Sleep state analysis system

Abstract

This system provides an analysis method that can understand the disease attributes of subjects based on their past and present clinical information and various data collected during examinations, thereby improving the accuracy of sleep stage analysis results. [Solution] A sleep state analysis system comprising a disease attribute estimation unit 40 that estimates the disease attributes of a subject from all or part of the measured values ​​of a heart rate sensor 21, measured values ​​of other sensors, and the subject's clinical information, and an algorithm modification unit 50 that modifies the algorithm for analyzing sleep stages from the disease attributes of the subject estimated by the disease attribute estimation unit 40, which calculates heart rate variability from at least the measured values ​​of the subject's heart rate sensor 21 and analyzes the subject's sleep stages from the heart rate variability.

Description

Technical Field

[0001] The present invention relates to a sleep state analysis system that analyzes a subject's sleep stage from the subject's heartbeat information and the like.

Background Art

[0002] As this type of sleep state analysis system, as disclosed in Patent Document 1 below, there is known a sleep determination device that determines which of a plurality of sleep stages a user corresponds to by using heart rate variability parameters based on heart rate data.

Prior Art Literature

Patent Literature

[0003]

Patent Document 1

Summary of the Invention

Problem to be Solved by the Invention

[0004] However, it has been known that when analyzing a subject's sleep stage, the analysis result of the sleep stage varies depending on the subject's health condition and the like, and establishment of an analysis system that can correctly reflect the subject's past and present clinical information and various information at the time of examination in the analysis result of the sleep stage is desired.

[0005] In view of the above circumstances, an object of the present invention is to provide an analysis system capable of grasping a subject's disease attribute based on the subject's past and present clinical information and various information at the time of examination, and improving the accuracy of a sleep stage analysis result.

Means for Solving the Problem

[0006] A sleep state analysis system that calculates heart rate variability from at least a measurement value obtained by a heart rate sensor of a subject, and analyzes the sleep stage of the subject from the heart rate variability, wherein A disease attribute estimation unit estimates the disease attributes of the subject from all or part of the measured values ​​of the heart rate sensor, the measured values ​​of other sensors, and the subject's clinical information. An algorithm modification unit that modifies the algorithm for analyzing sleep stages based on the disease attributes of the subject estimated by the disease attribute estimation unit. It is characterized by being equipped with [the following features].

[0007] According to the sleep state analysis system of the first invention, the disease attributes of a subject can be estimated from all or part of the measurements of the heart rate sensor, the measurements of other sensors, and the subject's clinical information, and the optimal analysis algorithm can be selected from various algorithms according to those disease attributes, thereby enabling more accurate sleep state analysis.

[0008] In other words, the sleep state analysis system of the first invention allows for the identification of the subject's disease attributes based on the subject's past and present clinical information and various information obtained during examinations, thereby improving the accuracy of the sleep stage analysis results.

[0009] The sleep state analysis system of the second invention is, in the first invention, The disease attribute estimation unit estimates sleep apnea syndrome from either or both of the airflow measurement value from the airflow sensor and the blood oxygen saturation from the SpO2 sensor, and if sleep apnea syndrome is estimated, the algorithm modification unit changes the algorithm for analyzing sleep stages to the sleep apnea syndrome-compatible algorithm. It is characterized by the following:

[0010] According to the sleep state analysis system of the second invention, sleep apnea syndrome can be estimated from either or both of the airflow measurement values ​​from the airflow sensor and the blood oxygen saturation from the SpO2 sensor for the subject. Based on this information, the algorithm for analyzing sleep stages can be changed to a sleep apnea syndrome-compatible algorithm, thereby enabling optimal sleep state analysis for the subject.

[0011] Thus, according to the sleep state analysis system of the second invention, it is possible to perform an optimal sleep state analysis for the subject based on the subject's examination information.

[0012] The sleep state analysis system of the third invention is, in the first invention, The disease attribute estimation unit determines whether the subject's disease attribute corresponds to dementia by referring to the subject's clinical information, and if so, the algorithm modification unit changes the algorithm for analyzing sleep stages to a dementia-specific algorithm. It is characterized by the following:

[0013] According to the sleep state analysis system of the third invention, it is possible to determine whether the subject's disease attribute corresponds to dementia by referring to the subject's clinical information, and based on that information, the algorithm for analyzing sleep stages can be changed to a dementia-specific algorithm, thereby enabling optimal sleep state analysis for the subject.

[0014] Thus, according to the sleep state analysis system of the third invention, it is possible to perform an optimal sleep state analysis for the subject based on the subject's clinical information.

[0015] The sleep state analysis system of the fourth invention is, in the first invention, The disease attribute estimation unit determines whether the subject's disease attribute corresponds to Parkinson's disease by referring to the subject's clinical information, and if so, the algorithm modification unit changes the algorithm for analyzing sleep stages to the Parkinson's disease-compatible algorithm. It is characterized by the following:

[0016] According to the sleep state analysis system of the fourth invention, it is possible to determine whether the subject's disease attribute corresponds to Parkinson's disease by referring to the subject's clinical information, and based on that information, the algorithm for analyzing sleep stages can be changed to a Parkinson's disease-compatible algorithm, thereby enabling optimal sleep state analysis for the subject.

[0017] As described above, according to the sleep state analysis system of the fourth aspect of the present invention, optimal sleep state analysis can be performed for a subject based on the subject's clinical information. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] [Figure 1] A block diagram showing the overall configuration of a sleep state analysis system according to an embodiment of the present invention. [Figure 2] An explanatory diagram showing one application example of the sleep state analysis system of Fig. 1. [Figure 3] An explanatory diagram showing another application example of the sleep state analysis system of Fig. 1. [Figure 4] An explanatory diagram showing still another application example of the sleep state analysis system of Fig. 1. [Figure 5] An explanatory diagram showing one example of sleep state analysis in the sleep state analysis system according to an embodiment of the present invention. MODE FOR CARRYING OUT THE INVENTION

[0019] The overall configuration of a sleep state analysis system according to an embodiment of the present invention will be described below with reference to Fig. 1.

[0020] As shown in Fig. 1, the sleep state analysis system 10 includes: a sensor unit 20 including a heart rate sensor 21 and the like; a clinical information storage unit 30; a disease attribute estimation unit 40; an algorithm changing unit 50; and a sleep stage analysis unit 60.

[0021] For convenience, the sensor unit 20 is described as a single sensor group; however, in practice, each sensor is attached to a different part of the subject and outputs detection signal data individually. Therefore, the sleep state analysis system 10 may be provided with a data storage unit or the like as needed.

[0022] The heart rate sensor 21 detects an electrocardiographic signal or a pulse wave signal of the subject by bringing its electrode terminals into contact with a plurality of locations on the body surface of the subject, and various known systems can be employed therefor.

[0023] The airflow sensor 22 is installed near the subject's head to detect the subject's respiratory airflow, and various known systems can be used.

[0024] The SpO2 sensor 23 is attached to the subject's fingertip or other body part and accurately detects the oxygen saturation (saturation) in the subject's blood using a sensor that detects using red light and infrared light, and various known systems can be employed.

[0025] The acceleration sensor 24 is attached to the limbs or other parts of the subject to detect the subject's physical activity, and various known systems can be used.

[0026] Past and present clinical information of the subject is stored in the clinical information storage unit 30 using an input unit (not shown) or the like.

[0027] The disease attribute estimation unit 40 estimates the subject's disease attributes from all or part of the measured values ​​of the heart rate sensor 21, the measured values ​​of other sensors such as the airflow sensor 22, and the subject's clinical information.

[0028] Here, among the disease attributes of the subjects, sleep apnea syndrome (hereinafter also referred to as SAS) is estimated from either the airflow measurement value of the airflow sensor 22 or the blood oxygen saturation of the SpO2 sensor 23, or both. For dementia and Parkinson's disease, the determination is made by referring to the clinical information stored in the clinical information storage unit 30. The criteria for determining SAS are made in accordance with the US SAS guidelines, etc. In addition, the subject's age, hypertension, diabetes, the severity of the estimated disease, and the severity of the relevant disease may also be considered as disease attributes.

[0029] The algorithm modification unit 50 is equipped with various algorithms optimized for the disease attributes of each subject, and the most suitable algorithm is selected based on the judgment of the disease attribute estimation unit 40.

[0030] Here, algorithm 1 is applied to healthy individuals, while algorithms 2-4 are applied to patients with sleep apnea syndrome (SAS), dementia, and Parkinson's disease, respectively. Various methods can be employed for the configuration of each algorithm stored in the algorithm modification unit 50.

[0031] The sleep stage analysis unit 60 employs a suitable algorithm selected by the algorithm change unit 50 to calculate the subject's heart rate variability from the heart rate sensor 21's measurements and analyzes the subject's sleep stage based on that heart rate variability. Furthermore, information regarding the subject's physical activity measured by the acceleration sensor 24 can also be incorporated into the sleep stage analysis as additional information.

[0032] The sleep stage analysis unit 60 can, for example, distinguish between four stages, including wakefulness, and the analysis is performed using a machine learning system that has been trained on the results of a large amount of analysis data.

[0033] The analysis results of the sleep stages are then output via an output / display unit (not shown in the diagram).

[0034] The above describes the configuration of the sleep state analysis system. In this configuration, the sleep state analysis system 10 may be an integrated device containing all the components, or it may be implemented as a system in which some or all of the components other than the sensor unit 20 are built on the cloud.

[0035] Next, we will explain in detail the application examples of the sleep state analysis system of this embodiment using Figures 2 to 4.

[0036] Figure 2 shows a configuration for estimating sleep apnea syndrome (SAS) from either or both of the airflow measurement values ​​from the airflow sensor 22 and the blood oxygen saturation values ​​from the SpO2 sensor 23.

[0037] Here, using either or both of the above airflow measurement values ​​and / or blood oxygen saturation values, the disease attribute estimation unit 40 estimates whether or not the subject has sleep apnea syndrome (SAS) according to the US SAS guidelines, and notifies the algorithm modification unit 50 of the estimation result. If SAS is estimated, the algorithm modification unit 50 adopts algorithm 2 as the applicable algorithm. If SAS is not estimated, algorithm 1 for healthy individuals is adopted.

[0038] In the former case, the sleep stage analysis unit 60 applies algorithm 2 based on the measurements of the heart rate sensor 21 to analyze the subject's sleep state, while in the latter case, algorithm 1 is applied.

[0039] Figure 3 shows the configuration in a case where dementia is suspected.

[0040] In Figure 3, the disease attribute estimation unit 40 determines whether the subject has dementia or not based on the subject's clinical information stored in the clinical information storage unit 30, and notifies the algorithm modification unit 50 of the determination result. If the determination is made, the algorithm modification unit 50 adopts algorithm 3 as the applicable algorithm. If the determination is not made, algorithm 1 for healthy individuals is adopted.

[0041] In the former case, the sleep stage analysis unit 60 applies algorithm 3 based on the measurements of the heart rate sensor 21 to analyze the subject's sleep state, while in the latter case, algorithm 1 is applied.

[0042] Figure 4 shows the configuration in a case where Parkinson's disease is suspected.

[0043] Similar to Figure 3, in Figure 4, the disease attribute estimation unit 40 determines whether the subject has Parkinson's disease based on the subject's clinical information stored in the clinical information storage unit 30, and notifies the algorithm modification unit 50 of the determination result. If the determination is made, the algorithm modification unit 50 adopts algorithm 4 as the applicable algorithm. If the determination is not made, algorithm 1 for healthy individuals is adopted.

[0044] In the former case, the sleep stage analysis unit 60 applies algorithm 4 based on the measurements of the heart rate sensor 21 to analyze the subject's sleep state, while in the latter case, algorithm 1 is applied.

[0045] Here, using Figure 5, an example of sleep state analysis in the sleep stage analysis unit 60 will be described.

[0046] Figure 5 shows the results of verifying the suitability of applying algorithms 1-4 to healthy individuals and individuals with each disease. Figure 5 also shows the relative accuracy of sleep stage accuracy based on sleep analysis.

[0047] As is clear from Figure 5, (A) Algorithm 1 is for healthy individuals, so the accuracy of sleep analysis for healthy individuals is good (1), but the accuracy decreases when used for patients with sleep apnea syndrome (SAS) (2). On the other hand, when algorithm 2, which is designed for SAS, is used for patients with SAS, the analysis can be performed with better accuracy than when used for healthy individuals (3).

[0048] Furthermore, (B) Algorithm 1 is designed for healthy individuals, so its sleep analysis accuracy is good for healthy individuals (1), but its accuracy decreases when used for dementia patients (4). On the other hand, using Algorithm 3, which is designed for dementia, on dementia patients allows for more accurate analysis than using it for healthy individuals (5).

[0049] Similarly, algorithm 1 (C) is designed for healthy individuals, so its sleep analysis accuracy is good for healthy individuals (1), but its accuracy decreases when used for Parkinson's disease patients (6). On the other hand, using algorithm 4, which is designed for Parkinson's disease, on Parkinson's disease patients allows for more accurate analysis than using it for healthy individuals (7).

[0050] The above describes the processing details of the sleep state analysis system of this embodiment. This sleep state analysis system allows for the understanding of the subject's disease attributes based on the subject's past and present clinical information and various information from examinations, thereby realizing an analysis system that can improve the accuracy of the sleep stage analysis results. [Explanation of symbols]

[0051] 1...Subject, 10...Sleep state analysis system, 21...Heart rate sensor, 22...Airflow sensor, 23...SpO2 sensor, 30...Clinical information storage unit, 40...Disease attribute estimation unit, 50...Algorithm modification unit, 60...Sleep stage analysis unit.

Claims

1. A sleep state analysis system that calculates heart rate variability from at least the heart rate sensor readings of a subject, and analyzes the subject's sleep stage from that heart rate variability, A disease attribute estimation unit estimates the disease attributes of the subject from all or part of the measured values ​​of the heart rate sensor, the measured values ​​of other sensors, and the subject's clinical information. An algorithm modification unit that modifies the algorithm for analyzing sleep stages based on the disease attributes of the subject estimated by the disease attribute estimation unit. A sleep state analysis system characterized by having the following features.

2. In the sleep state analysis system according to claim 1, The sleep state analysis system is characterized in that the disease attribute estimation unit estimates sleep apnea syndrome from either or both of the airflow measurement value of the airflow sensor and the blood oxygen saturation of the SpO2 sensor, and if sleep apnea syndrome is estimated, the algorithm modification unit changes the algorithm for analyzing sleep stages to a sleep apnea syndrome-compatible algorithm.

3. In the sleep state analysis system according to claim 1, The sleep state analysis system is characterized in that the disease attribute estimation unit determines whether the subject's disease attribute corresponds to dementia by referring to the subject's clinical information, and if so, the algorithm modification unit changes the algorithm for analyzing sleep stages to a dementia-specific algorithm.

4. In the sleep state analysis system according to claim 1, The sleep state analysis system is characterized in that the disease attribute estimation unit determines whether the subject's disease attribute corresponds to Parkinson's disease by referring to the subject's clinical information, and if so, the algorithm modification unit changes the algorithm for analyzing sleep stages to a Parkinson's disease-compatible algorithm.

Images