples included in the cardiac state signal acquired by the monitoring device 4 during a unit processing period immediately before the execution of the statistics calculation process are used. Therefore, for example, if the unit processing period is 2 seconds, the number of samples used in the statistics calculation process is all samples included in the cardiac state signal acquired by the monitoring device 4 during the most recent 2 seconds.

[0029] The abnormality estimation process is a process for estimating whether or not the cardiac condition of the estimation target 9 is abnormal. An abnormality of the estimation target by the abnormality estimation process is, for example, ventricular fibrillation. The abnormality estimation process includes a refractory period sample determination process, an out-of-range data determination process, and a ventricular abnormality determination process.

[0030] In order to facilitate understanding of the refractory period sample determination process, the out-of-range data determination process, and the ventricular abnormality determination process, a cardiac state quantity time series obtained from a heart in a normal state and a cardiac state quantity time series obtained from a heart in an abnormal state will be explained.

[0031] Fig. 2 is a diagram showing an example of a cardiac state quantity time series obtained from a heart in a normal state in an embodiment. More specifically, Fig. 2 is a diagram showing an example of a cardiac potential time series obtained from a heart in a normal state. The vertical axis of Fig. 2 indicates the potential of the cardiac potential, and the horizontal axis indicates time.

[0032] When the heart is beating normally, electrocardiogram waveforms including R waves are observed. The black circles in Fig. 2 indicate R waves. A, B, and C in Fig. 2 each indicate a type of activity period related to polarization when the heart is beating. Hereinafter, a period of type A will be referred to as period A. Hereinafter, a period of type B will be referred to as period B. Hereinafter, a period of type C will be referred to as period C.

[0033] Period A is the polarization period of the myocardium. In period A, mainly the R waveform is observed. Period B is the absolute refractory period. Period B is the period immediately after myocardial polarization. In period B, due to the principle of myocardium, if the condition of the heart is normal, there is no cardiac potential generation corresponding to the waveform. Period C is the relative refractory period. In period C, if the condition of the heart is normal, there is no waveform due to the constant rhythm of the beating trend. In other words, if the condition of the heart is normal, polarization is repeated periodically, but since period C is the period between the repeated polarizations, there is no waveform. Note that when periods B and C are not distinguished from each other, they are generally called refractory periods.

[0034] Thus, in the case of a time series of cardiac potentials obtained from a normal heart, it is possible to distinguish between periods A, B, and C. Also, in the case of a time series of cardiac potentials obtained from a normal heart, the voltage change from 0 millivolts is smaller in the refractory period of the cardiac potential (i.e., periods B and C) than in period A, in which an R wave occurs. The range of voltage change in period A of the time series of cardiac potentials obtained from a normal heart is generally referred to as the physiologically normal range of repolarization potential change.

[0035] Fig. 3 is a diagram showing an example of a cardiac state quantity time series obtained from a heart in an abnormal state in an embodiment. Specifically, Fig. 3 is a diagram showing an example of a time series of cardiac potentials obtained from a heart in an abnormal state. More specifically, Fig. 3 is a diagram showing an example of a time series of cardiac potentials obtained from a heart in a state of ventricular fibrillation. The vertical axis of Fig. 3 indicates the potential of the cardiac potential, and the horizontal axis indicates time.

[0036] A, B, and C in Fig. 3 respectively indicate period A, period B, and period C. Fig. 3 shows that in the cardiac potential during ventricular fibrillation, even in periods equivalent to the refractory periods in normal cardiac potential (i.e., periods B and C), the behavior of the cardiac potential falls outside the range of physiologically normal repolarization potential changes.

[0037] The abnormal condition estimation system 100 is a system that estimates whether the cardiac condition of the estimation target 9 is normal or abnormal based on the difference in behavior of the cardiac potential that exists between a normal heart and an abnormal heart. The out-of-range data determination process executed in the abnormal condition estimation system 100 is a process executed to quantify the degree of occurrence of out-of-range data in a section corresponding to the refractory period of a normal cardiac potential by using statistics.

[0038] The refractory period sample determination process, the out-of-range data determination process, and the ventricular abnormality determination process will each be described.

[0039] The refractory period sample determination process is a process for determining which samples in the cardiac state time series belong to the refractory period. For example, the refractory period sample determination process is a process for determining that a sample that satisfies a predetermined condition belongs to the refractory period.

[0040] The predetermined condition is, for example, a condition that the sample exceeds a predetermined threshold. The threshold is specifically a statistic of the mental state time series within a predetermined interval. The statistic is, for example, the sum of a predetermined representative value and a predetermined degree of dispersion. The statistic may be, for example, the difference between the predetermined representative value and a predetermined degree of dispersion. The representative value is, for example, the average value. The degree of dispersion is, for example, the standard deviation.

[0041] However, a sample of the cardiac state time series may momentarily exceed the threshold. Therefore, in the refractory period sample determination process, it may be determined whether or not a predetermined condition for the number of times the sample consecutively crosses the threshold is satisfied. In the refractory period sample determination process, if the predetermined condition for the number of times the sample consecutively crosses the threshold is satisfied, the sample is determined to belong to the refractory period.

[0042] The threshold value may be, for example, a mental state statistic obtained by a statistic calculation process.

[0043] For the sake of simplicity, the abnormal state estimation system 100 will be described taking as an example a case where the refractory period sample determination process is a process for determining which samples of the cardiac state time series belong to the refractory period based on the cardiac state statistics acquired by the statistics calculation process. Note that, when the refractory period sample determination process is a process for determining that a sample that satisfies a predetermined condition is a sample that belongs to the refractory period, the statistics calculation process does not necessarily need to be executed.

[0044] The out-of-range data determination process is executed on samples that have been determined to be samples belonging to the refractory period by the refractory period sample determination process (hereinafter referred to as "refractory period samples").

[0045] The out-of-range data determination process is a process for determining whether or not the value of each refractory period sample is outside a range corresponding to each time position (hereinafter referred to as the "threshold region"). The time position is the position of each sample in the cardiac state time series in the time axis direction. Hereinafter, a refractory period sample whose value (i.e., cardiac state quantity) is determined to be outside the threshold region by the out-of-range data determination process is referred to as out-of-range data.

[0046] A threshold region is a range having at least an upper limit and a lower limit. The upper limit of the threshold region is hereinafter referred to as the upper threshold. The lower limit of the threshold region is hereinafter referred to as the lower threshold.

[0047] The threshold region is determined for each unit processing period according to the distribution of refractory period samples within the unit processing period. For example, the upper threshold is (M+V) when the average value of the cardiac state quantity indicated by the refractory period samples within the unit processing period including the time position for which the threshold region is determined is M and the standard deviation is V. For example, the lower threshold is (MV) when the average value of the cardiac state quantity indicated by the refractory period samples within the unit processing period is M and the standard deviation is V.

[0048] The upper threshold and the lower threshold are not necessarily limited to the sum or difference of the mean value M and the standard deviation V. The upper threshold and the lower threshold may be the sum or difference of values ​​adjusted according to the detection sensitivity by multiplying the standard deviation V by a constant (correction value). The upper threshold and the lower threshold may be the result of conversion by a predetermined function with the mean value M and the standard deviation V as independent variables.

[0049] The upper and lower thresholds may be calculated based on the variance or gradient of the cardiac state quantity. The upper and lower thresholds may be calculated based on the amount of adjustment due to equipment or environmental data other than the biosignal, or continuity (presence or absence of missing observations). Being outside the threshold range means that the value is either less than the lower threshold or greater than the upper threshold.

[0050] FIG. 4 is a diagram showing an upper threshold, a lower threshold, a threshold region, and out-of-range data in an embodiment. FIG. 4 shows a cardiac potential time series as an example of a cardiac state time series. The horizontal axis of FIG. 4 indicates the elapsed time from the origin time. The vertical axis of FIG. 4 indicates the cardiac potential. FIG. 4 shows the upper threshold and the lower threshold. The upper threshold and the lower threshold in the example of FIG. 4 are an example of values ​​calculated using cardiac potential data for the most recent two seconds. Therefore, as shown in FIG. 4, the upper threshold and the lower threshold are not necessarily the same at all times.

[0051] In Fig. 4, the ranges of cardiac potentials indicated by D1, D2, and D3 are threshold regions at time T1, time T2, and time T3, respectively. As shown in Fig. 4, the ranges of cardiac potentials indicated by the threshold regions are not necessarily the same at all times. Fig. 4 shows a set of refractory period samples determined to be out-of-range data.

[0052] The ventricular abnormality determination process is a process for estimating the state of the ventricle based on the samples determined to be out-of-range data by the out-of-range data determination process. The ventricular abnormality determination process is a process for determining that the ventricular state is abnormal when a condition indicating a predetermined manner in which the peak period appears and indicating a manner in which the peak period appears when the ventricular state is abnormal (hereinafter referred to as a "peak period appearance condition") is satisfied.

[0053] The peak period is the peak determination target period during which the out-of-range data integrated time exceeds the threshold time. The out-of-range data integrated time is a value obtained for each peak determination target period. The out-of-range data integrated time is the integrated value of the occurrence time of samples that are determined to be out-of-range data by the out-of-range data determination process among the samples in each peak determination target period. In other words, the out-of-range data integrated time is the result of assigning a predetermined time width to each sample and multiplying the time width by the number of samples that are determined to be out-of-range data among the samples in each peak determination target period.

[0054] The peak determination period is a period of a predetermined length. The start time of the peak determination period is a time that satisfies a predetermined condition. The start time of the peak determination period is, for example, the end time of the immediately preceding peak determination period. The start time of the peak determination period may be, for example, a condition that a predetermined time has elapsed since the immediately preceding peak determination period.

[0055] The condition that the time is a predetermined time after the immediately preceding peak determination target period means that the peak determination target period is set periodically in the ventricular abnormality determination process. In the ventricular abnormality determination process, for example, first, the period from 0 milliseconds to 200 milliseconds in the cardiac state time series is set as the peak determination target period, and it is determined whether or not it is a peak period. In the ventricular abnormality determination process, the time 200 milliseconds is then set as a new 0 millisecond, and the subsequent 200 millisecond period is set as a new peak determination target period, and this process is repeated.

[0056] The threshold time is a predetermined reference time for detecting a value that does not occur in a cardiac state time series of a normal heart. More specifically, the threshold time is a predetermined reference time that is longer than an out-of-range data integration time in a cardiac state time series of a normal heart. Since the threshold time is longer than an out-of-range data integration time in a cardiac state time series of a normal heart, a peak determination target period in which the out-of-range data integration time exceeds the threshold time is a period in which an abnormal cardiac state time series appears.

[0057] For example, when the cardiac state time series is acquired at a sampling rate of 200 Hz and three points are out-of-range data, the length of the peak period is 15 milliseconds, which is three times 5 milliseconds. Note that the time interval between each sample in a time series with a sampling rate of 200 Hz is 5 milliseconds. When the cardiac state time series is acquired at a sampling rate of 200 Hz, the occurrence time of a sample, i.e., the predetermined time width given to a sample, is, for example, 5 milliseconds.

[0058] The length of the peak determination period is preferably approximately the same as the length of one beat of the heartbeat, and therefore the length of the peak determination period is, for example, 200 milliseconds.

[0059] The threshold time is, for example, a time longer than the occurrence time of an R wave of a normal heart, for example, 50 milliseconds.

[0060] An example of the process executed in the ventricular abnormality determination process will be specifically described in the case where the threshold time is 50 milliseconds, the length of the peak determination target period is 200 milliseconds, and the peak determination target period is set periodically every 200 milliseconds. In this case, the ventricular abnormality determination process executes a process of determining, as a peak period, a peak determination target period in which the cumulative time of refractory period samples in which the cardiac state quantity is determined to be outside the threshold region is 50 milliseconds or more, among each peak determination target period repeated periodically at 200 millisecond intervals.

[0061] The peak period appearance condition is, for example, a condition that a predetermined number of peak periods appear consecutively. When the peak period appearance condition is a condition that a predetermined number of peak periods appear consecutively, the heart of the estimation target 9 is in a state in which ventricular flutter or ventricular fibrillation has occurred. The number of consecutive peak periods is a predetermined value that is set in advance, but it is desirable that the value be determined by taking into account, for example, the frequency of erroneous determination and the time required to obtain a determination result.

[0062] More specifically, it is desirable that the value is one that achieves both a low frequency of erroneous judgment and a short time required to obtain the judgment result. The time required to obtain the judgment result is desirably a time that allows a notification to be made to prevent damage that may occur due to the heart of the estimation target 9 becoming abnormal. The number of consecutive peak periods is, for example, five times.

[0063] In this way, the first type cardiac state estimation process is a process for estimating the cardiac state of the estimation subject 9 based on the occurrence time of out-of-range data.

[0064] 5 is a diagram showing an example of a hardware configuration of the monitoring device 4 in an embodiment. The monitoring device 4 includes a control unit 41 including a processor 91 such as a CPU (Central Processing Unit) and a memory 92 connected by a bus, and executes a program. The monitoring device 4 functions as a device including the control unit 41, an input unit 42, a communication unit 43, a storage unit 44, and an output unit 45 by executing the program.

[0065] More specifically, the processor 91 reads out a program stored in the storage unit 44, and stores the read out program in the memory 92. The processor 91 executes the program stored in the memory 92, whereby the monitoring device 4 functions as a device including a control unit 41, an input unit 42, a communication unit 43, a storage unit 44, and an output unit 45.

[0066] The control unit 41 controls the operation of various functional units included in the monitoring device 4. The control unit 41 executes, for example, a cardiac state estimation process. The control unit 41 controls, for example, the operation of the output unit 45. The control unit 41 records, for example, various pieces of information generated by the execution of the cardiac state estimation process in the storage unit 44. The control unit 41 records, for example, a cardiac state time series indicated by a cardiac state signal input to the input unit 42 or the communication unit 43 in the storage unit 44.

[0067] The input unit 42 includes input devices such as a mouse, a keyboard, and a touch panel. The input unit 42 may be configured as an interface that connects these input devices to the monitoring device 4. The input unit 42 accepts input of various information to the monitoring device 4. For example, a heart state signal is input to the input unit 42. For example, environmental information may be input to the input unit 42.

[0068] The communication unit 43 includes a communication interface for connecting the monitoring device 4 to an external device. The communication unit 43 communicates with the external device via wired or wireless communication. The external device is, for example, a device that transmits the cardiac state signal. The transmitter of the cardiac state signal is, for example, the relay terminal 2. The external device is, for example, the control device 5. The communication unit 43 may communicate with the environmental sensor 3. When the communication unit 43 communicates with the environmental sensor 3, the communication unit 43 may acquire environmental information acquired by the environmental sensor 3 through communication with the environmental sensor 3.

[0069] The storage unit 44 is configured using a computer-readable storage medium device such as a magnetic hard disk device or a semiconductor storage device. The storage unit 44 stores various information related to the monitoring device 4. The storage unit 44 stores, for example, information input via the input unit 42 or the communication unit 43. The storage unit 44 stores, for example, various information generated by the execution of the cardiac state estimation process.

[0070] It should be noted that the mental state signal and the environmental information do not necessarily have to be input only to the input unit 42, and do not necessarily have to be input only to the communication unit 43. The mental state signal and the environmental information may be input from either the input unit 42 or the communication unit 43.

[0071] The output unit 45 outputs various information. The output unit 45 includes a display device such as a CRT (Cathode Ray Tube) display, a liquid crystal display, or an organic EL (Electro-Luminescence) display. The output unit 45 may be configured as an interface that connects these display devices to the monitoring device 4. The output unit 45 outputs, for example, information input to the input unit 42. The output unit 45 may display, for example, a result of execution of the cardiac state estimation process.

[0072] 6 is a diagram showing an example of a functional configuration of the control unit 41 in the embodiment. The control unit 41 includes a mental state time series acquisition unit 410, a mental state estimation unit 420, a memory control unit 430, a communication control unit 440, an output control unit 450, and an environmental information acquisition unit 460.

[0073] The cardiac state time-series acquiring section 410 repeatedly acquires the cardiac state time-series signal at a predetermined cycle via the input section 42 or the communication section 43. That is, the cardiac state time-series acquiring section 410 acquires the cardiac state time series.

[0074] The cardiac state estimation unit 420 estimates the cardiac state of the estimation target 9 based on the cardiac state time series indicated by the cardiac state signal acquired by the cardiac state time series acquisition unit 410. The cardiac state estimation unit 420 estimates the cardiac state of the estimation target 9, for example, by performing a cardiac state estimation process on the cardiac state time series indicated by the cardiac state signal acquired by the cardiac state time series acquisition unit 410. The cardiac state estimation process executed by the cardiac state estimation unit 420 is, for example, a first type cardiac state estimation process.

[0075] The memory control unit 430 records various information in the memory unit 44. The communication control unit 440 controls the operation of the communication unit 43. The communication control unit 440 controls the operation of the communication unit 43 to cause the communication unit 43 to transmit, for example, the estimation result of the mental state estimation unit 420 to the control device 5. The output control unit 450 controls the operation of the output unit 45. For example, the output control unit 450 controls the operation of the output unit 45 to cause the output unit 45 to output the estimation result of the mental state estimation unit 420.

[0076] The environmental information acquiring section 460 repeatedly acquires environmental information at a predetermined cycle via the input section 42 or the communication section 43. That is, the mental state time-series acquiring section 410 acquires environmental information.

[0077] 7 is a diagram showing an example of a hardware configuration of the control device 5 in the embodiment. The control device 5 includes a control unit 51 including a processor 93 such as a CPU and a memory 94 connected by a bus, and executes a program. The control device 5 functions as a device including the control unit 51, an input unit 52, a communication unit 53, a storage unit 54, and an output unit 55 by executing the program.

[0078] More specifically, the processor 93 reads out a program stored in the storage unit 44 and stores the read out program in the memory 94. The processor 93 executes the program stored in the memory 94, whereby the control device 5 functions as a device including a control unit 51, an input unit 52, a communication unit 53, a storage unit 54, and an output unit 55.

[0079] The control unit 51 controls the operation of various functional units included in the control device 5. The control unit 51 executes, for example, a notification determination process. The control unit 51 controls, for example, the operation of the communication unit 53. The control unit 51 controls, for example, the operation of the communication unit 53 to send a notification to a notification destination. The control unit 51 controls, for example, the operation of the output unit 55. The control unit 51 records, for example, various information generated by the execution of the notification determination process in the memory unit 54. The control unit 51 records, for example, information input to the input unit 52 or the communication unit 53 in the memory unit 54. The information input to the input unit 52 or the communication unit 53 is, for example, the estimation result of the mental state estimation unit 420.

[0080] The input unit 52 includes input devices such as a mouse, a keyboard, and a touch panel. The input unit 52 may be configured as an interface that connects these input devices to the control device 5. The input unit 52 accepts input of various information to the control device 5. To the input unit 52, for example, an estimation result from the mental state estimation unit 420 is input.

[0081] The communication unit 53 includes a communication interface for connecting the control device 5 to an external device. The communication unit 53 communicates with the external device via wired or wireless communication. The external device is, for example, the monitoring device 4. The external device is, for example, a predetermined notification destination.

[0082] The storage unit 54 is configured using a computer-readable storage medium device such as a magnetic hard disk device or a semiconductor storage device. The storage unit 54 stores various information related to the control device 5. The storage unit 54 stores, for example, information input via the input unit 52 or the communication unit 53. The storage unit 54 stores, for example, various information generated by the execution of a notification determination process.

[0083] It should be noted that the estimation result of the mental state estimation section 420 (i.e., the estimation result of the monitoring device 4) does not necessarily have to be input only to the input section 52, and does not necessarily have to be input only to the communication section 53. The estimation result of the mental state estimation section 420 may be input from either the input section 52 or the communication section 53.

[0084] The output unit 55 outputs various information. The output unit 55 includes a display device such as a CRT display, a liquid crystal display, or an organic EL display. The output unit 55 may be configured as an interface that connects these display devices to the control device 5. The output unit 55 outputs information input to the input unit 52, for example. The output unit 55 may display an estimation result input to the input unit 52 or the communication unit 53, for example. The output unit 55 may display a result of execution of a notification determination process, for example.

[0085] 8 is a diagram showing an example of a functional configuration of the control unit 51 in the embodiment. The control unit 51 includes an estimation result acquisition unit 510, a notification determination unit 520, a storage control unit 530, a communication control unit 540, and an output control unit 550.

[0086] The inference result acquisition unit 510 repeatedly acquires the inference result of the mental state estimation unit 420 input to the input unit 52 or the communication unit 53 at a predetermined cycle.

[0087] The notification determination unit 520 executes a notification determination process on the estimation result acquired by the estimation result acquisition unit 510. That is, the notification determination unit 520 determines whether the estimation result acquired by the estimation result acquisition unit 510 satisfies a notification criterion.

[0088] The memory control unit 530 records various information in the memory unit 54. The communication control unit 540 controls the operation of the communication unit 53.

[0089] The communication control unit 540 controls the operation of the communication unit 53, and causes the communication unit 53 to execute notification to, for example, a notification destination. The communication control unit 540 may cause the communication unit 53 to transmit a control signal for controlling the operation of the automobile 90, such as a signal instructing the automobile 90 to decelerate or a signal instructing the automobile 90 to stop.

[0090] The output control unit 550 controls the operation of the output unit 55. For example, the output control unit 550 controls the operation of the output unit 55 to cause the output unit 55 to output the determination result of the notification determination unit 520.

[0091] Fig. 9 is a flowchart showing an example of the flow of processing executed by the abnormal condition estimating system 100 of the embodiment. The abnormal condition estimating system 100 repeatedly executes the processing shown in the flowchart in Fig. 9 until a predetermined end condition is satisfied. The predetermined end condition is, for example, a condition that the supply of power to the biosignal acquiring device 1 is interrupted. The determination as to whether the end condition is satisfied is performed, for example, by the control unit 41. The determination as to whether the end condition is satisfied is performed, for example, by the cardiac state estimation unit 420. The determination as to whether the end condition is satisfied may be performed, for example, by the notification determination unit 520.

[0092] The cardiac state time series acquisition unit 410 acquires a cardiac state time series acquired from the estimation target 9 (step S101). Next, the cardiac state estimation unit 420 estimates the cardiac state of the estimation target 9 based on the cardiac state time series acquired in step S101 (step S102). Next, the notification determination unit 520 determines whether or not to notify the notification destination based on the estimation result in step S102 (step S103).

[0093] If it is determined that the notification should be sent to the notification destination (step S103: YES), the communication control unit 540 controls the operation of the communication unit 53 to notify the notification destination (step S104). After step S104, it is determined whether or not a termination condition is satisfied (step S105). If the termination condition is satisfied (step S105: YES), the process ends. If the termination condition is not satisfied (step S105: NO), the process returns to step S101.

[0094] If it is determined that the notification should not be sent to the notification destination (step S103: NO), the process proceeds to step S105.

[0095] The abnormal condition estimating system 100 of the embodiment configured in this manner estimates the cardiac state of the estimation target 9 only through processes with low amounts of calculations, such as processes for calculating statistics such as averages and deviations, processes for determining whether or not a threshold is exceeded, and processes for counting periods, frequency, etc. Therefore, the abnormal condition estimating system 100 can reduce the amount of calculations required for estimating the cardiac state.

[0096] Furthermore, the abnormal state estimation system 100 does not determine whether all samples in the cardiac state time series are out-of-range data, but determines whether samples in the refractory period are out-of-range data. Therefore, the abnormal state estimation system 100 can reduce the amount of calculation required to estimate cardiac abnormality compared to when all samples in the cardiac state time series are determined to be out-of-range data. Furthermore, the abnormal state estimation system 100 can perform highly accurate estimation that suppresses erroneous determination by avoiding determination in polarization sections where normal waveforms occur.

[0097] Furthermore, by including a notification determination unit 520 and a communication control unit 540, the abnormal condition estimation system 100 has a function of notifying a notification destination. Since it has the notification function, the abnormal condition estimation system 100 can appeal to the driver of the vehicle 90, such as a bus driver, to take action such as calling out to the driver or issuing an alarm, as necessary (i.e., in response to the determination of the notification determination unit 520). Therefore, the abnormal condition estimation system 100 can reduce the danger caused by the cardiac condition of the estimation target 9 being in an abnormal state.

[0098] Furthermore, since the abnormal condition estimation system 100 includes the notification determination unit 520 and the communication control unit 540, it is also possible to transmit a control signal to the automobile 90 directly, rather than to the driver, to cause the automobile 90 to decelerate or stop. Therefore, the abnormal condition estimation system 100 can reduce the risk caused by the cardiac condition of the estimation target 9 being in an abnormal state.

[0099] (First Modification) The cardiac state estimation unit 420 may execute a second type cardiac state estimation process as the cardiac state estimation process instead of the first type cardiac state estimation process. That is, the cardiac state estimation unit 420 may estimate the cardiac state of the estimation target 9 by executing the second type cardiac state estimation process on the cardiac state time series indicated by the cardiac state signal acquired by the cardiac state time series acquisition unit 410.

[0100] <Second type mental state estimation process> The second type cardiac state estimation process is a process for estimating the cardiac state of the estimation subject 9 based on the time interval RRI (RR-Interval) of an R wave in a cardiac state time series.

[0101] The second type cardiac state estimation process is a process for estimating that the cardiac state of the estimation target 9 is abnormal, for example, when the RRI in the cardiac state time series is smaller than a predetermined threshold, that is, an RRI lower limit threshold. The RRI lower limit threshold is, for example, the RRI of a person with a normal cardiac state during exercise. A value larger than the RRI of a person with a normal cardiac state during exercise is, for example, 600 ms.

[0102] When the RRI lower limit threshold is the RRI during exercise of a person with a normal heart condition, if the RRI of the cardiac state time series is smaller than the RRI lower limit threshold, there is a high possibility of ventricular tachycardia occurring. Therefore, by estimating the cardiac state based on whether the RRI in the cardiac state time series is smaller than the RRI lower limit threshold, it is possible to estimate whether the heart of the estimation target 9 is in an abnormal state in which ventricular tachycardia may occur.

[0103] In the second type cardiac state estimation process, when the RRI in the cardiac state time series is greater than an RRI upper threshold, which is a predetermined threshold different from the RRI lower threshold, it may be estimated that the cardiac state of the estimation target 9 is abnormal. When the cardiac state is abnormal, cardiac activity may decrease and the pulse rate may drop. In other words, when the cardiac state is abnormal, bradycardia may occur.

[0104] By estimating the state of the heart based on whether the RRI in the cardiac state time series is greater than the RRI upper threshold, it is possible to estimate whether the heart of the estimation target 9 is in an abnormal state in which bradycardia occurs. The RRI upper threshold is preferably a value that allows estimation of the occurrence of bradycardia, and is desirably, for example, 1000 ms or more.

[0105] In the second type cardiac state estimation process, the cardiac state of the estimation target 9 may be estimated using an RRI lower limit threshold and an RRI upper limit threshold.

[0106] The second type cardiac state estimation process may be a process for estimating the cardiac state of the estimation target 9 by further using environmental information. The environmental information used by the second type cardiac state estimation process to estimate the cardiac state of the estimation target 9 is, for example, information acquired by an inertial sensor such as an acceleration sensor or a gyro sensor, and is information indicating the acceleration of the estimation target 9 (hereinafter referred to as "detected target acceleration information"). In other words, when the second type cardiac state estimation process uses environmental information for the cardiac state of the estimation target 9, the environmental sensor 3 that provides the environmental information is, for example, an inertial sensor.

[0107] The RRI becomes small when the subject 9 moves, even if the cardiac state of the subject 9 is normal and not in a state where ventricular tachycardia occurs. Therefore, the second type cardiac state estimation process using not only the cardiac state time series but also the detected subject acceleration information can estimate the cardiac state of the subject 9 with higher accuracy than the second type cardiac state estimation process based only on the cardiac state time series.

[0108] In the second type cardiac state estimation process using not only the cardiac state time series but also the acceleration information of the detection target, a normality determination is performed. The normality determination is a process in which the cardiac state of the estimation target 9 is determined to be normal if the statistics obtained from the acceleration information of the detection target exceeds a threshold value that satisfies a predetermined condition, even if the RRI obtained from the cardiac state time series becomes smaller than a predetermined criterion.

[0109] Hereinafter, the predetermined condition satisfied by the threshold in the normal judgment is referred to as the normal threshold condition. The normal threshold condition is, for example, a condition of a predetermined value. In such a case, the threshold satisfying the normal threshold condition is a predetermined value. In the normal judgment, the heart state of the estimation target 9 is judged to be abnormal only when the RRI obtained from the cardiac state time series becomes smaller than a predetermined reference and the statistical quantity calculated based on the detection target acceleration information does not exceed the threshold satisfying the normal threshold condition.

[0110] This is because, even if the RRI becomes smaller, if the statistical amount obtained from the acceleration information of the detected object is large, the decrease in the RRI is likely not caused by an abnormality in the cardiac condition, but by the movement or motion of the estimated object 9.

[0111] The statistics obtained from the detection target acceleration information are specifically statistics of the distribution of sample values ​​indicated by a time series of information indicating the acceleration of the estimation target 9, which is information acquired by an inertial sensor.

[0112] The statistics in the normal judgment are, for example, values ​​obtained by integrating the absolute values ​​of the acceleration values ​​on three axes over a predetermined fixed time period. However, the statistics in the normal judgment may be any value calculated based on the detection target acceleration information, and are not limited to values ​​obtained by integrating the absolute values ​​of the acceleration values ​​on three axes over a predetermined fixed time period.

[0113] The normal threshold condition does not necessarily have to be a condition of a predetermined value that is determined in advance. The normal threshold condition may be, for example, a statistic calculated based on the detection target acceleration information in a past time period that satisfies a predetermined condition related to a period (hereinafter referred to as a "normal determination period condition"). The normal determination period condition is, for example, a condition of 3 seconds ago.

[0114] The normal threshold condition may be, for example, a condition that the value of the objective variable of a predetermined function having the detection object acceleration information in a past time section as an explanatory variable satisfies a normal determination period condition.

[0115] The environmental information used by the second type cardiac state estimation process to estimate the cardiac state of the estimation target 9 may include, for example, position information of the estimation target 9. The position information of the estimation target 9 is information acquired using a technology for acquiring position information, such as a global positioning system (GPS). That is, the environmental sensor 3 for acquiring position information is a device for acquiring the position information of the estimation target 9 using a technology for acquiring position information, such as a GPS of a smartphone equipped with a GPS function.

[0116] The presumed target 9 is often not engaged in strenuous exercise while driving the car 90. Therefore, when information indicating that the presumed target 9 is on the roadway or information indicating that the presumed target 9 is moving at a speed equal to or faster than a walking speed such as 50 km / h is acquired based on the location information, the decrease in RRI is not a decrease in RRI caused by the presumed target 9 exercising.

[0117] Therefore, a decrease in RRI when information indicating that the estimation target 9 is on the roadway or information indicating that the estimation target 9 is moving at a speed equal to or faster than a walking speed, such as 50 km per hour, is acquired based on the position information means that there is a high probability that the cardiac state of the estimation target 9 is abnormal. Therefore, when the second type cardiac state estimation process estimates the cardiac state of the estimation target 9 based also on the position information, it is possible to estimate the cardiac state of the estimation target 9 with higher accuracy than when estimating the cardiac state of the estimation target 9 without using the position information.

[0118] The abnormal condition estimating system 100 of the first modified example configured in this manner estimates the cardiac state of the estimation target 9 only through processes with low amounts of calculations, such as processes for calculating statistics such as averages and deviations, processes for determining whether or not a threshold is exceeded, and processes for counting periods, frequency, etc. Therefore, the abnormal condition estimating system 100 can reduce the amount of calculations required for estimating cardiac abnormalities.

[0119] (Second Modification) The cardiac state estimation unit 420 may execute a third type cardiac state estimation process as the cardiac state estimation process instead of the first type cardiac state estimation process. That is, the cardiac state estimation unit 420 may estimate the cardiac state of the estimation target 9 by executing the third type cardiac state estimation process on the cardiac state time series indicated by the cardiac state signal acquired by the cardiac state time series acquisition unit 410.

[0120] <Third type mental state estimation processing> The third type cardiac state estimation process is a process for executing a first type cardiac state estimation process, a second type cardiac state estimation process, and a first integrated estimation process. The first integrated estimation process is a process for estimating the cardiac state of the estimation target 9 based on the estimation result of the first type cardiac state estimation process and the estimation result of the second type cardiac state estimation process.

[0121] In ventricular fibrillation, which is one of the phenomena caused by cardiac abnormalities, the QRA waveform is irregular (see Non-Patent Document 2).

[0122] In the third type cardiac state estimation process, the first type cardiac state estimation process and the second type cardiac state estimation process are executed, and then the first integrated estimation process is executed. In the first integrated estimation process, the cardiac state of the estimation target 9 is estimated to be abnormal only when both the first type cardiac state estimation process and the second type cardiac state estimation process estimate that the cardiac state of the estimation target 9 is abnormal.

[0123] Therefore, in the first integrated estimation process, even if the cardiac condition is estimated to be abnormal by executing the first type cardiac state estimation process, if the estimation result by executing the second type cardiac state estimation process estimates the cardiac condition to be normal, the cardiac condition of the estimation target 9 is estimated to be normal.

[0124] In this way, the third type cardiac state estimation process is a process that estimates the cardiac state of the estimation subject 9 using the estimation results of both the first type cardiac state estimation process and the second type cardiac state estimation process, rather than just the estimation results of either the first type cardiac state estimation process or the second type cardiac state estimation process.

[0125] Therefore, the abnormal condition estimation system 100 of the second modified example configured as above can estimate the cardiac state of the estimation target 9 with higher accuracy than when the cardiac state of the estimation target 9 is estimated using the estimation result of either the first type cardiac state estimation process or the second type cardiac state estimation process. That is, the abnormal condition estimation system 100 of the second modified example estimates the cardiac state of the estimation target 9 based on two conditions, namely, the occurrence of a signal in the refractory period and the irregularity of the QRS waveform, and therefore can estimate the cardiac state of the estimation target 9 with high accuracy.

[0126] The occurrence of a signal during the refractory period means that the peak period appearance condition is satisfied.

[0127] In the second type cardiac state estimation process executed in the third type cardiac state estimation process, the cardiac state of the estimation target 9 does not necessarily need to be estimated based on whether the RRI exceeds a threshold value, and the cardiac state of the estimation target 9 may be estimated using the RRI statistics. That is, the cardiac state of the estimation target 9 may be estimated by estimating the irregularity of the QRS waveform in ventricular fibrillation as the RRI irregularity using the RRI statistics.

[0128] The RRI statistic may be, for example, the average, deviation, variance, median, absolute deviation, root mean square, percentile, maximum value, or minimum value of the RRI.

[0129] For example, when the RRI statistic is the average RRI, in the third type cardiac state estimation process, when the difference between the average values ​​of the repeatedly calculated RRO exceeds a predetermined threshold and the occurrence of a signal during the refractory period is confirmed, the cardiac state of the estimation target 9 is estimated to be abnormal.

[0130] The RRI statistic may be another statistic other than the average such as the deviation. Even if the RRI statistic is another statistic, the cardiac condition of the estimation target 9 is estimated to be abnormal depending on whether or not the difference between the repeatedly calculated statistics exceeds a predetermined threshold.

[0131] (Third Modification) The cardiac state estimation unit 420 may execute a fourth type cardiac state estimation process as the cardiac state estimation process instead of the first type cardiac state estimation process. That is, the cardiac state estimation unit 420 may estimate the cardiac state of the estimation target 9 by executing a fourth type cardiac state estimation process on the cardiac state time series indicated by the cardiac state signal acquired by the cardiac state time series acquisition unit 410.

[0132] <Fourth type of mental state estimation processing> The fourth type cardiac state estimation process is a process for executing a first type cardiac state estimation process, a second type cardiac state estimation process, and a second integrated estimation process. The second integrated estimation process is a process for estimating the cardiac state of the estimation target 9 based on the estimation results of the first type cardiac state estimation process and the second type cardiac state estimation process.

[0133] The second integrated estimation process is a process for estimating that the cardiac state of the estimation target 9 is abnormal, regardless of the estimation result of the first type cardiac state estimation process, when the estimation result of the second type cardiac state estimation process is abnormal. In this respect, the first integrated estimation process differs from the second integrated estimation process.

[0134] The second integrated estimation process is a process that estimates that the cardiac state of the estimation target 9 is abnormal when the estimation result of the second type cardiac state estimation process is normal even if the estimation result of the first type cardiac state estimation process is abnormal. The second integrated estimation process is a process that estimates that the cardiac state of the estimation target 9 is normal when the estimation result of the first type cardiac state estimation process is normal and the estimation result of the second type cardiac state estimation process is also normal.

[0135] In the fourth mental state inference process, the first mental state inference process and the second mental state inference process are executed, and then the second integrated inference process is executed.

[0136] In this way, the fourth type cardiac state estimation process is a process that estimates the cardiac state of the estimation subject 9 using the estimation results of both the first type cardiac state estimation process and the second type cardiac state estimation process, rather than just the estimation results of either the first type cardiac state estimation process or the second type cardiac state estimation process.

[0137] Therefore, the abnormal state estimation system 100 of the third modified example configured in this manner can estimate the cardiac state of the estimation target 9 with higher accuracy than when the cardiac state of the estimation target 9 is estimated using the estimation results of either the first type cardiac state estimation process or the second type cardiac state estimation process.

[0138] FIG. 10 is an explanatory diagram for explaining the effect of the fourth type cardiac state estimation process in the modified example. FIG. 10 shows a process in which ventricular tachycardia occurs in the estimation target 9 that has been generating a normal cardiac potential, followed by ventricular fibrillation. FIG. 10 shows that the heart is normal in the period from time position t0 to time position t1. FIG. 10 shows that tachycardia occurs in the period from time position t1 to time position t2. FIG. 10 shows that ventricular flutter or ventricular fibrillation occurs in the period from time position t2 to time position t3. The waveform in the period from time t2 to time t3 in FIG. 10 is an example of a waveform showing that the pulse becomes weak due to, for example, cardiopulmonary ischemia. FIG. 10 shows that the state transitions to cardiac arrest after time position t3.

[0139] The waveform surrounded by a frame A1 in Fig. 10 is an example of a waveform estimated to be abnormal by a second type of cardiac state estimation process. The waveform surrounded by a frame A2 in Fig. 10 is an example of a waveform estimated to be abnormal by a first type of cardiac state estimation process. The waveform surrounded by an area A3 in Fig. 10 is an example of a waveform leading to cardiac arrest.

[0140] (Fourth Modification) The cardiac state estimation unit 420 may execute a fifth type cardiac state estimation process as the cardiac state estimation process instead of the first type cardiac state estimation process. That is, the cardiac state estimation unit 420 may estimate the cardiac state of the estimation target 9 by executing the fifth type cardiac state estimation process on the cardiac state time series indicated by the cardiac state signal acquired by the cardiac state time series acquisition unit 410.

[0141] <Fifth type of mental state estimation processing> The fifth type cardiac state estimation process differs from the first type cardiac state estimation process to the fourth type cardiac state estimation process in that it estimates the state of cardiac arrest. The state of cardiac arrest is, for example, the state after time t3 in FIG. 10. The fifth type cardiac state estimation process is a process that executes the first type cardiac state estimation process, the second type cardiac state estimation process, the first integrated estimation process, the second integrated estimation process, and the cardiac arrest estimation process. In the fifth type cardiac state estimation process, the first integrated estimation process and the second integrated estimation process are executed after the first type cardiac state estimation process and the second type cardiac state estimation process, and then the cardiac arrest estimation process is executed.

[0142] The cardiac arrest estimation process is a process for estimating that the cardiac state of the estimation target 9 is in a cardiac arrest state when the deviation of the distribution of cardiac state quantities within a predetermined period after the abnormality occurrence time position is equal to or smaller than a predetermined threshold. The abnormality occurrence time position is the time position at which the cardiac state of the estimation target 9 is estimated to be abnormal by the first integrated estimation process or the second integrated estimation process.

[0143] During cardiac arrest, the cardiac potential changes due to pulsation are approximately zero, and are approximately equal to 0 mV. Therefore, the cardiac arrest estimation process includes a process of calculating the deviation of the distribution of cardiac state quantities every 200 ms, and a process of determining whether the calculated deviation is outside the range of ±15 mV.

[0144] Therefore, the abnormal condition estimating system 100 of the fourth modified example configured as above can detect cardiac arrest by executing the fifth type cardiac state estimating process.

[0145] (Fifth Modification) The cardiac state estimation unit 420 may shape the waveform of the cardiac state time series using a filter that performs various signal processing such as an analog filter, a digital filter such as FIR (Finite Impulse Response) or IR (Infinite Impulse Response), a moving average filter that applies a moving average, etc. For example, noise components included in the cardiac state time series are removed by using a filter.

[0146] The statistics calculated in the statistics calculation process are not limited to the average and deviation, but may be variance, average, median, absolute deviation, root mean square, percentile, maximum, minimum, or other statistics.

[0147] Regarding the setting of the peak determination target period, a new peak determination target period is set from 0 milliseconds every time the data is updated in accordance with the sampling rate, and the peak determination target periods may be set so as to overlap each other.

[0148] The peak period appearance condition does not necessarily have to include the condition that the peak periods are consecutive. Therefore, the peak period appearance condition may be, for example, a condition that four or more peak periods occur within 1000 milliseconds, regardless of whether they are consecutive or non-consecutive.

[0149] The monitoring device 4 may be implemented using a plurality of information processing devices communicably connected via a network. In this case, each functional unit of the monitoring device 4 may be distributed and implemented in the plurality of information processing devices.

[0150] The control device 5 may be implemented using a plurality of information processing devices communicably connected via a network. In this case, each functional unit included in the control device 5 may be distributed and implemented in the plurality of information processing devices.

[0151] The monitoring device 4 and the control device 5 do not necessarily have to be implemented as different devices. For example, the monitoring device 4 and the control device 5 may be implemented as a single device having both functions. For example, the control unit 41 may include a notification determination unit 520.

[0152] All or part of the functions of the abnormal state estimation system 100 may be realized using hardware such as an ASIC (Application Specific Integrated Circuit), a PLD (Programmable Logic Device), or an FPGA (Field Programmable Gate Array). The program may be recorded on a computer-readable recording medium. Examples of computer-readable recording media include portable media such as flexible disks, magneto-optical disks, ROMs, and CD-ROMs, and storage devices such as hard disks built into a computer system. The program may be transmitted via an electric communication line.

[0153] The monitoring device 4 is an example of a state estimation device.

[0154] Although an embodiment of the present invention has been described in detail above with reference to the drawings, the specific configuration is not limited to this embodiment, and designs that do not deviate from the gist of the present invention are also included. [Explanation of symbols]

[0155] 100...abnormal state estimation system, 1...biological signal acquisition device, 2...relay terminal, 3...environmental sensor, 4...monitoring device, 5...control device, 41...control unit, 42...input unit, 43...communication unit, 44...storage unit, 45...output unit, 410...heart state time series acquisition unit, 420...heart state estimation unit, 430...storage control unit, 440...communication control unit, 450...output control unit, 460...environmental information acquisition unit, 51...control unit, 52...input unit, 53...communication unit, 54...storage unit, 55...output unit, 510...estimation result acquisition unit, 520...notification determination unit, 530...storage control unit, 540...communication control unit, 550...output control unit, 91...processor, 92...memory, 93...processor, 94...memory, 9...estimation target, 90...automobile

Claims

1. a cardiac state time series acquisition unit that acquires a cardiac state time series that is a time series of cardiac state quantities that are quantities indicating a state of the heart of an object to be estimated; a cardiac state estimation unit which estimates the cardiac state of the estimation target based on the occurrence time of the out-of-range data by treating refractory period samples, which are refractory period samples among the cardiac state time series samples acquired by the cardiac state time series acquisition unit, as out-of-range data, the refractory period samples having values ​​outside a processing threshold region determined according to a distribution of the refractory period samples; A state estimation device comprising:

2. The cardiac state estimation unit further estimates the cardiac state of the estimation target based on a time interval RRI (RR-Interval) of an R wave in the cardiac state time series. The state estimation device according to claim 1 .

3. the cardiac state estimation unit estimates that the cardiac state of the estimation subject is abnormal when an estimation result of the cardiac state of the estimation subject based on a time interval RRI of an R wave in the cardiac state time series is an estimation result that the cardiac state of the estimation subject is abnormal and when an estimation result of the cardiac state of the estimation subject based on an occurrence time of the out-of-range data is also an estimation result that the cardiac state of the estimation subject is abnormal. The state estimation device according to claim 2 .

4. when an estimation result of the cardiac state of the estimation subject based on a time interval RRI of an R wave in the cardiac state time series indicates that the cardiac state of the estimation subject is abnormal, the cardiac state estimation unit estimates that the cardiac state of the estimation subject is abnormal regardless of a result of the estimation of the cardiac state of the estimation subject based on a time of occurrence of the out-of-range data. The state estimation device according to claim 2 .

5. The cardiac state estimation unit estimates that the cardiac state of the subject to be estimated is a state of cardiac arrest when a deviation in distribution of cardiac state quantities within a predetermined period after an abnormality occurrence time position, which is a time position at which the cardiac state of the subject to be estimated is abnormal based on either or both of an estimation result of the cardiac state of the subject to be estimated based on a time interval RRI of R waves in the cardiac state time series and an estimation result of the cardiac state of the subject to be estimated based on the occurrence time of the out-of-range data, is equal to or less than a predetermined threshold value, with the position of each sample in the cardiac state time series in the time axis direction being the time position. The state estimation device according to claim 2 .

6. A computer-implemented state estimation method, comprising: a cardiac state time series acquisition step of acquiring a cardiac state time series which is a time series of cardiac state quantities which are quantities indicating a state of the heart of an object to be estimated; a cardiac state estimation step of estimating a cardiac state of the estimation target based on a generation time of the out-of-range data, the out-of-range data being a refractory period sample out of the cardiac state time-series samples acquired in the cardiac state time-series acquisition step, the out-of-range data being a refractory period sample having a value outside a processing threshold region determined according to a distribution of the refractory period sample; A state estimation method having the following structure:

7. A program for causing a computer to function as the state estimation device according to any one of claims 1 to 5.

Images