本发明涉及判断人是否睡着的领域。本发明涉及用于根据在人身上测量到的测量信号来判断人是否睡着的装置。The present invention relates to the field of judging whether a person is asleep and to a device for judging whether a person is asleep based on a measurement signal measured on the person.

发明内容Summary of the invention

本发明人经过以下考虑：The inventors have considered the following:

为了判断人是否睡着，可以使用基于加速度的睡眠评分算法。这类算法使用人的加速度来计算人是否睡着。这种仅基于加速度的睡眠判断的问题在于，通常不可能仅使用基于加速度的数据将人不活动(醒着且不运动或仅有很少运动)的状态与人睡着了的状态分开。也就是说，当使用这样的算法时，不活动的人(例如，因为从平板电脑观看电影或在移动装置上浏览互联网)经常被这样的算法错误地判断或错误地归类为睡着了。换句话说，基于不活动的人的加速度数据，算法可能会错误地判断人睡着了(实际上没有睡着，只是不活动)。In order to determine whether a person is asleep, an acceleration-based sleep scoring algorithm can be used. This type of algorithm uses a person's acceleration to calculate whether a person is asleep. The problem with this type of sleep judgment based only on acceleration is that it is usually impossible to separate the state of a person being inactive (awake and not moving or with only a little movement) from the state of a person being asleep using only acceleration-based data. That is, when using such an algorithm, inactive people (for example, because they are watching a movie on a tablet or browsing the Internet on a mobile device) are often misjudged or misclassified as asleep by such an algorithm. In other words, based on the acceleration data of an inactive person, the algorithm may mistakenly determine that the person is asleep (when in fact they are not asleep, just inactive).

因此，通过用于确定在人身上测量到的加速度信号来判断人是否睡着的装置，可能会错误地将人的非活动状态(即人是醒着的但运动减少)判断或归类为人的睡眠状态(即人在睡觉/睡着了)。另一个问题是，夜间的短暂唤醒有时无法确定(无法识别或辨别)。Therefore, a device for determining whether a person is asleep by determining acceleration signals measured on a person may erroneously determine or classify a person's inactive state (i.e., the person is awake but has reduced movement) as a person's sleeping state (i.e., the person is sleeping/asleep). Another problem is that brief awakenings during the night are sometimes not determined (cannot be identified or distinguished).

鉴于上述情况，本发明旨在改进对人是否睡着的判断。一个目标可能是改进这种判断，以区分人不活动但仍然醒着(未睡着)的状态和人睡着(不是醒着)的状态。另一个目标可以是提供一种用于判断人是否睡着的装置，该装置根据上述描述进行了改进。In view of the above, the present invention aims to improve the judgment of whether a person is asleep. One goal may be to improve this judgment to distinguish between a state where the person is inactive but still awake (not asleep) and a state where the person is asleep (not awake). Another goal may be to provide a device for judging whether a person is asleep, which is improved according to the above description.

该目标通过所附独立权利要求的主题来实现。有利的实现方式在从属权利要求中进一步定义。This object is achieved by the subject matter of the attached independent claims. Advantageous implementations are further defined in the dependent claims.

本发明的第一方面提供了一种用于判断人是否睡着的装置。所述装置用于接收心脏信号和/或温度信号。所述心脏信号是与人的心率和/或人的血容量变化相关的信号。所述温度信号与人的温度相关。此外，所述装置用于根据所述心脏信号的幅度和/或根据关于所述温度信号的幅度变化的变量来判断人是否睡着。A first aspect of the present invention provides a device for determining whether a person is asleep. The device is used to receive a heart signal and/or a temperature signal. The heart signal is a signal related to a person's heart rate and/or a change in a person's blood volume. The temperature signal is related to a person's temperature. In addition, the device is used to determine whether a person is asleep based on the amplitude of the heart signal and/or based on a variable related to the amplitude change of the temperature signal.

换句话说，第一方面提出使用与人的心率和/或血容量变化相关的心脏信号的幅度来判断人是否睡着。附加地或替代地，第一方面提出使用关于与人的温度相关的温度信号的幅度变化的变量来判断人是否睡着。判断人是否睡着了可以称为判断人是否处于睡眠状态(即人的睡眠状态是否存在)。当人未睡着时，人是醒着的(即人处于清醒状态)。因此，判断人是否睡着可以称为判断人是睡着了还是醒着的。In other words, the first aspect proposes to use the amplitude of the heart signal related to the change of the heart rate and/or blood volume of the person to determine whether the person is asleep. Additionally or alternatively, the first aspect proposes to use a variable about the amplitude change of the temperature signal related to the temperature of the person to determine whether the person is asleep. Determining whether the person is asleep can be referred to as determining whether the person is in a sleeping state (i.e., whether the sleeping state of the person exists). When the person is not asleep, the person is awake (i.e., the person is in a wakeful state). Therefore, determining whether the person is asleep can be referred to as determining whether the person is asleep or awake.

人的皮肤温度是由身体的体温调节机制控制的。在睡眠开始时(即人睡着了时)，身体热调节随着身体活动的减少而降低，而在睡眠结束时(人醒着时)，身体热调节会增加。通常，睡着时，身体产热会因心率降低而降低，并且人的远端温度会升高，热量损失也会增加。由于这两种变化，核心体温(core body temperature，CBT)降低。CBT也可以称为人的近端温度。因此，在人的睡眠开始时(即人睡着了时),CBT降低，而人的远端温度(例如，手指温度)升高。远端温度(例如，人的四肢温度)和近端温度(CBT)及其梯度(DPG，即远端温度和近端温度之间的梯度)可能是睡眠潜伏期的可靠预测因子。外周血管扩张程度越高，入睡时间所需的越短。A person's skin temperature is controlled by the body's thermoregulatory mechanism. At the beginning of sleep (i.e., when a person falls asleep), body thermal regulation decreases as physical activity decreases, while at the end of sleep (when a person is awake), body thermal regulation increases. Normally, when asleep, body heat production decreases due to a decrease in heart rate, and a person's distal temperature increases, and heat loss increases. Due to these two changes, the core body temperature (CBT) decreases. CBT can also be called a person's proximal temperature. Therefore, at the beginning of a person's sleep (i.e., when a person falls asleep), CBT decreases, while a person's distal temperature (e.g., finger temperature) increases. Distal temperature (e.g., the temperature of a person's limbs) and proximal temperature (CBT) and their gradient (DPG, i.e., the gradient between distal and proximal temperatures) may be reliable predictors of sleep latency. The higher the degree of peripheral vasodilation, the shorter the time required to fall asleep.

因此，使用心脏信号的幅度和/或与温度信号的幅度变化相关的变量可以判断人是否睡着。此外，这能够改进睡眠状态的判断，以区分醒着的人的非活动状态和睡着了的人的睡眠状态。Therefore, using the amplitude of the heart signal and/or a variable related to the amplitude change of the temperature signal can determine whether a person is asleep. In addition, this can improve the judgment of the sleep state to distinguish between the inactive state of an awake person and the sleep state of a asleep person.

心脏信号是在人的身体上测量的，例如，是在人的手指、手腕、手和/或手臂上测量的。也就是说，血容量变化可以是人在例如手指、手腕、手和/或手臂上的血容量变化。术语“心脏活动”和“心脏运行”可用作“心脏的运行”的同义词。因此，术语“指示心脏活动的信号”和“指示心脏运行的信号”可以用作术语“心脏信号”的同义词。The cardiac signal is measured on the human body, for example, on the human finger, wrist, hand and/or arm. That is, the blood volume change may be a blood volume change of the human, for example, on the finger, wrist, hand and/or arm. The terms "heart activity" and "heart operation" may be used as synonyms of "heart operation". Therefore, the terms "signal indicating heart activity" and "signal indicating heart operation" may be used as synonyms of the term "heart signal".

心脏信号的幅度是心脏信号随时间变化的峰值和谷值之间的差值。幅度变化表示心脏信号的峰谷差随时间变化或改变的程度，例如，在一段时间内。一段时间内的幅度变化越大，这个时间段内的幅度变化程度或改变程度就越大，反之亦然。心脏信号的幅度变化可以称为心脏信号的幅度的时间变化。The amplitude of the heart signal is the difference between the peak and valley values of the heart signal over time. The amplitude variation indicates the degree to which the peak-to-valley difference of the heart signal varies or changes over time, for example, over a period of time. The greater the amplitude variation over a period of time, the greater the degree of amplitude variation or change over this period of time, and vice versa. The amplitude variation of the heart signal can be referred to as the time variation of the amplitude of the heart signal.

所述装置可以用于从传感器接收心脏信号，该传感器用于生成心脏信号。这样的传感器可以称为心功能传感器。心脏信号指示人的心率和/或血容量变化。心脏信号可以包括一系列峰值或脉冲。The device may be used to receive a cardiac signal from a sensor that is used to generate a cardiac signal. Such a sensor may be referred to as a cardiac function sensor. The cardiac signal indicates changes in a person's heart rate and/or blood volume. The cardiac signal may include a series of peaks or pulses.

与使用根据人的心率数据的心率变异性(heart rate variability，HRV)来判断人是否睡着相比，使用心脏信号的幅度是有利的。与计算心脏信号的幅度相比，计算HRV更容易出错。也就是说，为了计算HRV，需要检测心脏信号的每个峰值。也就是说，缺少心率的一个或多个峰值会降低所计算的HRV的正确性(或质量)，从而可能根据HRV导致错误的睡眠状态判断。相反，心脏信号的幅度可以在几分钟的时间范围内检测。因此，在计算心脏信号的幅度时，可以取几个心脏搏动的平均值。因此，心脏信号的一个或多个峰值的缺失不太可能对所计算的幅度产生很大的影响。Compared with using heart rate variability (HRV) based on a person's heart rate data to determine whether a person is asleep, using the amplitude of the heart signal is advantageous. Calculating HRV is more prone to error than calculating the amplitude of the heart signal. That is, in order to calculate HRV, each peak of the heart signal needs to be detected. That is, the lack of one or more peaks of the heart rate will reduce the correctness (or quality) of the calculated HRV, which may lead to an erroneous judgment of the sleep state based on HRV. In contrast, the amplitude of the heart signal can be detected within a time range of several minutes. Therefore, when calculating the amplitude of the heart signal, the average of several heart beats can be taken. Therefore, the absence of one or more peaks of the heart signal is unlikely to have a significant impact on the calculated amplitude.

温度信号的幅度变化是温度信号的幅度随时间的变化。该变化指示温度信号的幅度随时间变化或改变的程度，例如，在一段时间内。一段时间内温度信号的幅度变化越大，该时间段内温度信号的幅度变化的程度或幅度改变的程度就越大，反之亦然。温度信号的幅度变化可以称为温度信号的幅度的时间变化。The amplitude variation of a temperature signal is the variation of the amplitude of the temperature signal over time. The variation indicates the degree to which the amplitude of the temperature signal varies or changes over time, for example, over a period of time. The greater the amplitude variation of the temperature signal over a period of time, the greater the degree of amplitude variation or the degree of amplitude change of the temperature signal over that period of time, and vice versa. The amplitude variation of a temperature signal may be referred to as the temporal variation of the amplitude of the temperature signal.

在某些情况下，当人不活动，但仍醒着(即未睡着)时，人的远端皮肤温度就会升高。例如，有时当人在晚上不活动时，皮肤温度在人睡着之前就已经升高了。因此，与直接使用人的温度(例如，人的温度的绝对值或温度信号的实际幅度)相比，使用与温度信号的幅度变化相关的变量对于判断人是否睡着是有利的。In some cases, when a person is inactive but still awake (i.e., not asleep), the person's distal skin temperature increases. For example, sometimes when a person is inactive at night, the skin temperature increases before the person falls asleep. Therefore, compared to directly using the person's temperature (e.g., the absolute value of the person's temperature or the actual amplitude of the temperature signal), using a variable related to the amplitude change of the temperature signal is advantageous for determining whether the person is asleep.

人的温度可以是皮肤区域(也可以是远端皮肤区域)的温度或者人的体温。换句话说，人的温度可以是皮肤温度(例如，远端皮肤温度)或者体温(例如，人的身体核心温度或近端温度)。例如，可以在肢体(可以称为四肢)处测量人的温度，例如，在人的手指、手腕、手和/或手臂上测量。在这种情况下，温度是远端皮肤温度。所述装置可以用于从温度传感器接收温度信号，该温度传感器用于生成温度信号。温度信号指示人的温度(例如，皮肤温度，也可以是远端皮肤温度)。因此，当前时刻的温度信号的幅度(即当前幅度)可以等于或表示人在当前时刻的温度(即当前温度)。换句话说，温度信号的幅度可以是人的温度值。The temperature of a person may be the temperature of a skin area (or a distal skin area) or the body temperature of the person. In other words, the temperature of a person may be the skin temperature (e.g., the distal skin temperature) or the body temperature (e.g., the core temperature or the proximal temperature of the body of the person). For example, the temperature of a person may be measured at a limb (which may be referred to as a limb), for example, at a finger, wrist, hand, and/or arm of the person. In this case, the temperature is the distal skin temperature. The device may be used to receive a temperature signal from a temperature sensor, which is used to generate a temperature signal. The temperature signal indicates the temperature of the person (e.g., the skin temperature, or the distal skin temperature). Therefore, the amplitude of the temperature signal at the current moment (i.e., the current amplitude) may be equal to or represent the temperature of the person at the current moment (i.e., the current temperature). In other words, the amplitude of the temperature signal may be the temperature value of the person.

所述装置可以用于根据心脏信号计算心脏信号的幅度。所述装置可以用于根据温度信号计算与温度信号的幅度变化相关的变量。可选地，计算心脏信号的幅度可以是计算指示心脏信号的幅度的变量。本文关于心脏信号幅度的描述相应地对于指示心脏信号幅度的变量是有效的。The device may be used to calculate the amplitude of the cardiac signal from the cardiac signal. The device may be used to calculate a variable related to the amplitude change of the temperature signal from the temperature signal. Optionally, calculating the amplitude of the cardiac signal may be calculating a variable indicating the amplitude of the cardiac signal. The description herein about the amplitude of the cardiac signal is correspondingly valid for the variable indicating the amplitude of the cardiac signal.

所述装置可以包括或者是用于执行本文所述的装置的功能的处理器、微处理器、控制器、微控制器、现场可编程门阵列(field-programmable gate array，FPGA)、专用集成电路或其任何组合。所述装置可以是计算机。所述计算机可以包括至少一个处理器和至少一个数据存储器。The device may include or be a processor, microprocessor, controller, microcontroller, field-programmable gate array (FPGA), application specific integrated circuit or any combination thereof for performing the functions of the device described herein. The device may be a computer. The computer may include at least one processor and at least one data storage.

人可以被称为装置的用户或使用者。A person may be referred to as a user or operator of a device.

在所述第一方面的一种实现方式中，所述心脏信号为光电体积描记(photoplethysmographic，PPG)信号。In an implementation of the first aspect, the cardiac signal is a photoplethysmographic (PPG) signal.

所述装置可以用于从PPG传感器接收PPG信号，所述PPG传感器用于生成PPG信号。由于人的心功能的变化，PPG信号的幅度会发生变化。PPG信号幅度在睡眠开始时(即人睡着了时)(显著)增大，在睡眠结束时(即人醒着时)降低。The device may be used to receive a PPG signal from a PPG sensor, which is used to generate a PPG signal. Due to changes in a person's cardiac function, the amplitude of the PPG signal may change. The PPG signal amplitude increases (significantly) at the beginning of sleep (i.e. when the person is asleep) and decreases at the end of sleep (i.e. when the person is awake).

在所述第一方面的一种实现方式中，所述装置用于通过计算后续时间窗口的心脏信号的平均幅度来计算或确定心脏信号的幅度。In an implementation of the first aspect, the apparatus is configured to calculate or determine the amplitude of the cardiac signal by calculating an average amplitude of the cardiac signal in subsequent time windows.

换句话说，为了计算或确定所述心脏信号的幅度，所述装置可以用于针对两个或更多后续时间窗口计算心脏信号的平均幅度。In other words, in order to calculate or determine the amplitude of the cardiac signal, the apparatus may be configured to calculate an average amplitude of the cardiac signal for two or more subsequent time windows.

这能够抵消心脏信号(例如，PPG信号)中可能存在的由于运动伪像而产生的噪声。也就是说，人的运动可能会对心脏信号的幅度产生影响。由于心脏信号的幅度应该仅受人的心脏的运行影响，因此人的运动对心脏信号的幅度的影响代表噪声。计算后续时间窗口的心脏信号的平均幅度可以滤除心脏信号中的运动(即运动伪像)引起的噪声。每个时间窗口可以具有例如一分钟(1分钟)的持续时间。This can cancel out noise due to motion artifacts that may be present in the heart signal (e.g., PPG signal). That is, the motion of a person may have an effect on the amplitude of the heart signal. Since the amplitude of the heart signal should only be affected by the operation of the person's heart, the effect of the person's motion on the amplitude of the heart signal represents noise. Calculating the average amplitude of the heart signal for subsequent time windows can filter out the noise caused by motion (i.e., motion artifacts) in the heart signal. Each time window can have a duration of, for example, one minute (1 min).

在所述第一方面的一种实现方式中，所述装置用于在所述时间窗口中的一个时间窗口期间，所述心脏信号的幅度变化(峰值和谷值差值的变化)和/或所述心脏信号的峰值之间的时间距离的变化在心功能的正常生理限制内的情况下，确定所述心脏信号在所述时间窗口期间的信号质量等于或大于所述信号质量的阈值。此外，所述装置可以用于仅针对所述心脏信号的信号质量等于或大于所述信号质量的所述阈值的时间窗口，计算所述心脏信号的所述平均幅度。In an implementation of the first aspect, the device is used to determine that the signal quality of the cardiac signal during the time window is equal to or greater than the threshold value of the signal quality when the change in the amplitude of the cardiac signal (the change in the difference between the peak and the valley value) and/or the change in the time distance between the peak values of the cardiac signal during one of the time windows is within the normal physiological limit of cardiac function. In addition, the device can be used to calculate the average amplitude of the cardiac signal only for the time window in which the signal quality of the cardiac signal is equal to or greater than the threshold value of the signal quality.

换句话说，可以在时间窗口期间对心脏信号进行关于信号质量的评分。心脏信号的信号质量越低，心脏信号中存在的噪声(运动伪影)就越多，反之亦然。通过仅针对心脏信号的信号质量等于或大于信号质量的阈值(即具有足够的信号质量)的时间窗口计算心脏信号的平均幅度，可以针对这样的时间窗口计算心脏信号的平均幅度。换句话说，可以在心脏信号的信号质量足够高的时间内，即在心脏信号中没有太多的噪声(例如，由于运动伪像)的时间内，计算或确定心脏信号的幅度。这样就保证了心脏信号的幅度是由于人的心脏的运行而不是由于人的运动而引起的。In other words, the cardiac signal can be scored regarding signal quality during a time window. The lower the signal quality of the cardiac signal, the more noise (motion artifacts) there is in the cardiac signal, and vice versa. By calculating the average amplitude of the cardiac signal only for time windows in which the signal quality of the cardiac signal is equal to or greater than a threshold value of the signal quality (i.e., having sufficient signal quality), the average amplitude of the cardiac signal can be calculated for such time windows. In other words, the amplitude of the cardiac signal can be calculated or determined within a time in which the signal quality of the cardiac signal is high enough, i.e., within a time in which there is not much noise (e.g., due to motion artifacts) in the cardiac signal. This ensures that the amplitude of the cardiac signal is due to the operation of the human heart and not due to the movement of the human.

只要心脏信号的幅度变化和/或心脏信号的峰值之间的时间距离变化在心功能正常生理限制范围内，心脏信号的幅度主要就是由心脏引起的。也就是说，其他因素(例如，人的运动)对心脏信号幅度的影响可以忽略不计。As long as the amplitude variation of the cardiac signal and/or the temporal distance variation between the peak values of the cardiac signal are within the normal physiological limits of cardiac function, the amplitude of the cardiac signal is mainly caused by the heart. That is, the influence of other factors (e.g., human motion) on the amplitude of the cardiac signal can be ignored.

可选地，为了确定在所述时间窗口中的一个时间窗口期间的心脏信号的信号质量，所述装置可以用于确定所述时间窗口期间的心脏信号的后续峰值的数量以及心脏信号的后续峰值之间的时间距离。Optionally, in order to determine the signal quality of the heart signal during one of the time windows, the apparatus may be configured to determine a number of subsequent peaks of the heart signal during the time window and a time distance between subsequent peaks of the heart signal.

所述装置可以用于比较时间窗口期间的心脏信号的峰值。所述装置可以用于在峰值彼此变化的程度大于峰值变化的阈值的情况下，确定所述时间窗口期间的心脏信号的信号质量低于信号质量的阈值。所述装置可以用于将时间窗口期间的心脏信号的峰谷差(可称为大小)彼此进行比较。所述装置可以用于在所述时间窗口期间的所述心脏信号的幅度变化的程度大于幅度变化的阈值的情况下，确定所述时间窗口期间的所述心脏信号的信号质量低于信号质量的阈值。换句话说，所述装置可以用于通过确定所述心脏信号的峰值和谷值在所述时间窗口期间彼此相似的程度来确定所述时间窗口期间的所述心脏信号的信号质量。此外或替代地，所述装置可以用于确定时间窗口期间的心脏信号的峰值之间的时间距离(持续时间)。所述装置可以用于在所述时间窗口期间所述峰值之间的时间距离变化的程度大于时间距离变化的阈值的情况下，确定所述时间窗口期间的所述心脏信号的信号质量低于信号质量的阈值。The device may be used to compare the peak values of the cardiac signal during the time window. The device may be used to determine that the signal quality of the cardiac signal during the time window is lower than the threshold value of the signal quality when the degree of variation of the peak values from each other is greater than the threshold value of the variation of the peak values. The device may be used to compare the peak-to-valley differences (which may be referred to as magnitudes) of the cardiac signal during the time window with each other. The device may be used to determine that the signal quality of the cardiac signal during the time window is lower than the threshold value of the signal quality when the degree of variation of the amplitude of the cardiac signal during the time window is greater than the threshold value of the variation of the amplitude. In other words, the device may be used to determine the signal quality of the cardiac signal during the time window by determining the degree to which the peak values and valley values of the cardiac signal are similar to each other during the time window. In addition or alternatively, the device may be used to determine the time distance (duration) between the peak values of the cardiac signal during the time window. The device may be used to determine that the signal quality of the cardiac signal during the time window is lower than the threshold value of the signal quality when the degree of variation of the time distance between the peak values during the time window is greater than the threshold value of the variation of the time distance.

在所述第一方面的一种实现方式中，所述装置用于在所述心脏信号的幅度增大到超过所述心脏信号的幅度的第一阈值的情况下，判断人睡着了。此外，所述装置可以用于在所述心脏信号的幅度降低到低于所述心脏信号的幅度的第二阈值的情况下，判断人醒着。In an implementation of the first aspect, the device is used to determine that the person is asleep when the amplitude of the heart signal increases to exceed a first threshold of the amplitude of the heart signal. In addition, the device can be used to determine that the person is awake when the amplitude of the heart signal decreases to below a second threshold of the amplitude of the heart signal.

换句话说，所述装置可以用于在所述心脏信号的幅度大于所述心脏信号的幅度的第一阈值的情况下，判断人睡着了，在所述心脏信号的幅度小于所述心脏信号的幅度的第二阈值的情况下，判断人醒着。心脏信号的幅度的第一阈值可以等于或大于心脏信号的幅度的第二阈值。In other words, the device can be used to determine that the person is asleep when the amplitude of the heart signal is greater than a first threshold of the amplitude of the heart signal, and to determine that the person is awake when the amplitude of the heart signal is less than a second threshold of the amplitude of the heart signal. The first threshold of the amplitude of the heart signal can be equal to or greater than the second threshold of the amplitude of the heart signal.

心脏信号(例如，PPG信号)的幅度的第一阈值和第二阈值可以取决于心功能传感器如何安装在人身上(或由人佩戴)，用于测量人的心功能并生成心脏信号。心功能传感器的可能改变的佩戴状况例如是传感器佩戴在身体上的位置(例如，身体四肢上的位置)以及传感器佩戴的紧密程度。在装置开始判断人是否睡着之前，当人在当前佩戴状况下醒着时，可以确定心脏信号的平均幅度。根据该确定的心脏信号的平均幅度，可以确定心脏信号的幅度的第一阈值和第二阈值。可选地，所述装置可以用于在开始睡眠状态确定之前针对当前佩戴状况确定当人醒着时心脏信号的平均幅度。这可以是在装置开始判断人是否睡着之前(例如，当人激活或打开用于睡眠状态判断的装置时)由装置执行的设置过程。该装置可以用于确定心脏信号的平均幅度、心脏信号的幅度的第一阈值和第二阈值。由此，可以根据心脏信号和/或温度信号的幅度来判断人的清醒状态，如本文稍后所述。所述装置可以用于存储和/或接收心脏信号的幅度的第一阈值和第二阈值。The first threshold and the second threshold of the amplitude of the heart signal (e.g., PPG signal) may depend on how the heart function sensor is mounted on a person (or worn by a person) to measure the heart function of the person and generate a heart signal. The wearing condition of the heart function sensor that may change is, for example, the position of the sensor worn on the body (e.g., the position on the limbs of the body) and the tightness of the sensor wearing. Before the device starts to determine whether the person is asleep, when the person is awake in the current wearing condition, the average amplitude of the heart signal can be determined. According to the determined average amplitude of the heart signal, the first threshold and the second threshold of the amplitude of the heart signal can be determined. Optionally, the device can be used to determine the average amplitude of the heart signal when the person is awake for the current wearing condition before starting the sleep state determination. This can be a setting process performed by the device before the device starts to determine whether the person is asleep (e.g., when the person activates or turns on the device for sleep state determination). The device can be used to determine the average amplitude of the heart signal, the first threshold and the second threshold of the amplitude of the heart signal. Thus, the awake state of the person can be determined according to the amplitude of the heart signal and/or the temperature signal, as described later in this article. The device can be used to store and/or receive the first threshold and the second threshold of the amplitude of the heart signal.

在所述第一方面的一种实现方式中，与温度信号的幅度变化相关的变量指示温度信号的幅度变化、温度信号的幅度变化的频率和温度信号的幅度变化的速率中的至少一个。In an implementation of the first aspect, the variable related to the amplitude change of the temperature signal indicates at least one of the amplitude change of the temperature signal, the frequency of the amplitude change of the temperature signal, and the rate of the amplitude change of the temperature signal.

温度信号的幅度变化的频率表示在一段时间内，温度信号的幅度变化发生的次数(即多久一次)。在一段时间内，温度信号的幅度变化的频率越大，在该时间段内，温度信号的幅度变化发生的次数就越多(即更频繁)，反之亦然。The frequency of the amplitude change of the temperature signal indicates the number of times (i.e., how often) the amplitude change of the temperature signal occurs within a period of time. The greater the frequency of the amplitude change of the temperature signal within a period of time, the more times (i.e., more frequently) the amplitude change of the temperature signal occurs within the period of time, and vice versa.

温度信号的幅度变化的速率指示温度信号的幅度变化发生的速度。在一段时间内，温度信号的幅度变化的速率越大，温度信号的幅度变化在该时间段内发生的速度越快，反之亦然。The rate of change of the amplitude of the temperature signal indicates the speed at which the amplitude change of the temperature signal occurs. In a period of time, the greater the rate of change of the amplitude of the temperature signal, the faster the amplitude change of the temperature signal occurs in the period of time, and vice versa.

如上所述，在睡眠开始时(即人睡着了时)，身体热调节随着身体活动的减少而降低，而在睡眠结束时(人醒着时)，身体热调节会增加。身体热调节(热控制)的降低(减少或削弱)可以明显地看出与温度信号的幅度变化相关的变量的减少。该变量可以称为时域变量。As mentioned above, at the beginning of sleep (i.e. when a person is asleep), body thermal regulation decreases as body activity decreases, while at the end of sleep (when a person is awake), body thermal regulation increases. The decrease (reduction or weakening) of body thermal regulation (thermal control) can be clearly seen as a decrease in the variable associated with the amplitude change of the temperature signal. This variable can be called a time domain variable.

在所述第一方面的一种实现方式中，所述装置用于根据所述温度信号，通过计算以下各项中的至少一项来计算与所述温度信号的幅度变化相关的所述变量：滑动时间窗口内的温度信号导数的至少两个绝对值之和；针对时间窗口的所述温度信号的傅里叶变换；在滑动时间窗口或固定阶跃时间窗口期间的所述温度信号的最大值和最小值。In an implementation of the first aspect, the device is used to calculate the variable related to the amplitude change of the temperature signal based on the temperature signal by calculating at least one of the following items: the sum of at least two absolute values of the temperature signal derivative within a sliding time window; the Fourier transform of the temperature signal for the time window; the maximum and minimum values of the temperature signal during a sliding time window or a fixed step time window.

换句话说，所述装置可以用于根据所述温度信号，通过计算滑动时间窗口内的温度信号导数的至少两个绝对值之和来计算所述变量。附加地或可替代地，所述装置可以用于根据所述温度信号，通过计算针对时间窗口的所述温度信号的傅里叶变换来计算所述变量。附加地或可替代地，所述装置可以用于根据所述温度信号，通过计算滑动时间窗口或固定阶跃时间窗口内的温度信号的最大值和最小值来计算变量。In other words, the device may be used to calculate the variable based on the temperature signal by calculating the sum of at least two absolute values of the derivative of the temperature signal within a sliding time window. Additionally or alternatively, the device may be used to calculate the variable based on the temperature signal by calculating the Fourier transform of the temperature signal for a time window. Additionally or alternatively, the device may be used to calculate the variable based on the temperature signal by calculating the maximum and minimum values of the temperature signal within a sliding time window or a fixed step time window.

术语“滚动时间窗口”可以用作术语“滑动时间窗口”的同义词。The term "tumbling time window" may be used as a synonym for the term "sliding time window".

滑动时间窗口内的温度信号导数的至少两个绝对值(即两个或两个以上绝对值)之和可以指示温度信号的幅度变化的速率，也可以指示温度信号的幅度变化。时间窗口的温度信号的傅里叶变换可以指示温度信号的幅度变化的频率，也可以指示温度信号的幅度变化。温度信号在滑动时间窗口或固定阶跃时间窗口内的最大值和最小值可以指示温度信号的幅度变化。The sum of at least two absolute values (i.e., two or more absolute values) of the temperature signal derivative within the sliding time window can indicate the rate of change of the amplitude of the temperature signal, and can also indicate the amplitude change of the temperature signal. The Fourier transform of the temperature signal in the time window can indicate the frequency of the amplitude change of the temperature signal, and can also indicate the amplitude change of the temperature signal. The maximum and minimum values of the temperature signal within the sliding time window or the fixed step time window can indicate the amplitude change of the temperature signal.

计算滑动时间窗口内的温度信号导数的至少两个绝对值之和可以如下进行：取当前时间窗口的温度信号的导数，然后取针对当前时间窗口计算的温度信号导数的绝对值。接下来，将所计算的温度信号导数的绝对值(针对滑动时间窗口的两个或两个以上当前时间窗口计算)求和。Calculating the sum of at least two absolute values of the temperature signal derivative within the sliding time window can be performed as follows: taking the derivative of the temperature signal for the current time window, then taking the absolute value of the temperature signal derivative calculated for the current time window. Next, summing the absolute values of the calculated temperature signal derivatives (calculated for two or more current time windows of the sliding time window).

在所述第一方面的一种实现方式中，所述装置用于在与所述温度信号的幅度变化相关的所述变量降低到低于所述变量的第一阈值的情况下，判断人睡着了。此外，所述装置可以用于在与所述温度信号的幅度变化相关的变量增大到超过所述变量的第二阈值的情况下，判断人醒着。In an implementation of the first aspect, the device is used to determine that the person is asleep when the variable related to the amplitude change of the temperature signal decreases to a value lower than a first threshold value of the variable. In addition, the device can be used to determine that the person is awake when the variable related to the amplitude change of the temperature signal increases to a value higher than a second threshold value of the variable.

换句话说，所述装置可以用于在与所述温度信号的幅度变化相关的变量小于变量的第一阈值的情况下，判断人睡着了，在与所述温度信号的幅度变化相关的变量大于变量的第二阈值的情况下，判断人醒着。变量的第一阈值可以等于或小于变量的第二阈值。所述装置可以用于存储和/或接收与温度信号的幅度变化相关的变量的第一阈值和第二阈值。In other words, the device can be used to determine that the person is asleep when the variable related to the amplitude change of the temperature signal is less than a first threshold value of the variable, and to determine that the person is awake when the variable related to the amplitude change of the temperature signal is greater than a second threshold value of the variable. The first threshold value of the variable can be equal to or less than the second threshold value of the variable. The device can be used to store and/or receive the first threshold value and the second threshold value of the variable related to the amplitude change of the temperature signal.

与温度信号的幅度变化相关的变量的第一阈值和第二阈值可以根据(或取决于)机器学习模型来确定，该机器学习模型可由所述装置用于至少根据变量来判断人是否睡着，如本文稍后所述。可选地，所述装置可以用于确定与温度信号的幅度变化相关的变量的第一阈值和第二阈值。例如，在机器学习模型包括或者是决策树模型的情况下，第一阈值和第二阈值可以取决于其他变量及其阈值，可选地，是否使用第一阈值和第二阈值可以取决于这些其他变量及其阈值。The first threshold and the second threshold of the variable related to the amplitude change of the temperature signal can be determined according to (or depending on) a machine learning model, which can be used by the device to determine whether a person is asleep at least based on the variable, as described later in this document. Optionally, the device can be used to determine the first threshold and the second threshold of the variable related to the amplitude change of the temperature signal. For example, in the case where the machine learning model includes or is a decision tree model, the first threshold and the second threshold can depend on other variables and their thresholds, and optionally, whether to use the first threshold and the second threshold can depend on these other variables and their thresholds.

所述装置可以用于在满足以下条件中的至少一个条件的情况下，判断人睡着了：温度信号的幅度变化降低到低于温度信号的幅度变化的第一阈值。温度信号的幅度变化的频率降低到低于温度信号的幅度变化的频率的第一阈值。温度信号的幅度变化的速率降低到低于温度信号的幅度变化的速率的第一阈值。The device can be used to determine that a person has fallen asleep when at least one of the following conditions is met: the amplitude change of the temperature signal decreases to a value lower than a first threshold value of the amplitude change of the temperature signal. The frequency of the amplitude change of the temperature signal decreases to a value lower than a first threshold value of the frequency of the amplitude change of the temperature signal. The rate of the amplitude change of the temperature signal decreases to a value lower than a first threshold value of the rate of the amplitude change of the temperature signal.

所述装置可以用于在满足以下条件中的至少一个条件的情况下判断人醒着：温度信号的幅度变化增大到超过温度信号的幅度变化的第二阈值。温度信号的幅度变化的频率增大到超过温度信号的幅度变化的频率的第二阈值。温度信号的幅度变化的速率增大到超过温度信号的幅度变化的速率的第二阈值。The device can be used to determine that the person is awake when at least one of the following conditions is met: the amplitude change of the temperature signal increases to exceed the second threshold value of the amplitude change of the temperature signal. The frequency of the amplitude change of the temperature signal increases to exceed the second threshold value of the frequency of the amplitude change of the temperature signal. The rate of the amplitude change of the temperature signal increases to exceed the second threshold value of the rate of the amplitude change of the temperature signal.

上述相应的第一阈值可以等于或小于上述相应的第二阈值。所述装置可以用于存储和/或接收相应的第一阈值和第二阈值。The corresponding first threshold may be equal to or less than the corresponding second threshold. The apparatus may be configured to store and/or receive the corresponding first threshold and the second threshold.

在所述第一方面的一种实现方式中，所述装置用于根据所述心脏信号的所述幅度和与所述温度信号的幅度变化相关的所述变量中的至少一个，使用经训练的机器学习模型来判断人是否睡着。根据包括多个数据集的训练数据，训练所述经训练的机器学习模型。多个数据集中的每个数据集包括睡眠状态变量，指示人是否在相应时间睡着了，与以下内容相关联：在所述相应时间的所述心脏信号的所述幅度和与所述温度信号的幅度变化相关的所述变量的中至少一个。In an implementation of the first aspect, the device is used to determine whether a person is asleep using a trained machine learning model based on the amplitude of the heart signal and at least one of the variables associated with the amplitude change of the temperature signal. The trained machine learning model is trained based on training data including multiple data sets. Each of the multiple data sets includes a sleep state variable indicating whether the person is asleep at a corresponding time, which is associated with: at least one of the amplitude of the heart signal and the variables associated with the amplitude change of the temperature signal at the corresponding time.

机器学习模型(根据训练数据进行训练，即经训练的机器学习模型)可以包括或者是决策树模型、随机森林模型、神经网络模型、深度神经网络模型或隐马尔可夫模型。可以替代地使用本领域已知的任何其他机器学习模型。所述训练数据的数据集取决于所述装置是否用于根据所述心脏信号的幅度，或根据与所述温度信号的幅度变化相关的变量，或根据所述心脏信号的幅度和与所述温度信号的幅度变化相关的变量，使用经训练的机器学习模型来判断人的睡眠状态。在上述第一替代方案的情况下，所述多个数据集中的每个数据集包括睡眠状态变量，所述睡眠状态变量指示与所述心脏信号的幅度相关联的人是否在相应时间睡着了。在上述第二替代方案的情况下，多个数据集中的每个数据集包括与在相应时间与温度信号的幅度变化相关的变量相关联的睡眠状态变量。在上述最后一个替代方案的情况下，多个数据集中的每个数据集包括与心脏信号的幅度和在相应时间与温度信号的幅度变化相关的变量相关联的睡眠状态变量。The machine learning model (trained according to the training data, i.e., the trained machine learning model) may include or be a decision tree model, a random forest model, a neural network model, a deep neural network model, or a hidden Markov model. Any other machine learning model known in the art may be used instead. The data set of the training data depends on whether the device is used to judge the sleep state of a person based on the amplitude of the heart signal, or based on the variable associated with the amplitude change of the temperature signal, or based on the amplitude of the heart signal and the variable associated with the amplitude change of the temperature signal, using the trained machine learning model. In the case of the first alternative scheme described above, each of the multiple data sets includes a sleep state variable, which indicates whether the person associated with the amplitude of the heart signal fell asleep at the corresponding time. In the case of the second alternative scheme described above, each of the multiple data sets includes a sleep state variable associated with a variable associated with the amplitude change of the temperature signal at the corresponding time. In the case of the last alternative scheme described above, each of the multiple data sets includes a sleep state variable associated with the amplitude of the heart signal and a variable associated with the amplitude change of the temperature signal at the corresponding time.

例如，可以通过迭代执行优化算法，来使用训练数据训练机器学习模型。例如，首先，在优化算法的迭代中，将训练数据的相应数据集的心脏信号的幅度和/或与温度信号的幅度变化相关的变量输入到机器学习模型中。机器学习模型的输出由机器学习模型根据心脏信号的幅度和/或与相应数据集的温度信号的幅度变化相关的变量计算，是指示人是否睡着的睡眠状态变量。将计算出的睡眠状态变量(针对相应数据集计算的)与相应数据集的睡眠状态变量(即基本事实)进行比较。接下来，可以通过调整机器学习模型来减少它们之间的差异(即误差)。例如，可以通过调整机器学习模型的权重来减少该差异(即误差)，以处理相应数据集的输入数据。在调整机器学习模型之后，可以使用训练数据的多个数据集中的另一个数据集执行进一步的迭代。当迭代的计算误差小于误差的阈值时，可以停止迭代执行优化算法。For example, the machine learning model can be trained using training data by iteratively executing the optimization algorithm. For example, first, in the iteration of the optimization algorithm, the amplitude of the heart signal of the corresponding data set of the training data and/or the variables related to the amplitude change of the temperature signal are input into the machine learning model. The output of the machine learning model is calculated by the machine learning model based on the amplitude of the heart signal and/or the variables related to the amplitude change of the temperature signal of the corresponding data set, which is a sleep state variable indicating whether the person is asleep. The calculated sleep state variable (calculated for the corresponding data set) is compared with the sleep state variable (i.e., the basic facts) of the corresponding data set. Next, the difference (i.e., error) between them can be reduced by adjusting the machine learning model. For example, the difference (i.e., error) can be reduced by adjusting the weights of the machine learning model to process the input data of the corresponding data set. After adjusting the machine learning model, further iterations can be performed using another data set in the multiple data sets of the training data. When the iterative calculation error is less than the error threshold, the iterative execution of the optimization algorithm can be stopped.

睡眠状态变量可以是分类结果，指示人是否睡着。每个数据集的睡眠状态变量是相应数据集(在相应时间内)的基本事实。睡眠状态变量可以是二元变量，其中，二元变量的第一值(例如，零“0”或一“1”)可以指示人正在睡眠，二元变量的第二值(例如，分别为一“1”或零“0”)可以指示人没有睡，即人醒着。或者，睡眠状态变量可以是数字范围的数字(例如，介于0％和100％之间的百分比值)，其中，大于数字范围的阈值的数字(例如，阈值百分比)可以指示人正在睡眠，而小于该范围的阈值的数字(例如，阈值百分比)可以指示人没睡着，而是醒着的。对于反之亦然的情况也可能如此，即，大于阈值的数字可以指示人是醒着的，并且小于阈值的数字可以指示人是睡着的。The sleep state variable can be a classification result, indicating whether the person is asleep. The sleep state variable of each data set is the basic fact of the corresponding data set (at the corresponding time). The sleep state variable can be a binary variable, wherein the first value of the binary variable (e.g., zero "0" or one "1") can indicate that the person is sleeping, and the second value of the binary variable (e.g., one "1" or zero "0", respectively) can indicate that the person is not asleep, that is, the person is awake. Alternatively, the sleep state variable can be a number in a digital range (e.g., a percentage value between 0% and 100%), wherein a number greater than a threshold value of the digital range (e.g., a threshold percentage) can indicate that the person is sleeping, and a number less than the threshold value of the range (e.g., a threshold percentage) can indicate that the person is not asleep, but awake. This may also be true for the vice versa, that is, a number greater than a threshold value can indicate that the person is awake, and a number less than a threshold value can indicate that the person is asleep.

在所述第一方面的一种实现方式中，所述装置用于接收与人的加速度相关的加速度信号。此外，该装置可以用于根据加速度信号计算人的活动程度。此外，该装置可以用于除心脏信号的幅度和与温度信号的幅度变化相关的变量中的至少一个之外，根据活动程度判断人是否睡着。In an implementation of the first aspect, the device is used to receive an acceleration signal related to the acceleration of a person. In addition, the device can be used to calculate the activity level of the person based on the acceleration signal. In addition, the device can be used to determine whether the person is asleep based on the activity level in addition to at least one of the amplitude of the heart signal and the variable related to the amplitude change of the temperature signal.

人的加速度可以是或者可以包括人的一个或多个身体部位的加速度。例如，人的加速度可以是人的身体区域的加速度，这个人安装或佩戴了用于生成加速度信号的加速度计。The acceleration of a person may be or may include the acceleration of one or more body parts of a person. For example, the acceleration of a person may be the acceleration of a body region of a person who is equipped with or wearing an accelerometer for generating an acceleration signal.

例如，该装置可以用于根据加速度信号使用算法计算人的活动程度。这种算法可以称为睡眠评分算法或活动算法。这种算法的示例包括Cole-Kripke算法[Cole RJ,KripkeDF,Gruen W,Mullaney DJ,GillinJC.Automatic sleep/wake identification fromwrist activity.Sleep.1992年10月；15(5)：461-9.]、UCSD评分算法[Jean-Louis G,Kripke DF,Mason WJ,Elliott JA,Youngstedt SD.Sleep estimation from wristmovement quantified by different actigraphic modalities.J NeurosciMethods.2001年2月15日；105(2):185-91.doi:10.1016/s0165-0270(00)00364-2.PMID:11275275.]和Sedeh等人的睡眠算法[Sadeh A,Sharkey KM,Carskadon MA.Activity-based sleep-wake identification:an empirical test of methodologicalissues.Sleep.1994年4月；17(3):201-7.]。For example, the device can be used to calculate the activity level of a person using an algorithm based on the acceleration signal. Such an algorithm can be called a sleep scoring algorithm or an activity algorithm. Examples of such algorithms include the Cole-Kripke algorithm [Cole RJ, KripkeDF, Gruen W, Mullaney DJ, GillinJC. Automatic sleep/wake identification from wrist activity. Sleep. 1992 Oct; 15(5): 461-9.], the UCSD scoring algorithm [Jean-Louis G, Kripke DF, Mason WJ, Elliott JA, Youngstedt SD. Sleep estimation from wristmovement quantified by different actigraphic modalities. J NeurosciMethods. 2001 Feb 15; 105(2): 185-91. doi: 10.1016/s0165-0270(00)00364-2. PMID: 11275275.], and the sleep algorithm of Sedeh et al. [Sadeh A, Sharkey KM, Carskadon MA. Activity-based sleep-wake identification: an empirical test of methodological issues. Sleep. 1994 April; 17(3): 201-7.].

例如，Cole-Kripke算法首先根据加速度变化来计算活动程度。术语“活动水平”可以用作术语“活动程度”的同义词。然后，可以计算时间窗口(例如，一分钟窗口)内的活动程度之和。接下来，可以使用例如以下公式为每个时间窗口(例如，一分钟窗口或一分钟周期)计算经调整的活动值：For example, the Cole-Kripke algorithm first calculates the activity level based on the acceleration change. The term "activity level" can be used as a synonym for the term "activity level". Then, the sum of the activity levels within a time window (e.g., a one-minute window) can be calculated. Next, an adjusted activity value can be calculated for each time window (e.g., a one-minute window or a one-minute period) using, for example, the following formula:

总活度＝E0+E1·0.2+(E-1)·0.2+E2·0.04+(E-2)·0.04。Total activity = E0 + E1·0.2 + (E-1)·0.2 + E2·0.04 + (E-2)·0.04.

在上述公式中，E0是感兴趣的时间窗口中的活动程度，E1是后一个时间窗口(例如，在一分钟窗口的情况下，晚一分钟)的活动程度，并且E-1是前一个时间窗口的活动程度(例如，在一分钟窗口的情况下，早一分钟)，依此类推。如果给定时间窗口(例如，1分钟时间窗口或1分钟周期)内的总活动小于或等于唤醒阈值，则该时间窗口(可称为周期)被评分为睡着。如果给定时间窗口期间的总活动大于唤醒阈值，则该时间窗口被评分为醒着。In the above formula, E0 is the activity level in the time window of interest, E1 is the activity level in the following time window (e.g., one minute later in the case of a one-minute window), and E-1 is the activity level in the previous time window (e.g., one minute earlier in the case of a one-minute window), and so on. If the total activity within a given time window (e.g., a 1-minute time window or a 1-minute cycle) is less than or equal to the wake threshold, then the time window (which may be referred to as a cycle) is scored as asleep. If the total activity during a given time window is greater than the wake threshold, then the time window is scored as awake.

活动程度(活动水平)可以是介于0％和100％之间的百分比值，其中，百分比值越大，人的活动就越大(例如，人移动的越多)，反之亦然。可选地，计算活动程度可以是计算指示活动程度的变量。本文关于活动程度的描述相应地对于指示活动程度的变量是有效的。The activity level may be a percentage value between 0% and 100%, wherein the greater the percentage value, the greater the activity of the person (e.g., the more the person moves), and vice versa. Alternatively, calculating the activity level may be calculating a variable indicating the activity level. The description herein regarding the activity level is correspondingly valid for the variable indicating the activity level.

在所述第一方面的一种实现方式中，所述装置用于在所述活动程度降低到低于所述活动程度的第一阈值的情况下，判断人睡着了。此外，所述装置用于在所述活动程度增大到超过所述活动程度的第二阈值的情况下，判断人醒着。In an implementation of the first aspect, the device is used to determine that the person is asleep when the activity level decreases to a first threshold value below the activity level. In addition, the device is used to determine that the person is awake when the activity level increases to a second threshold value above the activity level.

换句话说，所述装置可以用于在所述活动程度小于所述活动程度的第一阈值的情况下，判断人睡着了，并且在所述活动程度大于所述活动程度的第二阈值的情况下，判断人醒着。活动程度的第一阈值可以等于或小于活动程度的第二阈值。所述装置可以用于存储和/或接收所述活动程度的第一阈值和第二阈值。In other words, the device may be used to determine that the person is asleep when the activity level is less than a first threshold value of the activity level, and to determine that the person is awake when the activity level is greater than a second threshold value of the activity level. The first threshold value of the activity level may be equal to or less than the second threshold value of the activity level. The device may be used to store and/or receive the first threshold value and the second threshold value of the activity level.

在所述第一方面的一种实现方式中，所述装置用于接收与人的温度相关的温度信号。此外，所述装置可以用于除心脏信号的幅度和与温度信号的幅度变化相关的变量中的至少一个之外，根据温度信号的幅度判断人是否睡着。In an implementation of the first aspect, the device is used to receive a temperature signal related to the temperature of the person. In addition, the device can be used to determine whether the person is asleep based on the amplitude of the temperature signal in addition to at least one of the amplitude of the heart signal and the variable related to the amplitude change of the temperature signal.

如上所述，温度信号可以是装置可选地可接收的温度信号。上述关于温度信号的描述相应地是有效的。As mentioned above, the temperature signal may be a temperature signal that the device may optionally receive. The above description regarding the temperature signal is correspondingly valid.

在所述第一方面的一种实现方式中，所述装置用于在所述温度信号的幅度增大到超过所述温度信号的幅度的第一阈值的情况下，判断人睡着了。此外，所述装置可以用于在所述温度信号的幅度降低到低于所述温度信号的幅度的第二阈值的情况下，判断人醒着。In an implementation of the first aspect, the device is used to determine that the person is asleep when the amplitude of the temperature signal increases to exceed a first threshold of the amplitude of the temperature signal. In addition, the device can be used to determine that the person is awake when the amplitude of the temperature signal decreases to below a second threshold of the amplitude of the temperature signal.

换句话说，所述装置可以用于在所述温度信号的幅度大于所述温度信号的幅度的第一阈值的情况下，判断人睡着了，在所述温度信号的幅度小于所述温度信号的幅度的第二阈值的情况下，判断人醒着。温度信号的幅度的第一阈值可以等于或大于温度信号的幅度的第二阈值。所述装置可以用于存储和/或接收温度信号的幅度的第一阈值和第二阈值。In other words, the device can be used to determine that the person is asleep when the amplitude of the temperature signal is greater than a first threshold of the amplitude of the temperature signal, and to determine that the person is awake when the amplitude of the temperature signal is less than a second threshold of the amplitude of the temperature signal. The first threshold of the amplitude of the temperature signal can be equal to or greater than the second threshold of the amplitude of the temperature signal. The device can be used to store and/or receive the first threshold and the second threshold of the amplitude of the temperature signal.

在所述第一方面的一种实现方式中，所述装置用于根据所述心脏信号的所述幅度和与所述温度信号的幅度变化相关的所述变量中的至少一个，并且根据所述温度信号的所述活动程度和所述幅度中的至少一个，使用经训练的机器学习模型来判断人是否睡着。根据包括多个数据集的训练数据，训练所述经训练的机器学习模型。多个数据集中的每个数据集包括睡眠状态变量，指示人是否在相应时间睡着了，与以下内容相关联：在所述相应时间的所述心脏信号的所述幅度和与所述温度信号的幅度变化相关的所述变量的中至少一个，In an implementation of the first aspect, the device is used to determine whether a person is asleep based on the amplitude of the heart signal and at least one of the variables associated with the amplitude change of the temperature signal, and based on the activity level and at least one of the amplitude of the temperature signal, using a trained machine learning model. The trained machine learning model is trained based on training data including multiple data sets. Each of the multiple data sets includes a sleep state variable indicating whether a person is asleep at a corresponding time, and is associated with: the amplitude of the heart signal at the corresponding time and at least one of the variables associated with the amplitude change of the temperature signal,

与在所述相应时间的所述温度信号的所述活动程度和所述幅度中的至少一个相关联。is associated with at least one of the activity level and the amplitude of the temperature signal at the corresponding time.

上述关于经训练的机器学习模型的描述可以相应地是有效的。机器学习模型(根据训练数据进行训练，即经训练的机器学习模型)可以包括或者是决策树模型、随机森林模型、神经网络模型、深度神经网络模型或隐马尔可夫模型。可以替代地使用本领域已知的任何其他机器学习模型。训练数据的数据集取决于装置使用哪些输入来使用经训练的机器学习模型判断人是否睡着。所述多个数据集中的每个数据集包括睡眠状态变量，所述睡眠状态变量指示人是否在相应时间睡着了，与使用的相应输入相关联。如上所述，相应输入是或包括心脏信号的幅度和与温度信号的幅度变化相关的变量中的至少一个以及活动程度和温度信号的幅度中的至少一个。The above description of the trained machine learning model may be correspondingly valid. The machine learning model (trained according to the training data, i.e., the trained machine learning model) may include or be a decision tree model, a random forest model, a neural network model, a deep neural network model, or a hidden Markov model. Any other machine learning model known in the art may be used instead. The data set of the training data depends on which inputs the device uses to use the trained machine learning model to determine whether a person is asleep. Each of the multiple data sets includes a sleep state variable, which indicates whether a person is asleep at a corresponding time, and is associated with the corresponding input used. As described above, the corresponding input is or includes at least one of the amplitude of the heart signal and a variable related to the amplitude change of the temperature signal, and at least one of the activity level and the amplitude of the temperature signal.

如上所述，可以执行机器学习模型的训练，例如通过迭代执行优化算法。可以如上文所述实现优化算法。As described above, training of the machine learning model can be performed, for example, by iteratively executing an optimization algorithm. The optimization algorithm can be implemented as described above.

为了实现根据本发明的第一方面所述的装置，上述第一方面的部分或全部实现方式和可选特征可以相互结合。In order to implement the device according to the first aspect of the present invention, part or all of the implementation methods and optional features of the first aspect may be combined with each other.

本发明的第二方面提供了一种用于判断人是否睡着的系统。该系统包括如上所述的根据第一方面所述的装置。此外，所述系统包括：心功能传感器，用于生成所述心脏信号；和/或温度传感器，用于生成所述温度信号。The second aspect of the present invention provides a system for determining whether a person is asleep. The system comprises the apparatus according to the first aspect as described above. In addition, the system comprises: a cardiac function sensor for generating the cardiac signal; and/or a temperature sensor for generating the temperature signal.

心功能传感器可称为用于测量心功能的传感器。可选地，心功能传感器可以是包括多个基本传感器的传感器系统。心功能传感器可以用于测量(例如，非侵入地测量)人的心率和血容量变化，以生成心脏信号。可选地，温度传感器可以是包括多个基本传感器的传感器系统。温度传感器可以用于测量(例如，非侵入地测量)人的温度，以生成温度信号。该系统可以称为用于检测人是否睡着的系统。The cardiac function sensor may be referred to as a sensor for measuring cardiac function. Alternatively, the cardiac function sensor may be a sensor system including a plurality of basic sensors. The cardiac function sensor may be used to measure (e.g., non-invasively measure) a person's heart rate and blood volume changes to generate a cardiac signal. Alternatively, the temperature sensor may be a sensor system including a plurality of basic sensors. The temperature sensor may be used to measure (e.g., non-invasively measure) a person's temperature to generate a temperature signal. The system may be referred to as a system for detecting whether a person is asleep.

在所述第二方面的一种实现方式中，所述心功能传感器包括光电体积描记(photoplethysmographic，PPG)传感器。In an implementation of the second aspect, the cardiac function sensor includes a photoplethysmographic (PPG) sensor.

在所述第二方面的一种实现方式中，所述系统包括加速度计，用于生成加速度信号。可选地，加速度计可以是包括多个基本传感器的传感器系统。加速度计可以用于测量(例如，非侵入地)测量人的加速度，以生成加速度信号。In an implementation of the second aspect, the system includes an accelerometer for generating an acceleration signal. Optionally, the accelerometer may be a sensor system including a plurality of basic sensors. The accelerometer may be used to measure (e.g., non-invasively) a person's acceleration to generate an acceleration signal.

在所述第二方面的一种实现方式中，所述系统是人可穿戴的设备。In an implementation of the second aspect, the system is a wearable device.

该设备可以例如佩戴在人的手指、手腕、手和/或手臂上。该设备可以是例如可由人佩戴的手表(例如，智能手表)、戒指或衣服(例如，戒指、腕带、手镯等)。根据一种实现方式，所述系统的装置可以是第一设备的一部分或者可以是第一设备，并且上述系统的一个或多个可选传感器可以是第二设备的一部分。第一设备可以是便携式设备或用户端设备，例如，智能手机、平板电脑、笔记本电脑等；第一设备可以是固定设备，例如，台式电脑。第二设备可以是手表、智能手表、衣服等，第二设备可以是人可穿戴的。The device can be worn, for example, on a person's finger, wrist, hand and/or arm. The device can be, for example, a watch (e.g., a smart watch), a ring or a piece of clothing (e.g., a ring, a wristband, a bracelet, etc.) that can be worn by a person. According to one implementation, the apparatus of the system can be part of a first device or can be a first device, and one or more optional sensors of the above system can be part of a second device. The first device can be a portable device or a user-end device, such as a smart phone, a tablet computer, a laptop computer, etc.; the first device can be a fixed device, such as a desktop computer. The second device can be a watch, a smart watch, a piece of clothing, etc., and the second device can be wearable by a person.

以上对第一方面所述的装置的描述相应地对于第二方面所述的系统是有效的。上述第二方面所述的系统的描述相应地对于上述第一方面所述的装置是有效的。The above description of the device described in the first aspect is correspondingly valid for the system described in the second aspect. The above description of the system described in the second aspect is correspondingly valid for the above device described in the first aspect.

第二方面的系统及其实现方式和可选特征实现了与第一方面所述的装置及其相应的实现方式和相应的可选特征相同的优点。The system of the second aspect and its implementation and optional features achieve the same advantages as the device of the first aspect and its corresponding implementation and corresponding optional features.

为了实现根据本发明的第二方面所述的系统，如上所述，第二方面的部分或全部实现方式和可选特征可以相互组合。In order to implement the system according to the second aspect of the present invention, as described above, part or all of the implementations and optional features of the second aspect may be combined with each other.

本发明的第三方面提供了一种用于判断人是否睡着的方法。该方法包括接收心脏信号和/或温度信号。心脏信号与人的心率和/或人的血容量变化相关。所述温度信号与人的温度相关。此外，所述方法包括根据所述心脏信号的幅度和/或根据关于所述温度信号的幅度变化的变量来判断人是否睡着。换句话说，心脏信号与人的心功能相关。A third aspect of the present invention provides a method for determining whether a person is asleep. The method includes receiving a heart signal and/or a temperature signal. The heart signal is related to the person's heart rate and/or the change in the person's blood volume. The temperature signal is related to the person's temperature. In addition, the method includes determining whether the person is asleep based on the amplitude of the heart signal and/or based on a variable related to the amplitude change of the temperature signal. In other words, the heart signal is related to the person's heart function.

上述第一方面的装置的描述相应地对于第三方面的方法是有效的。The above description of the apparatus of the first aspect is correspondingly valid for the method of the third aspect.

在所述第三方面的一种实现方式中，所述心脏信号为光电体积描记(photoplethysmographic，PPG)信号。In an implementation of the third aspect, the cardiac signal is a photoplethysmographic (PPG) signal.

在所述第三方面的一种实现方式中，所述方法包括：通过计算后续时间窗口的心脏信号的平均幅度来计算或确定心脏信号的幅度。In an implementation of the third aspect, the method includes: calculating or determining the amplitude of the cardiac signal by calculating the average amplitude of the cardiac signal in subsequent time windows.

在所述第三方面的一种实现方式中，所述方法包括：在所述时间窗口中的一个时间窗口期间，所述心脏信号的幅度变化和/或所述心脏信号的峰值之间的时间距离变化在心功能的正常生理限制内的情况下，确定所述心脏信号在所述时间窗口期间的信号质量等于或大于所述信号质量的阈值。此外，所述方法可以包括仅针对所述心脏信号的信号质量等于或大于所述信号质量的所述阈值的时间窗口，计算所述心脏信号的所述平均幅度。In an implementation of the third aspect, the method includes: determining that the signal quality of the cardiac signal during one of the time windows is equal to or greater than the threshold value of the signal quality when the amplitude change of the cardiac signal and/or the time distance change between the peaks of the cardiac signal during one of the time windows is within the normal physiological limits of cardiac function. In addition, the method may include calculating the average amplitude of the cardiac signal only for time windows in which the signal quality of the cardiac signal is equal to or greater than the threshold value of the signal quality.

在所述第三方面的一种实现方式中，所述方法包括：在所述心脏信号的幅度增大到超过所述心脏信号的幅度的第一阈值的情况下，判断人睡着了。此外，所述方法可以包括：在所述心脏信号的幅度降低到低于所述心脏信号的幅度的第二阈值的情况下，判断人醒着。In an implementation of the third aspect, the method includes: when the amplitude of the heart signal increases to exceed a first threshold of the amplitude of the heart signal, judging that the person is asleep. In addition, the method may include: when the amplitude of the heart signal decreases to below a second threshold of the amplitude of the heart signal, judging that the person is awake.

在所述第三方面的一种实现方式中，与温度信号的幅度变化相关的变量指示温度信号的幅度变化、温度信号的幅度变化的频率和温度信号的幅度变化的速率中的至少一个。In an implementation of the third aspect, the variable related to the amplitude change of the temperature signal indicates at least one of the amplitude change of the temperature signal, the frequency of the amplitude change of the temperature signal, and the rate of the amplitude change of the temperature signal.

在所述第三方面的一种实现方式中，所述方法包括：根据所述温度信号，通过计算以下各项中的至少一项来计算与所述温度信号的幅度变化相关的所述变量：滑动时间窗口内的温度信号导数的至少两个绝对值之和；针对时间窗口的所述温度信号的傅里叶变换；在滑动时间窗口或固定阶跃时间窗口期间的所述温度信号的最大值和最小值。In an implementation of the third aspect, the method includes: based on the temperature signal, calculating the variable related to the amplitude change of the temperature signal by calculating at least one of the following items: the sum of at least two absolute values of the temperature signal derivative within a sliding time window; the Fourier transform of the temperature signal for the time window; the maximum and minimum values of the temperature signal during a sliding time window or a fixed step time window.

在所述第三方面的一种实现方式中，所述方法包括：在与所述温度信号的幅度变化相关的所述变量降低到低于所述变量的第一阈值的情况下，判断人睡着了。此外，所述方法可以包括：在与所述温度信号的幅度变化相关的变量增大到超过所述变量的第二阈值的情况下，判断人醒着。In an implementation of the third aspect, the method includes: when the variable related to the amplitude change of the temperature signal decreases to a value lower than a first threshold value of the variable, judging that the person is asleep. In addition, the method may include: when the variable related to the amplitude change of the temperature signal increases to a value higher than a second threshold value of the variable, judging that the person is awake.

在所述第三方面的一种实现方式中，所述方法包括：根据所述心脏信号的所述幅度和与所述温度信号的幅度变化相关的所述变量中的至少一个，使用经训练的机器学习模型来判断人是否睡着。根据包括多个数据集的训练数据，训练所述经训练的机器学习模型。多个数据集中的每个数据集包括睡眠状态变量，指示人是否在相应时间睡着了，与以下内容相关联：在所述相应时间的所述心脏信号的所述幅度和与所述温度信号的幅度变化相关的所述变量的中至少一个。In an implementation of the third aspect, the method includes: using a trained machine learning model to determine whether a person is asleep based on at least one of the amplitude of the heart signal and the variable associated with the amplitude change of the temperature signal. The trained machine learning model is trained based on training data including multiple data sets. Each of the multiple data sets includes a sleep state variable indicating whether a person is asleep at a corresponding time, which is associated with: at least one of the amplitude of the heart signal and the variable associated with the amplitude change of the temperature signal at the corresponding time.

在所述第三方面的一种实现方式中，所述方法包括：接收与人的加速度相关的加速度信号。此外，该方法可以包括：根据加速度信号计算人的活动程度。此外，该方法可以包括：除心脏信号的幅度和与温度信号的幅度变化相关的变量中的至少一个之外，根据活动程度判断人是否睡着。In an implementation of the third aspect, the method includes: receiving an acceleration signal related to the acceleration of a person. In addition, the method may include: calculating the activity level of the person based on the acceleration signal. In addition, the method may include: determining whether the person is asleep based on the activity level in addition to at least one of the amplitude of the heart signal and the variable related to the amplitude change of the temperature signal.

在所述第三方面的一种实现方式中，所述方法包括：在所述活动程度降低到低于所述活动程度的第一阈值的情况下，判断人睡着了。此外，所述方法可以包括：在所述活动程度增大到超过所述活动程度的第二阈值的情况下，判断人醒着。In an implementation of the third aspect, the method includes: when the activity level decreases to a level below a first threshold of the activity level, determining that the person is asleep. In addition, the method may include: when the activity level increases to a level above a second threshold of the activity level, determining that the person is awake.

在所述第三方面的一种实现方式中，所述方法包括：接收与人的温度相关的温度信号。此外，该方法可以包括：除心脏信号的幅度和与温度信号的幅度变化相关的变量中的至少一个之外，根据温度信号的幅度判断人是否睡着。In an implementation of the third aspect, the method includes: receiving a temperature signal related to the temperature of the person. In addition, the method may include: determining whether the person is asleep based on the amplitude of the temperature signal in addition to at least one of the amplitude of the heart signal and a variable related to the amplitude change of the temperature signal.

在所述第三方面的一种实现方式中，所述方法包括：在所述温度信号的幅度增大到超过所述温度信号的幅度的第一阈值的情况下，判断人睡着了。此外，所述方法可以包括：在所述温度信号的幅度降低到低于所述温度信号的幅度的第二阈值的情况下，判断人醒着。In an implementation of the third aspect, the method includes: when the amplitude of the temperature signal increases to exceed a first threshold of the amplitude of the temperature signal, judging that the person is asleep. In addition, the method may include: when the amplitude of the temperature signal decreases to below a second threshold of the amplitude of the temperature signal, judging that the person is awake.

在所述第三方面的一种实现方式中，所述方法包括：根据所述心脏信号的所述幅度和与所述温度信号的幅度变化相关的所述变量中的至少一个，并且根据所述温度信号的所述活动程度和所述幅度中的至少一个，使用经训练的机器学习模型来判断人是否睡着。根据包括多个数据集的训练数据，训练所述经训练的机器学习模型。多个数据集中的每个数据集包括睡眠状态变量，指示人是否在相应时间睡着了，与以下内容相关联：在所述相应时间的所述心脏信号的所述幅度和与所述温度信号的幅度变化相关的所述变量的中至少一个，In an implementation of the third aspect, the method includes: judging whether a person is asleep based on the amplitude of the heart signal and at least one of the variables associated with the amplitude change of the temperature signal, and based on the activity level and at least one of the amplitude of the temperature signal, using a trained machine learning model. The trained machine learning model is trained based on training data including multiple data sets. Each of the multiple data sets includes a sleep state variable indicating whether a person is asleep at a corresponding time, and is associated with: the amplitude of the heart signal at the corresponding time and at least one of the variables associated with the amplitude change of the temperature signal,

第三方面的方法及其实现方式和可选特征实现了与第一方面所述的装置及其相应的实现方式和相应的可选特征相同的优点。The method of the third aspect and its implementation and optional features achieve the same advantages as the device of the first aspect and its corresponding implementation and corresponding optional features.

为了实现根据本发明的第三方面所述的方法，上述第三方面的部分或全部所述实现方式和可选特征可以相互结合。In order to implement the method according to the third aspect of the present invention, part or all of the implementation modes and optional features of the third aspect may be combined with each other.

本发明的第四方面提供了一种计算机程序，包括用于在处理器上实现时执行根据第三方面或其任何实现方式所述的方法的程序代码。A fourth aspect of the present invention provides a computer program comprising program code for executing the method according to the third aspect or any implementation thereof when implemented on a processor.

本发明的第五方面提供了一种计算机程序，包括用于执行根据第三方面或其任何实现方式所述的方法的程序代码。A fifth aspect of the present invention provides a computer program, comprising program code for executing the method according to the third aspect or any implementation manner thereof.

本发明的第六方面提供了一种计算机，包括存储器和处理器，用于存储和执行程序代码，以执行根据第三方面或其任何实现方式所述的方法。A sixth aspect of the present invention provides a computer, comprising a memory and a processor, for storing and executing program code to perform the method according to the third aspect or any implementation thereof.

本发明的第七方面提供了一种非瞬时性存储介质，该非瞬时性存储介质存储可执行程序代码，可执行程序代码在由处理器执行时，促使执行根据第三方面或其任何实现方式所述的方法。A seventh aspect of the present invention provides a non-transitory storage medium, which stores executable program code, and when the executable program code is executed by a processor, it causes the execution of the method described in the third aspect or any implementation thereof.

本发明的第八方面提供了一种计算机可读存储介质，所述计算机可读存储介质存储可执行程序代码，所述可执行程序代码在由处理器执行时，促使执行根据所述第三方面或其任何实现方式所述的方法。An eighth aspect of the present invention provides a computer-readable storage medium, which stores executable program code. When the executable program code is executed by a processor, it prompts the execution of the method described in the third aspect or any implementation thereof.

第四方面的计算机程序、第五方面的计算机程序、第六方面的计算机、第七方面的非瞬时性存储介质和第八方面的计算机可读存储介质都实现了与第一方面所述的装置及其相应的实现方式和相应的可选特征相同的优点。The computer program of the fourth aspect, the computer program of the fifth aspect, the computer of the sixth aspect, the non-transitory storage medium of the seventh aspect, and the computer-readable storage medium of the eighth aspect all achieve the same advantages as the device described in the first aspect and its corresponding implementation methods and corresponding optional features.

需要注意的是，本申请中描述的所有设备、元件、单元和模块可以在软件或硬件元件或其任何类型的组合中实现。本申请中描述的各种实体所执行的所有步骤以及所描述的各种实体要执行的功能均意在指相应实体用于执行相应步骤和功能。虽然在以下具体实施例的描述中，外部实体执行的具体功能或步骤没有在执行具体步骤或功能的实体的具体详述元件的描述中反映，但是技术人员应清楚，这些方法和功能可以通过相应的硬件或软件元件或其任何组合实现。It should be noted that all devices, elements, units and modules described in this application can be implemented in software or hardware elements or any type of combination thereof. All steps performed by the various entities described in this application and the functions to be performed by the various entities described are intended to refer to the corresponding entities for performing the corresponding steps and functions. Although in the description of the following specific embodiments, the specific functions or steps performed by the external entity are not reflected in the description of the specific detailed elements of the entity performing the specific steps or functions, it should be clear to the technician that these methods and functions can be implemented by corresponding hardware or software elements or any combination thereof.

附图说明BRIEF DESCRIPTION OF THE DRAWINGS

结合所附附图，具体实施例的以下描述阐述本发明的上述各个方面及实现方式，在附图中：The following description of specific embodiments, in conjunction with the accompanying drawings, illustrates the above-mentioned various aspects and implementations of the present invention, in which:

图1示出了根据本发明的实施例的用于判断人是否睡着的装置(参见图1A)和根据本发明的实施例的系统(参见图1B)；FIG1 shows a device for determining whether a person is asleep according to an embodiment of the present invention (see FIG1A ) and a system according to an embodiment of the present invention (see FIG1B );

图2示出了根据本发明的实施例的用于判断人是否睡着的装置(参见图2A)和根据本发明的实施例的系统(参见图2B)；FIG2 shows a device for determining whether a person is asleep according to an embodiment of the present invention (see FIG2A ) and a system according to an embodiment of the present invention (see FIG2B );

图3示出了根据本发明的两个实施例的用于判断人是否睡着的方法；FIG3 shows a method for determining whether a person is asleep according to two embodiments of the present invention;

图4示出了根据本发明的实施例的用于描述如何判断人睡着了的心脏信号的幅度的示例；FIG4 shows an example of how to determine the amplitude of a heart signal when a person falls asleep according to an embodiment of the present invention;

图5示出了根据本发明的实施例的用于描述如何判断人醒着的心脏信号的幅度的示例；FIG5 shows an example for describing how to determine the amplitude of a heart signal of a person who is awake according to an embodiment of the present invention;

图6示出了根据本发明的实施例的在人睡着了的情况下的温度信号和温度信号的处理的示例；FIG6 shows an example of a temperature signal and processing of the temperature signal in a case where a person falls asleep according to an embodiment of the present invention;

图7示出了根据本发明的实施例的在人醒着的情况下的温度信号和温度信号的处理的示例；FIG. 7 shows an example of a temperature signal and processing of the temperature signal when a person is awake according to an embodiment of the present invention;

图8示出了图4和图6的曲线图以及显示人的活动程度的曲线图和显示睡眠状态变量的曲线图，该睡眠状态变量指示人是否睡着了。FIG. 8 shows the graphs of FIGS. 4 and 6 together with a graph showing the activity level of a person and a graph showing a sleep state variable indicating whether the person is asleep.

在附图中，相应的元件标有相同的附图标记。In the figures, corresponding elements are provided with the same reference numerals.

具体实施方式Detailed ways

图1示出了根据本发明的实施例的用于判断人是否睡着的装置(参见图1A)和根据本发明的实施例的系统(参见图1B)。图1A的装置是根据上述第一方面的装置的示例。图1B的系统是根据上述第二方面的系统的示例。FIG1 shows a device for determining whether a person is asleep according to an embodiment of the present invention (see FIG1A ) and a system according to an embodiment of the present invention (see FIG1B ). The device of FIG1A is an example of a device according to the first aspect described above. The system of FIG1B is an example of a system according to the second aspect described above.

如图1A所示，根据一种替代方案，用于判断人是否睡着的装置1可以用于接收指示心脏运行的信号S1，其中，指示心脏运行的信号S1与人的心功能相关。指示心脏运行的信号S1与在测量位置的人的心率和(人的)血容量变化相关。装置1可以用于根据指示心脏运行的信号S1的幅度判断人是否睡着。可选地，指示心脏运行的信号S1是光电体积描记(photoplethysmographic，PPG)信号。指示心脏运行的信号S1可以是任何其他已知的心脏信号。装置1可以用于根据指示心脏运行的信号S1计算心脏信号的幅度。As shown in FIG. 1A , according to an alternative, the device 1 for determining whether a person is asleep can be used to receive a signal S1 indicating heart operation, wherein the signal S1 indicating heart operation is related to the heart function of the person. The signal S1 indicating heart operation is related to the heart rate of the person and the change in blood volume (of the person) at the measurement position. The device 1 can be used to determine whether a person is asleep based on the amplitude of the signal S1 indicating heart operation. Optionally, the signal S1 indicating heart operation is a photoplethysmographic (PPG) signal. The signal S1 indicating heart operation can be any other known heart signal. The device 1 can be used to calculate the amplitude of the heart signal based on the signal S1 indicating heart operation.

根据另一种替代方案，用于判断人是否睡着的装置1可以用于接收温度信号S2，其中，温度信号S2与人的温度相关。装置1可以用于根据关于温度信号S2的幅度变化的变量来判断人是否睡着。与温度信号S2的幅度变化相关的变量可以指示温度信号S2的幅度变化、温度信号S2的幅度变化的频率、温度信号S2的幅度变化的速率中的至少一个。装置1可以用于根据温度信号S2计算与温度信号S2的幅度变化相关的变量。为了计算变量，装置1可以用于计算滑动时间窗口内的温度信号导数的至少两个绝对值之和。附加地或可替代地，装置1可以用于计算针对时间窗口的温度信号S2的傅里叶变换。附加地或可替代地，装置1可以用于计算滑动时间窗口或固定阶跃时间窗口内的温度信号S2的最大值和最小值。According to another alternative, the device 1 for determining whether a person is asleep can be used to receive a temperature signal S2, wherein the temperature signal S2 is related to the temperature of the person. The device 1 can be used to determine whether a person is asleep based on a variable regarding the amplitude change of the temperature signal S2. The variable related to the amplitude change of the temperature signal S2 can indicate at least one of the amplitude change of the temperature signal S2, the frequency of the amplitude change of the temperature signal S2, and the rate of the amplitude change of the temperature signal S2. The device 1 can be used to calculate the variable related to the amplitude change of the temperature signal S2 based on the temperature signal S2. In order to calculate the variable, the device 1 can be used to calculate the sum of at least two absolute values of the derivative of the temperature signal within a sliding time window. Additionally or alternatively, the device 1 can be used to calculate the Fourier transform of the temperature signal S2 for the time window. Additionally or alternatively, the device 1 can be used to calculate the maximum and minimum values of the temperature signal S2 within a sliding time window or a fixed step time window.

根据另一替代方案，用于判断人是否睡着的装置1可以用于接收指示心脏运行的信号S1和温度信号S2，并且根据指示心脏运行的信号S1的幅度和与温度信号S2的幅度变化相关的变量判断人是否睡着。According to another alternative, the device 1 for determining whether a person is asleep can be used to receive a signal S1 indicating heart operation and a temperature signal S2, and determine whether a person is asleep based on the amplitude of the signal S1 indicating heart operation and a variable related to the amplitude change of the temperature signal S2.

该装置可以是或者可以包括处理电路(图1中未示出)。处理电路用于执行、进行或启动本文所述的用于判断人是否睡着的装置的各种操作。处理电路可以用于执行、进行或启动根据第一方面的装置的各种操作，如上所述。处理电路可以包括硬件和/或可以由软件控制。硬件可以包括模拟电路或数字电路，或既包括模拟电路又包括数字电路。数字电路可以包括专用集成电路(application-specific integrated circuit，ASIC)、现场可编程阵列(field-programmable array，FPGA)、数字信号处理器(digital signal processor，DSP)或多用途处理器等组件。The device may be or may include a processing circuit (not shown in FIG. 1 ). The processing circuit is used to execute, perform or start various operations of the device for determining whether a person is asleep as described herein. The processing circuit may be used to execute, perform or start various operations of the device according to the first aspect, as described above. The processing circuit may include hardware and/or may be controlled by software. The hardware may include analog circuits or digital circuits, or both analog circuits and digital circuits. The digital circuit may include components such as an application-specific integrated circuit (ASIC), a field-programmable array (FPGA), a digital signal processor (DSP) or a multi-purpose processor.

该装置还可以包括存储器电路(与处理电路相关联，也可以是处理电路的一部分)，存储器电路存储可由处理电路可选地在软件(图1中未示出)的控制下执行的一条或多条指令。例如，存储器电路可以包括存储可执行软件代码的非瞬时性存储介质，该可执行软件代码在由处理电路执行时，使处理电路执行、进行或启动本文所述的操作或方法。非瞬时性存储介质可以存储可执行软件代码，可执行软件代码在由处理电路执行时，使处理电路执行、进行或启动根据本文所述的第三方面的方法。The apparatus may also include a memory circuit (associated with the processing circuit or may be part of the processing circuit), the memory circuit storing one or more instructions that may be executed by the processing circuit, optionally under the control of software (not shown in FIG. 1). For example, the memory circuit may include a non-transitory storage medium storing executable software code that, when executed by the processing circuit, causes the processing circuit to perform, conduct or initiate the operations or methods described herein. The non-transitory storage medium may store executable software code that, when executed by the processing circuit, causes the processing circuit to perform, conduct or initiate the method according to the third aspect described herein.

关于图1A的装置的进一步细节，参考根据本发明的第一方面的装置的上述描述。For further details of the apparatus of FIG. 1A , reference is made to the above description of the apparatus according to the first aspect of the invention.

如图1B所示，系统2包括图1A的装置1，如上所述。根据一种替代方案，在装置1使用心脏信号S1来判断人的睡眠状态的情况下，系统2还可以包括心功能传感器3，用于生成指示心脏运行的信号S1。心功能传感器3用于向装置1提供指示心脏运行的信号S1，如图1B所示。心功能传感器3可以是或者可以包括光电体积描记(photoplethysmographic，PPG)传感器。心功能传感器3可以以不同的方式实现。根据另一替代方案，在装置1使用温度信号S2来判断人的睡眠状态的情况下，系统2还可以包括温度传感器4，用于生成温度信号S2。温度传感器4用于向装置1提供温度信号S2，如图1B所示。根据另一替代方案，在装置1使用指示心脏运行的信号S1和温度信号S2来判断人的睡眠状态的情况下，系统2还可以包括心功能传感器3和温度传感器4。As shown in FIG. 1B , system 2 includes the device 1 of FIG. 1A , as described above. According to an alternative, in the case where the device 1 uses the heart signal S1 to determine the sleep state of a person, the system 2 may further include a heart function sensor 3 for generating a signal S1 indicating the operation of the heart. The heart function sensor 3 is used to provide the device 1 with a signal S1 indicating the operation of the heart, as shown in FIG. 1B . The heart function sensor 3 may be or may include a photoplethysmographic (PPG) sensor. The heart function sensor 3 may be implemented in different ways. According to another alternative, in the case where the device 1 uses the temperature signal S2 to determine the sleep state of a person, the system 2 may further include a temperature sensor 4 for generating a temperature signal S2. The temperature sensor 4 is used to provide the temperature signal S2 to the device 1, as shown in FIG. 1B . According to another alternative, in the case where the device 1 uses the signal S1 indicating the operation of the heart and the temperature signal S2 to determine the sleep state of a person, the system 2 may further include a heart function sensor 3 and a temperature sensor 4.

可选地，所述系统2是人可穿戴的设备。设备2可以例如佩戴在人的手指、手腕、手和/或手臂上。设备2可以是例如可由人佩戴的手表(例如，智能手表)、戒指或衣服(例如，戒指、腕带、手镯等)。根据一实施例，系统2的装置1可以是第一设备的一部分或者可以是第一设备，并且上述系统2的一个或多个传感器3和4可以是第二设备的一部分。第一设备可以是便携式设备或用户端设备，例如，智能手机、平板电脑、笔记本电脑等；第一设备可以是固定设备，例如，台式电脑。第二设备可以是人可穿戴的设备，例如，手表、智能手表、衣服等。Optionally, the system 2 is a wearable device for a person. The device 2 can be worn, for example, on a person's finger, wrist, hand and/or arm. The device 2 can be, for example, a watch (e.g., a smart watch), a ring or clothing (e.g., a ring, a wristband, a bracelet, etc.) that can be worn by a person. According to one embodiment, the apparatus 1 of the system 2 can be part of a first device or can be the first device, and one or more sensors 3 and 4 of the above-mentioned system 2 can be part of a second device. The first device can be a portable device or a user-end device, such as a smart phone, a tablet computer, a laptop computer, etc.; the first device can be a fixed device, such as a desktop computer. The second device can be a wearable device for a person, such as a watch, a smart watch, clothing, etc.

关于图1B的系统的进一步细节，参考根据本发明的第二方面的系统的上述描述。For further details of the system of FIG. 1B , reference is made to the above description of the system according to the second aspect of the invention.

图2示出了根据本发明的实施例的用于判断人是否睡着的装置(参见图2A)和根据本发明的实施例的系统(参见图2B)。图2A的装置1对应于具有附加可选特征的图1A的装置1，图2B的系统2对应于具有附加可选特征的图1B的系统2。因此，图1A的装置1的上述描述对图2A的装置1是有效的，图1B的系统2的上述描述对图2B的系统2是有效的。在下文中，主要描述关于图1的附加可选特性。FIG. 2 shows an apparatus for determining whether a person is asleep according to an embodiment of the present invention (see FIG. 2A ) and a system according to an embodiment of the present invention (see FIG. 2B ). The apparatus 1 of FIG. 2A corresponds to the apparatus 1 of FIG. 1A with additional optional features, and the system 2 of FIG. 2B corresponds to the system 2 of FIG. 1B with additional optional features. Therefore, the above description of the apparatus 1 of FIG. 1A is valid for the apparatus 1 of FIG. 2A , and the above description of the system 2 of FIG. 1B is valid for the system 2 of FIG. 2B . In the following, the additional optional features of FIG. 1 are mainly described.

如图2A所示，除指示心脏运行的信号S1和/或温度信号S2之外，装置1还可以使用与人的加速度相关的加速度信号S3，用于判断人是否睡着。此时，装置1可以用于接收加速度信号S3，并且根据加速度信号S3计算人的活动程度。此外，装置1可以用于除指示心脏运行的信号S1的幅度和与温度信号S2的幅度变化相关的变量中的至少一个之外，根据活动程度判断人是否睡着。As shown in FIG2A , in addition to the signal S1 indicating the operation of the heart and/or the temperature signal S2, the device 1 can also use an acceleration signal S3 related to the acceleration of the person to determine whether the person is asleep. In this case, the device 1 can be used to receive the acceleration signal S3 and calculate the activity level of the person based on the acceleration signal S3. In addition, the device 1 can be used to determine whether the person is asleep based on the activity level in addition to at least one of the amplitude of the signal S1 indicating the operation of the heart and the variable related to the amplitude change of the temperature signal S2.

附加地或替代地，装置1可以用于接收温度信号S2，并且装置1可以用于除指示心脏运行的信号S1的幅度和与温度信号S2的幅度变化相关的变量中的至少一个之外，根据温度信号S2的幅度判断人是否睡着。Additionally or alternatively, the device 1 can be used to receive a temperature signal S2, and the device 1 can be used to determine whether a person is asleep based on the amplitude of the temperature signal S2 in addition to the amplitude of the signal S1 indicating heart operation and at least one of the variables related to the amplitude change of the temperature signal S2.

因此，根据替代方案，该装置可以用于除指示心脏运行的信号S1的幅度和与温度信号S2的幅度变化相关的变量中的至少一个之外，根据(使用加速度信号S3计算的)活动程度判断人是否睡着。根据另一替代方案，装置1可以用于除指示心脏运行的信号S1的幅度和与温度信号S2的幅度变化相关的变量中的至少一个之外，根据温度信号S2的幅度判断人是否睡着。根据另一替代方案，装置1可以用于除指示心脏运行的信号S1的幅度和与温度信号S2的幅度变化相关的变量中的至少一个之外，根据活动程度和温度信号S2的幅度来判断人是否睡着。Therefore, according to an alternative, the device can be used to determine whether a person is asleep based on the activity level (calculated using the acceleration signal S3), in addition to the amplitude of the signal S1 indicating the operation of the heart and at least one of the variables related to the amplitude change of the temperature signal S2. According to another alternative, the device 1 can be used to determine whether a person is asleep based on the amplitude of the temperature signal S2, in addition to the amplitude of the signal S1 indicating the operation of the heart and at least one of the variables related to the amplitude change of the temperature signal S2. According to another alternative, the device 1 can be used to determine whether a person is asleep based on the activity level and the amplitude of the temperature signal S2, in addition to the amplitude of the signal S1 indicating the operation of the heart and at least one of the variables related to the amplitude change of the temperature signal S2.

换句话说，图2A的装置1可以用于根据指示心脏运行的信号S1的幅度和与温度信号S2的幅度变化相关的变量中的至少一个，并且根据活动程度和温度信号S2的幅度中的至少一个，判断人是否睡着。In other words, the device 1 of Figure 2A can be used to determine whether a person is asleep based on at least one of the amplitude of the signal S1 indicating the operation of the heart and the variable related to the amplitude change of the temperature signal S2, and based on at least one of the activity level and the amplitude of the temperature signal S2.

关于图2A的装置1的进一步细节，参考图1A的装置1的上述描述以及根据本发明的第一方面的装置的上述描述。For further details of the apparatus 1 of FIG. 2A , reference is made to the above description of the apparatus 1 of FIG. 1A and to the above description of the apparatus according to the first aspect of the present invention.

如图2B所示，系统2包括图2A的装置1，如上所述。在装置1使用加速度信号S3(除指示心脏运行的信号S1和/或温度信号S2之外)来判断人的睡眠状态的情况下，系统2还可以包括加速度计5，用于生成加速度信号S3。加速度计5用于向装置1提供加速度信号S3，如图2B所示。在装置1用于除指示心脏运行的信号S1的幅度和/或与温度信号S2的幅度变化相关的变量之外，根据温度信号S2的幅度确定系统2包括温度传感器S2。As shown in FIG2B , the system 2 includes the device 1 of FIG2A , as described above. In the case where the device 1 uses the acceleration signal S3 (in addition to the signal S1 indicating the operation of the heart and/or the temperature signal S2) to determine the sleep state of a person, the system 2 may also include an accelerometer 5 for generating the acceleration signal S3. The accelerometer 5 is used to provide the acceleration signal S3 to the device 1, as shown in FIG2B . In the case where the device 1 is used to determine the system 2 includes a temperature sensor S2 based on the amplitude of the temperature signal S2 in addition to the amplitude of the signal S1 indicating the operation of the heart and/or a variable related to the amplitude change of the temperature signal S2.

关于图2B的系统的进一步细节，参考图1B的系统的上述描述以及根据本发明的第二方面的系统的上述描述。For further details of the system of FIG. 2B , reference is made to the above description of the system of FIG. 1B and to the above description of the system according to the second aspect of the invention.

图3示出了根据本发明的两个实施例的用于判断人是否睡着的方法。图3A的方法和图3B的方法分别是根据本发明的第三方面的方法的示例，如上所述。图1和图2的装置1可以用于执行图3A和图3B的用于判断人是否睡着的方法。图1和图2的上述描述可以相应地对于图3A和图3B的方法是有效的。FIG3 shows a method for determining whether a person is asleep according to two embodiments of the present invention. The method of FIG3A and the method of FIG3B are examples of methods according to the third aspect of the present invention, as described above. The apparatus 1 of FIG1 and FIG2 can be used to perform the method for determining whether a person is asleep of FIG3A and FIG3B. The above description of FIG1 and FIG2 can be valid for the method of FIG3A and FIG3B accordingly.

在图3A的用于判断人是否睡着的方法的步骤30中，可以接收或获得心脏信号和/或温度信号。心脏信号与人的心率和在人的测量位置(例如，手指、手腕、手和/或手臂)的血容量变化相关。所述温度信号与人的温度相关。此外，在图3A的方法的步骤30中，可以可选地接收或获得与人的加速度相关的加速度信号。在步骤30之后的步骤31中，该方法包括根据心脏信号的幅度和/或关于温度信号的幅度变化的变量，并且可选地根据温度信号的幅度和/或基于可选的加速度信号计算的人的活动程度，判断人是否睡着。In step 30 of the method for determining whether a person is asleep in FIG. 3A , a heart signal and/or a temperature signal may be received or obtained. The heart signal is related to the heart rate of the person and the change in blood volume at a measurement location of the person (e.g., a finger, wrist, hand, and/or arm). The temperature signal is related to the temperature of the person. In addition, in step 30 of the method of FIG. 3A , an acceleration signal related to the acceleration of the person may be optionally received or obtained. In step 31 after step 30, the method includes determining whether the person is asleep based on the amplitude of the heart signal and/or a variable regarding the amplitude change of the temperature signal, and optionally based on the amplitude of the temperature signal and/or the activity level of the person calculated based on an optional acceleration signal.

图3B的方法的步骤300对应于图3A的方法的步骤30。换句话说，在步骤300中，可以接收或获得传感器测量。因此，图3A的方法的步骤30的上述描述对于图3B的方法的步骤300是有效的。在步骤300之后的步骤301中，在步骤300中接收或获得心脏信号的情况下，可以根据心脏信号计算心脏信号的幅度。附加地或可替代地，在步骤300之后的步骤301中，在步骤300中接收或获得温度信号的情况下，可以根据温度信号来计算关于温度信号的幅度变化的变量。如上所述，与温度信号的幅度变化相关的变量可以表示温度信号的幅度变化、温度信号的幅度变化的频率、温度信号的幅度变化的速率中的至少一个。例如，为了根据温度信号计算与温度信号的幅度变化相关的变量，可以计算滑动时间窗口内的温度信号导数的至少两个绝对值之和。此外或替代地，可以计算针对时间窗口的温度信号的傅里叶变换。此外或替代地，可以计算滑动时间窗口或固定阶跃时间窗口内的温度信号的最大值和最小值。可选地，在步骤300之后的步骤301中，在步骤300中接收或获得加速度信号的情况下，可以根据加速度信号计算人的活动程度。因此，在步骤301中，可以执行基于传感器测量的特征提取。Step 300 of the method of FIG. 3B corresponds to step 30 of the method of FIG. 3A . In other words, in step 300, a sensor measurement may be received or obtained. Therefore, the above description of step 30 of the method of FIG. 3A is valid for step 300 of the method of FIG. 3B . In step 301 after step 300, in the case where a cardiac signal is received or obtained in step 300, the amplitude of the cardiac signal may be calculated from the cardiac signal. Additionally or alternatively, in step 301 after step 300, in the case where a temperature signal is received or obtained in step 300, a variable regarding the amplitude change of the temperature signal may be calculated from the temperature signal. As described above, the variable associated with the amplitude change of the temperature signal may represent at least one of the amplitude change of the temperature signal, the frequency of the amplitude change of the temperature signal, and the rate of the amplitude change of the temperature signal. For example, in order to calculate the variable associated with the amplitude change of the temperature signal from the temperature signal, the sum of at least two absolute values of the derivative of the temperature signal within a sliding time window may be calculated. Additionally or alternatively, a Fourier transform of the temperature signal for a time window may be calculated. Additionally or alternatively, the maximum and minimum values of the temperature signal within a sliding time window or a fixed step time window may be calculated. Optionally, in step 301 after step 300, in the case where an acceleration signal is received or obtained in step 300, the activity level of the person may be calculated according to the acceleration signal. Therefore, in step 301, feature extraction based on sensor measurement may be performed.

在步骤301之后的步骤302中，可以根据心脏信号的幅度和关于温度信号的幅度变化的变量中的至少一个，并且可选地根据人的活动程度和温度信号的幅度中的至少一个，使用经训练的机器学习模型来判断人是否睡着。换句话说，在步骤302中，可以将心脏信号的幅度和/或与温度信号的幅度变化以及可选地人的活动程度相关的变量和/或温度信号的幅度作为输入提供给经训练的机器学习模型(即，输入到机器学习模型)，以便判断人是否睡着。为此，经训练的机器学习模型可以根据上述输入计算或提供睡眠状态变量，作为输出，其中，睡眠状态变量指示人是否睡着。换句话说，经训练的机器学习模型可以根据在相应时间的上述输入在相应时间计算睡眠状态变量，指示人在相应时间是否睡着。In step 302 after step 301, a trained machine learning model may be used to determine whether a person is asleep based on at least one of the amplitude of the heart signal and the variable related to the amplitude change of the temperature signal, and optionally based on at least one of the person's activity level and the amplitude of the temperature signal. In other words, in step 302, the amplitude of the heart signal and/or the variable related to the amplitude change of the temperature signal and optionally the person's activity level and/or the amplitude of the temperature signal may be provided as input to a trained machine learning model (i.e., input to the machine learning model) to determine whether the person is asleep. To this end, the trained machine learning model may calculate or provide a sleep state variable as an output based on the above input, wherein the sleep state variable indicates whether the person is asleep. In other words, the trained machine learning model may calculate a sleep state variable at a corresponding time based on the above input at the corresponding time, indicating whether the person is asleep at the corresponding time.

根据包括多个数据集的训练数据训练经训练的机器学习模型，其中，多个数据集中的每个数据集包括睡眠状态变量，指示人是否在相应时间睡着了，与相应时间的上述输入相关联。因此，多个数据集中的每个数据集可以包括相应时间的睡眠状态变量，与在相应时间的心脏信号的幅度和与温度信号的幅度变化相关的变量的中至少一个，至少一个相关联，并且可选地与在相应时间的人的活动程度和温度信号的幅度中的至少一个相关联。每个数据集的睡眠状态变量是基本事实。The trained machine learning model is trained based on training data including multiple data sets, wherein each of the multiple data sets includes a sleep state variable indicating whether the person is asleep at the corresponding time, associated with the above-mentioned input at the corresponding time. Therefore, each of the multiple data sets may include a sleep state variable at the corresponding time, associated with at least one of the amplitude of the heart signal at the corresponding time and a variable related to the amplitude change of the temperature signal, and optionally associated with at least one of the activity level of the person and the amplitude of the temperature signal at the corresponding time. The sleep state variable of each data set is the basic fact.

睡眠状态变量可以是二元变量，其中，二元变量的一个值指示睡眠状态(人睡着了)，二元变量的另一个值指示清醒状态(人没有睡着，但醒着)。或者，睡眠状态变量可以是百分比值，其中，高于或低于百分比阈值的百分比值指示睡眠状态，分别低于或高于百分比阈值的百分比值指示清醒状态。睡眠状态变量可以以不同的方式实现。The sleep state variable may be a binary variable, wherein one value of the binary variable indicates a sleep state (the person is asleep) and another value of the binary variable indicates a wake state (the person is not asleep but awake). Alternatively, the sleep state variable may be a percentage value, wherein a percentage value above or below a percentage threshold indicates a sleep state and a percentage value below or above the percentage threshold, respectively, indicates a wake state. The sleep state variable may be implemented in different ways.

图3B的方法中使用的机器学习模型可以是任何已知的机器学习模型，例如，决策树模型、随机森林模型、神经网络模型、深度神经网络模型或隐马尔可夫模型等。在以下描述中，仅通过示例的方式假设在图3B的方法的步骤302中使用的经训练的机器学习模型是经训练的决策树模型。关于决策树模型的描述相应地对于任何其他机器学习模型是有效的，即在使用另一个机器学习模型的情况下。The machine learning model used in the method of FIG3B can be any known machine learning model, for example, a decision tree model, a random forest model, a neural network model, a deep neural network model, or a hidden Markov model, etc. In the following description, it is assumed by way of example only that the trained machine learning model used in step 302 of the method of FIG3B is a trained decision tree model. The description of the decision tree model is correspondingly valid for any other machine learning model, that is, when another machine learning model is used.

在下表中，示出了在步骤300中可以接收或获得的信号以及用于判断人是否睡着的经训练的机器学习模型(可从上述信号中获得)的输入。此外，还示出了每个输入在睡眠开始时(即人睡着了时)和睡眠结束时(即人醒着时)的行为。In the table below, the signals that can be received or obtained in step 300 and the inputs of the trained machine learning model (which can be obtained from the above signals) for determining whether a person is asleep are shown. In addition, the behavior of each input at the beginning of sleep (i.e., when the person is asleep) and the end of sleep (i.e., when the person is awake) is also shown.

信号Signal输入enter睡眠开始时的行为Behavior at sleep onset睡眠结束时的行为Behavior at the end of sleep心脏信号Heart Signals心脏信号的幅度Amplitude of the heart signal增大Increase降低reduce温度信号Temperature signal关于温度信号的幅度变化的变量Variables regarding the amplitude change of the temperature signal降低reduce增大Increase温度信号Temperature signal温度信号的幅度Amplitude of temperature signal增大Increase降低reduce加速度信号Acceleration signal人的活动程度The level of human activity降低reduce增大Increase

关于上表，心脏信号的幅度的增大可以指示人睡着了，而心脏信号的幅度的降低可以指示人醒着。例如，在心脏信号的幅度增大到超过心脏信号的幅度的第一阈值的情况下，可以判断人睡着了。在心脏信号的幅度降低到低于心脏信号的幅度的第二阈值的情况下，可以判断人醒着。上述第一阈值可以等于或大于上述第二阈值。With respect to the above table, an increase in the amplitude of the heart signal may indicate that the person is asleep, while a decrease in the amplitude of the heart signal may indicate that the person is awake. For example, in the case where the amplitude of the heart signal increases to exceed a first threshold of the amplitude of the heart signal, it may be determined that the person is asleep. In the case where the amplitude of the heart signal decreases to below a second threshold of the amplitude of the heart signal, it may be determined that the person is awake. The above-mentioned first threshold may be equal to or greater than the above-mentioned second threshold.

与温度信号的幅度变化相关的变量的降低可以指示人睡着了，而该变量的增大可以指示人醒着。例如，可以判断在变量降低到低于变量的第一阈值的情况下，人睡着了。如果变量增大到超过变量的第二阈值，则可以判断人醒着。上述第一阈值可以等于或小于上述第二阈值。A decrease in a variable associated with a change in the amplitude of the temperature signal may indicate that the person is asleep, while an increase in the variable may indicate that the person is awake. For example, it may be determined that the person is asleep when the variable decreases below a first threshold of the variable. If the variable increases to exceed a second threshold of the variable, it may be determined that the person is awake. The first threshold may be equal to or less than the second threshold.

温度信号的幅度增大可以指示人睡着了，温度信号的幅度降低可以指示人醒着。例如，如果温度信号的幅度增大到超过温度信号幅度的第一阈值，则可以判断人睡着了。如果温度信号的幅度降低到低于温度信号的幅度的第二阈值，则可以判断人醒着。上述第一阈值可以等于或大于上述第二阈值。An increase in the amplitude of the temperature signal may indicate that the person is asleep, and a decrease in the amplitude of the temperature signal may indicate that the person is awake. For example, if the amplitude of the temperature signal increases to exceed a first threshold of the amplitude of the temperature signal, it may be determined that the person is asleep. If the amplitude of the temperature signal decreases to below a second threshold of the amplitude of the temperature signal, it may be determined that the person is awake. The first threshold may be equal to or greater than the second threshold.

人的活动程度的降低可以指示人睡着了，人的活动程度的增大可以指示人醒着。例如，在人的活动程度降低到低于人的活动程度的第一阈值的情况下，可以判断人睡着了。在人的活动程度增大到超过人的活动程度的第二阈值的情况下，可以判断人醒着。上述第一阈值可以等于或小于上述第二阈值。A decrease in the activity level of a person may indicate that the person is asleep, and an increase in the activity level of a person may indicate that the person is awake. For example, when the activity level of a person decreases to a level lower than a first threshold of the activity level of the person, the person may be judged to be asleep. When the activity level of a person increases to a level higher than a second threshold of the activity level of the person, the person may be judged to be awake. The first threshold may be equal to or less than the second threshold.

可以训练在图3B的方法的步骤302中使用的经训练的机器学习模型，以组合关于上表中指示的和上面描述的不同可能输入的行为的上述信息，用于基于上述输入判断人是否睡着。这改进了判断，从而有助于克服将不活动的人醒着时错误地判断为睡着的问题。The trained machine learning model used in step 302 of the method of Figure 3B can be trained to combine the above information about the behavior of the different possible inputs indicated in the above table and described above for judging whether a person is asleep based on the above inputs. This improves the judgment, thereby helping to overcome the problem of misjudging an inactive person as asleep when they are awake.

在步骤302中，可以将心脏信号的幅度和/或与温度信号的幅度变化相关的变量可以输入到经训练的决策树模型中，作为输入。可选地，还可以将温度信号的幅度和/或人的活动程度输入到经训练的决策树模型中，作为进一步输入。经训练的决策树模型根据所提供的输入输出分类结果，该分类结果是醒着的或者是睡着的。可以通过指示人是否睡着的上述睡眠状态变量的形式提供分类结果。可以以预定义的间隔或窗口进行分类。上述信号在当前窗口和相邻窗口中的值可以用于计算该窗口的相应输入。使用包括多个数据集的训练数据训练经训练的决策树模型，其中，每个数据集包括输入到决策树模型的输入，与正确的分类结果(基本事实)相关联。In step 302, the amplitude of the heart signal and/or a variable related to the amplitude change of the temperature signal can be input into the trained decision tree model as an input. Optionally, the amplitude of the temperature signal and/or the activity level of the person can also be input into the trained decision tree model as a further input. The trained decision tree model outputs a classification result based on the input provided, and the classification result is awake or asleep. The classification result can be provided in the form of the above-mentioned sleep state variable indicating whether the person is asleep. Classification can be performed in predefined intervals or windows. The values of the above-mentioned signals in the current window and adjacent windows can be used to calculate the corresponding input of the window. The trained decision tree model is trained using training data including multiple data sets, wherein each data set includes an input input to the decision tree model, which is associated with the correct classification result (basic facts).

在图3B的方法的步骤302之后的可选步骤303中，可以可选地使用形态滤波器进一步处理经训练的机器学习模型输出的分类结果。也就是说，可以使用形态滤波器进一步处理经训练的机器学习模型的输出。形态滤波器可以包括根据以下等式的关闭运算和打开运算：In an optional step 303 following step 302 of the method of FIG. 3B , the classification result output by the trained machine learning model may optionally be further processed using a morphological filter. That is, the output of the trained machine learning model may be further processed using a morphological filter. The morphological filter may include a closing operation and an opening operation according to the following equation:

在这个等式中，“y₁”是来自分类器的二进制信号，“L_p”是以分钟为单位的窗口长度，“y_e”是形态滤波的二进制信号输出。形态关闭运算用"·"算子表示，可以用形态腐蚀和膨胀⊕运算表示为In this equation, "y₁ " is the binary signal from the classifier, "L_p " is the window length in minutes, and "y_e " is the binary signal output by the morphological filter. The morphological closing operation is represented by the "·" operator and can be replaced by morphological erosion. The expansion ⊕ operation is expressed as

形态打开运算由“”运算表示，可以用形态腐蚀/>和膨胀⊕运算表示为The morphological opening operation is performed by "Operation means that morphological corrosion can be used/> The expansion ⊕ operation is expressed as

形态滤波的目的是从机器学习模型的输出中去除短暂的虚假状态变化。The goal of morphological filtering is to remove short-lived spurious state changes from the output of a machine learning model.

因此，步骤303的输出是经训练的机器学习模型的滤波输出，指示人是否睡着。例如，步骤303的输出可以是滤波的睡眠状态变量，指示人是否睡着。无论可选的滤波步骤303是否存在，在最后，图3B的方法输出睡眠/清醒状态，指示人是睡着还是醒着。Thus, the output of step 303 is the filtered output of the trained machine learning model, indicating whether the person is asleep. For example, the output of step 303 can be a filtered sleep state variable, indicating whether the person is asleep. Regardless of whether the optional filtering step 303 is present, at the end, the method of FIG. 3B outputs a sleep/wake state, indicating whether the person is asleep or awake.

关于图3A和图3B的方法的进一步细节，参考根据本发明的第三方面的方法的上述描述。For further details of the method of Figures 3A and 3B, reference is made to the above description of the method according to the third aspect of the present invention.

图4示出了根据本发明的实施例的用于描述如何判断人睡着了的心脏信号的幅度的示例。FIG. 4 shows an example for describing how to determine the amplitude of a heart signal when a person falls asleep according to an embodiment of the present invention.

图4的底部曲线图显示了心脏信号的幅度，即(心脏信号与人身上的测量位置的血容量变化相关)。在人身上的测量位置例如可以是在人的手指、手腕、手和/或手臂上。如上所述，可以从心脏信号计算或获得心脏信号的幅度。y轴以任意单位指示心脏信号的幅度。x轴指示以分钟(min)为单位的时间。图4所示的心脏信号是PPG信号。这只是作为示例，并不限制本发明。图4的顶部的两个曲线图分别显示了不同时间点的心脏信号的滤波版本(例如，通过高通滤波器进行滤波)，其中，y轴表示滤波后的PPG信号的幅度，x轴表示以分钟为单位的时间。The bottom graph of FIG. 4 shows the amplitude of the cardiac signal, i.e. (the cardiac signal is related to the change in blood volume at the measurement location on the human body). The measurement location on the human body can be, for example, on the finger, wrist, hand and/or arm of the person. As described above, the amplitude of the cardiac signal can be calculated or obtained from the cardiac signal. The y-axis indicates the amplitude of the cardiac signal in arbitrary units. The x-axis indicates the time in minutes (min). The cardiac signal shown in FIG. 4 is a PPG signal. This is only an example and does not limit the present invention. The two graphs at the top of FIG. 4 respectively show filtered versions of the cardiac signal at different time points (for example, filtered by a high-pass filter), wherein the y-axis represents the amplitude of the filtered PPG signal and the x-axis represents the time in minutes.

图4左上方的曲线图显示了人是醒着的且不活动时的经滤波的PPG信号。图4右上方的曲线图显示了当人正在睡觉(即人睡着了)时经滤波的PPG信号。因此，在大约50分钟时，人是醒着的且不活动，在大约112分钟时，人睡着了。因此，心脏信号的幅度增大(右上方的曲线图中所示的幅度大于左上方的曲线图中所示的幅度)可以指示人睡着了。The upper left graph of FIG. 4 shows the filtered PPG signal when the person is awake and inactive. The upper right graph of FIG. 4 shows the filtered PPG signal when the person is sleeping (i.e., the person is asleep). Thus, at approximately 50 minutes, the person is awake and inactive, and at approximately 112 minutes, the person is asleep. Therefore, an increase in the amplitude of the heart signal (the amplitude shown in the upper right graph is greater than the amplitude shown in the upper left graph) can indicate that the person is asleep.

图4中所示的PPG信号的幅度与人身上的测量位置(手指、手腕、手和/或手臂)的血容量变化相关，可以从高通滤波的PPG信号计算。例如，首先可以从高通滤波的PPG信号中搜索所有PPG信号峰值。然后，可以将后续峰值幅度相互比较。人的正常心功能产生的峰值具有相似(非常相似)的峰值幅度，峰值之间的持续时间仅在生理心率变异性限制内变化。另一方面，由运动伪像(由于人的运动)引起的峰值通常具有很大的幅度变化。该幅度变化可以用于判断信号质量是否足够好，以计算由于心率引起的平均PPG信号方差。The amplitude of the PPG signal shown in FIG4 is related to the blood volume variation at the measurement location on the person (finger, wrist, hand and/or arm) and can be calculated from the high-pass filtered PPG signal. For example, all PPG signal peaks can first be searched from the high-pass filtered PPG signal. Then, the subsequent peak amplitudes can be compared to each other. Normal cardiac function of a person produces peaks with similar (very similar) peak amplitudes, and the duration between peaks varies only within the physiological heart rate variability limits. On the other hand, peaks caused by motion artifacts (due to the person's movement) usually have a large amplitude variation. This amplitude variation can be used to determine whether the signal quality is good enough to calculate the average PPG signal variance due to the heart rate.

因此，为了计算PPG信号的幅度，可以计算在后续时间窗口内由心功能引起的平均幅度。这能够丢弃PPG信号中表示受噪声干扰(即，受人的运动干扰)的数据的部分，而仅使用PPG信号中表示幅度变化是由于心功能引起的数据的部分，因此，也就是表示与心功能相关的数据的部分。使用PPG信号的幅度来判断睡眠状态的一个优点是，PPG信号的平均幅度比例如心率变异性(heart rate variability，HRV)值更容易计算。因此，使用PPG信号的幅度提供了可靠的算法运算(例如，比使用HRV值的算法运算更可靠)。Therefore, in order to calculate the amplitude of the PPG signal, the average amplitude caused by cardiac function in the subsequent time window can be calculated. This can discard the part of the PPG signal that represents data disturbed by noise (i.e., disturbed by human movement), and only use the part of the PPG signal that represents the data whose amplitude changes are due to cardiac function, and therefore, the part that represents data related to cardiac function. One advantage of using the amplitude of the PPG signal to determine the sleep state is that the average amplitude of the PPG signal is easier to calculate than, for example, the heart rate variability (HRV) value. Therefore, using the amplitude of the PPG signal provides a reliable algorithm operation (for example, more reliable than an algorithm operation using HRV values).

图5示出了根据本发明的实施例的用于描述如何判断人醒着的心脏信号的幅度的示例。FIG. 5 shows an example for describing how to determine the amplitude of a heart signal of a person who is awake according to an embodiment of the present invention.

与图4的两个顶部曲线图一样，图5的两个曲线图分别显示了与不同时间点的测量位置(例如，在人的手指、手腕、手和/或手臂上)的血容量变化相关的心脏信号的滤波版本(例如，通过高通滤波器进行滤波)，其中，y轴表示滤波后的PPG信号的幅度，x轴表示以分钟为单位的时间。图5所示的心脏信号是PPG信号。这只是作为示例，并不限制本发明。As with the two top graphs of FIG. 4 , the two graphs of FIG. 5 respectively show filtered versions (e.g., filtered by a high-pass filter) of a heart signal associated with changes in blood volume at a measurement location (e.g., on a person's finger, wrist, hand, and/or arm) at different points in time, wherein the y-axis represents the amplitude of the filtered PPG signal and the x-axis represents time in minutes. The heart signal shown in FIG. 5 is a PPG signal. This is only for example and does not limit the present invention.

图5的曲线图显示了当用户醒来时心脏信号的幅度是如何变化的。根据图5的左曲线图，例如，在大约464分钟的时间，人睡着了，并且根据图5的右曲线图，例如，在大约526分钟的时间，人醒着。因此，图5的左曲线图显示了人醒着之前睡着时滤波后的PPG信号。图5的右曲线图示出了在人已经醒来并且例如正在进行晨间活动之后滤波的PPG信号。因此，心脏信号的幅度降低(左曲线图中所示的幅度大于右曲线图中所示的幅度)可能指示人醒着。根据图4的两个顶部曲线图和图5的两个曲线图中所示的心脏信号(例如，PPG信号)的幅度，可以确定心脏信号的幅度的第一阈值和第二阈值，其中，在幅度高于第一阈值的情况下，人睡着了，在幅度低于第二阈值的情况下，人是醒着的。第一阈值可以等于或大于第二阈值。The graph of FIG. 5 shows how the amplitude of the heart signal changes when the user wakes up. According to the left graph of FIG. 5 , for example, at a time of about 464 minutes, the person fell asleep, and according to the right graph of FIG. 5 , for example, at a time of about 526 minutes, the person was awake. Therefore, the left graph of FIG. 5 shows the filtered PPG signal when the person was asleep before waking up. The right graph of FIG. 5 shows the filtered PPG signal after the person has woken up and is, for example, performing morning activities. Therefore, a decrease in the amplitude of the heart signal (the amplitude shown in the left graph is greater than the amplitude shown in the right graph) may indicate that the person is awake. According to the amplitude of the heart signal (e.g., PPG signal) shown in the two top graphs of FIG. 4 and the two graphs of FIG. 5 , a first threshold and a second threshold of the amplitude of the heart signal can be determined, wherein, in the case where the amplitude is higher than the first threshold, the person is asleep, and in the case where the amplitude is lower than the second threshold, the person is awake. The first threshold may be equal to or greater than the second threshold.

图6示出了根据本发明的实施例的在人睡着了的情况下的温度信号和温度信号的处理的示例。FIG. 6 shows an example of a temperature signal and processing of the temperature signal in a case where a person falls asleep according to an embodiment of the present invention.

图6的顶部曲线图示出了在人醒着且不活动的时间和人睡着的时间(例如，在晚上)与人的皮肤温度相关的温度信号的示例。也就是说，图6示出了人睡着时的温度信号的示例。温度信号可以由测量人的四肢(例如，手指、手腕、手和/或手臂)的皮肤温度的温度传感器生成。图6的底部曲线图示出了与温度信号的幅度随时间的变化相关的变量的示例，其中，该变量是根据温度信号(例如，顶部曲线图所示)通过计算滑动时间窗口内的温度信号导数的至少两个(N≥2)绝对值之和来计算的。图6的顶部曲线图的y轴指示以℃为单位的温度信号的幅度，图6的底部曲线图的y轴指示与温度的幅度变化相关的变量的值。图6的顶部和底部曲线图的x轴指示以分钟为单位的时间。The top graph of FIG. 6 shows an example of a temperature signal associated with a person's skin temperature during a time when the person is awake and inactive and during a time when the person is asleep (e.g., at night). That is, FIG. 6 shows an example of a temperature signal when a person is asleep. The temperature signal can be generated by a temperature sensor measuring the skin temperature of a person's limbs (e.g., fingers, wrists, hands, and/or arms). The bottom graph of FIG. 6 shows an example of a variable associated with a change in the amplitude of a temperature signal over time, wherein the variable is calculated based on the temperature signal (e.g., as shown in the top graph) by calculating the sum of at least two (N≥2) absolute values of the derivative of the temperature signal within a sliding time window. The y-axis of the top graph of FIG. 6 indicates the amplitude of the temperature signal in ° C, and the y-axis of the bottom graph of FIG. 6 indicates the value of the variable associated with the amplitude change of the temperature. The x-axes of the top and bottom graphs of FIG. 6 indicate time in minutes.

如图6所示，x轴(时间轴)分为四个部分。第1部分对应于人不活动的时间段(例如，一动不动地在平板电脑上观看电影)。第2部分对应于人准备睡觉的时间段。第3部分对应于人睡着了(例如，躺在床上等待睡觉)的时间段。第4部分对应于人睡着了的时间段。因此，温度信号的幅度的增大(在图6的顶部曲线图的第3部分中示出)可以指示人睡着了。此外，与温度信号的幅度变化相关的变量的降低(在图6的底部曲线图的第3部分中示出)可以指示人睡着了。As shown in FIG6 , the x-axis (time axis) is divided into four parts. Part 1 corresponds to a time period when a person is inactive (e.g., watching a movie on a tablet computer without moving). Part 2 corresponds to a time period when a person is preparing to go to bed. Part 3 corresponds to a time period when a person is asleep (e.g., lying in bed waiting to go to sleep). Part 4 corresponds to a time period when a person is asleep. Therefore, an increase in the amplitude of the temperature signal (shown in Part 3 of the top graph of FIG6 ) can indicate that a person is asleep. In addition, a decrease in a variable associated with a change in the amplitude of the temperature signal (shown in Part 3 of the bottom graph of FIG6 ) can indicate that a person is asleep.

图7示出了根据本发明的实施例的在人醒着的情况下的温度信号和温度信号的处理的示例。FIG. 7 shows an example of a temperature signal and processing of the temperature signal when a person is awake according to an embodiment of the present invention.

图7的顶部曲线图示出了在人睡着的时间和人醒着的时间(例如，在早晨)与人的皮肤温度相关的温度信号的示例。也就是说，图7显示了人醒着时的温度信号的示例。温度信号可以由测量人的四肢(例如，手指、手腕、手和/或手臂)的皮肤温度的温度传感器生成。图7的底部曲线图示出了与温度信号的幅度随时间的变化相关的变量的示例，其中，该变量是根据温度信号(例如，顶部曲线图所示)通过计算滑动时间窗口内的温度信号导数的至少两个(N≥2)绝对值之和来计算的。图7的顶部曲线图的y轴指示以℃为单位的温度信号的幅度，图7的底部曲线图的y轴指示与温度信号的幅度变化相关的变量的值。图7的顶部和底部曲线图的x轴指示以分钟为单位的时间。The top graph of FIG. 7 shows an example of a temperature signal associated with a person's skin temperature when the person is asleep and when the person is awake (e.g., in the morning). That is, FIG. 7 shows an example of a temperature signal when the person is awake. The temperature signal can be generated by a temperature sensor measuring the skin temperature of a person's limbs (e.g., fingers, wrists, hands, and/or arms). The bottom graph of FIG. 7 shows an example of a variable associated with a change in the amplitude of a temperature signal over time, wherein the variable is calculated based on the temperature signal (e.g., as shown in the top graph) by calculating the sum of at least two (N≥2) absolute values of the derivative of the temperature signal within a sliding time window. The y-axis of the top graph of FIG. 7 indicates the amplitude of the temperature signal in ° C, and the y-axis of the bottom graph of FIG. 7 indicates the value of the variable associated with the change in the amplitude of the temperature signal. The x-axis of the top and bottom graphs of FIG. 7 indicates time in minutes.

如图7所示，x轴(时间轴)分为两部分。图7的第1部分对应于人正在睡觉(即人睡着了)的时间段。图7的第2部分对应于人醒来(因此是醒着的)的时间段。因此，温度信号的幅度的降低(在图7的顶部曲线图的第2部分中示出)可以指示人醒着。此外，与温度信号的幅度变化相关的变量的增大(在图7的底部曲线图的第2部分中示出)可以指示人醒着。术语“唤醒”可以用作术语“醒着”的同义词。As shown in FIG. 7 , the x-axis (time axis) is divided into two parts. Part 1 of FIG. 7 corresponds to a time period when a person is sleeping (i.e., the person is asleep). Part 2 of FIG. 7 corresponds to a time period when a person is awake (and therefore awake). Therefore, a decrease in the amplitude of the temperature signal (shown in Part 2 of the top graph of FIG. 7 ) may indicate that the person is awake. In addition, an increase in a variable associated with a change in the amplitude of the temperature signal (shown in Part 2 of the bottom graph of FIG. 7 ) may indicate that the person is awake. The term “wake” may be used as a synonym for the term “awake”.

图8的曲线图(A)显示了人在一段时间内的活动程度(活动水平)。因此，曲线图(A)的y轴表示人的活动程度。图8的曲线图(B)对应于图6的顶部曲线图，图8的曲线图(C)对应于图6的底部曲线图。因此，为了描述图8的曲线图(B)和(C)，参考图6的上述描述。图8的曲线图(D)示出了与在人身上的测量位置(例如，在手指、手腕、手和/或手臂上)的血容量变化相关的心脏信号的幅度。心脏信号是PPG信号，这只是作为示例。图8的曲线图(D)的y轴指示心脏信号(PPG信号)的幅度。图8的曲线图(D)对应于图4的底部曲线图。因此，图4的描述相应地对于图8的曲线图(D)是有效的。图8的底部曲线图(E)示出了随时间的睡眠状态变量，其中，睡眠状态变量指示人是否睡着了(即睡着了或醒着)。底部曲线图(E)的y轴指示睡眠状态变量的值。根据图8，所述睡眠状态变量可以是二元变量，其中，睡眠状态变量的低值(零值，“0”)指示人睡着了，睡眠状态变量的高值(一值，“1”)指示人是醒着的(未睡着)。反之亦然，即低值可以指示清醒状态，高值可以指示睡眠状态(图8中未示出)。曲线图(E)的实线示出了基本事实，即人实际上是醒着的还是睡着了。曲线图(E)的虚线示出了当如上所述根据本发明执行用于判断人是否睡着了的方法时生成的睡眠状态变量。曲线图(E)的虚线示出了仅使用人的活动程度来判断人是否睡着了时生成的睡眠状态变量。图8的每个曲线图的x轴指示以分钟为单位的时间。The graph (A) of FIG8 shows the activity level (activity level) of a person over a period of time. Therefore, the y-axis of the graph (A) represents the activity level of the person. The graph (B) of FIG8 corresponds to the top graph of FIG6, and the graph (C) of FIG8 corresponds to the bottom graph of FIG6. Therefore, in order to describe the graphs (B) and (C) of FIG8, reference is made to the above description of FIG6. The graph (D) of FIG8 shows the amplitude of the heart signal associated with the change of blood volume at the measurement position on the person (e.g., on the finger, wrist, hand and/or arm). The heart signal is a PPG signal, which is just an example. The y-axis of the graph (D) of FIG8 indicates the amplitude of the heart signal (PPG signal). The graph (D) of FIG8 corresponds to the bottom graph of FIG4. Therefore, the description of FIG4 is valid for the graph (D) of FIG8 accordingly. The bottom graph (E) of FIG8 shows the sleep state variable over time, wherein the sleep state variable indicates whether the person is asleep (i.e., asleep or awake). The y-axis of the bottom graph (E) indicates the value of the sleep state variable. According to Figure 8, the sleep state variable can be a binary variable, wherein a low value (zero value, "0") of the sleep state variable indicates that the person is asleep, and a high value (one value, "1") of the sleep state variable indicates that the person is awake (not asleep). Vice versa, that is, a low value can indicate an awake state, and a high value can indicate a sleep state (not shown in Figure 8). The solid line of the graph (E) shows the basic fact, that is, whether the person is actually awake or asleep. The dotted line of the graph (E) shows the sleep state variable generated when the method for determining whether a person is asleep is performed according to the present invention as described above. The dotted line of the graph (E) shows the sleep state variable generated when only the person's activity level is used to determine whether the person is asleep. The x-axis of each graph of Figure 8 indicates time in minutes.

如图8所示，曲线图的x轴(时间轴)分为四个部分。第1部分对应于人不活动的时间段(例如，一动不动地在平板电脑上观看电影)。第2部分对应于人准备睡觉的时间段。第3部分对应于人睡着了(例如，躺在床上等待睡觉)的时间段。第4部分对应于人睡着了的时间段。As shown in FIG8 , the x-axis (time axis) of the graph is divided into four parts. Part 1 corresponds to a time period when the person is inactive (e.g., watching a movie on a tablet without moving). Part 2 corresponds to a time period when the person is preparing to go to bed. Part 3 corresponds to a time period when the person is asleep (e.g., lying in bed waiting to go to sleep). Part 4 corresponds to a time period when the person is asleep.

从图8可以得出，人的活动程度的降低(在图8的曲线图(A)的第3部分中示出)可以指示人睡着了。此外，温度信号的幅度的增大(在图8的曲线图(B)的第3部分中示出)可以指示人睡着了。此外，与温度信号的幅度变化相关的变量的降低(在图8的曲线图(C)的第3部分中示出)可以指示人睡着了。此外，心脏信号的幅度的增大(在图8的曲线图(D)的第3部分中，这在图4的两个顶部曲线图中指示)可以指示人睡着了。It can be concluded from FIG. 8 that a decrease in the activity level of a person (shown in the third part of the graph (A) of FIG. 8 ) can indicate that the person has fallen asleep. In addition, an increase in the amplitude of the temperature signal (shown in the third part of the graph (B) of FIG. 8 ) can indicate that the person has fallen asleep. In addition, a decrease in the variable associated with the amplitude change of the temperature signal (shown in the third part of the graph (C) of FIG. 8 ) can indicate that the person has fallen asleep. In addition, an increase in the amplitude of the heart signal (in the third part of the graph (D) of FIG. 8 , which is indicated in the two top graphs of FIG. 4 ) can indicate that the person has fallen asleep.

如曲线图(E)所示，其中，虚线指示仅根据人的活动程度判断睡眠状态，这种判断可能错误地判断人睡着了。即，如曲线图(A)所示，在第1部分和第2部分中，人的活动程度已经发生变化。结果，由曲线图(E)的虚线指示的睡眠状态变量在第1部分的时间段期间错误地从高值(指示人醒着)改变到低值(指示人正在睡觉)，因为在该时间段期间活动程度降低。此外，由曲线图(E)的虚线指示的睡眠状态变量在第2部分的时间段期间错误地从低值改变到高值，然后再次改变到低值，因为在该时间段期间，活动程度首先增大然后再次降低。因此，仅使用活动程度来判断人是否睡着了会产生错误的结果，因为仅根据人的活动程度，可能会错误地将人的清醒状态(其中，人不活动(仅有很少的运动或没有运动))判断为人睡着了。如上所述，在第1部分和第2部分中，人仍然醒着(这在曲线图(E)中由指示基本事实的实线表示)。As shown in the graph (E), where the dotted line indicates that the sleep state is judged only based on the activity level of the person, this judgment may erroneously judge that the person is asleep. That is, as shown in the graph (A), the activity level of the person has changed in Part 1 and Part 2. As a result, the sleep state variable indicated by the dotted line of the graph (E) erroneously changes from a high value (indicating that the person is awake) to a low value (indicating that the person is sleeping) during the time period of Part 1 because the activity level decreases during this time period. In addition, the sleep state variable indicated by the dotted line of the graph (E) erroneously changes from a low value to a high value during the time period of Part 2, and then changes to a low value again because during this time period, the activity level first increases and then decreases again. Therefore, using only the activity level to judge whether a person is asleep will produce an erroneous result, because based only on the activity level of the person, the awake state of the person (where the person is inactive (only a little movement or no movement)) may be erroneously judged as a person asleep. As described above, in Part 1 and Part 2, the person is still awake (this is indicated by the solid line indicating the basic fact in the graph (E)).

根据本发明，可以通过使用心脏信号的幅度和/或关于温度信号的幅度变化的变量来判断人是否睡着了，从而克服这个问题。此外，根据本发明，可以可选地使用活动程度和/或温度信号的幅度来判断人是否睡着了。因此，由本发明的用于判断由图8的曲线图(E)中的虚线指示的人是否睡着了的方法生成的睡眠状态变量仅在第4部分开始时从高值(指示人醒着)改变到低值(指示人睡着了)。在第1部分和第2部分的时间段内，曲线图(E)中的虚线不会因为人的活动程度的变化而改变。According to the present invention, this problem can be overcome by using the amplitude of the heart signal and/or a variable about the amplitude change of the temperature signal to determine whether a person is asleep. In addition, according to the present invention, the activity level and/or the amplitude of the temperature signal can be optionally used to determine whether a person is asleep. Therefore, the sleep state variable generated by the method of the present invention for determining whether the person indicated by the dotted line in the curve diagram (E) of Figure 8 is asleep only changes from a high value (indicating that the person is awake) to a low value (indicating that the person is asleep) at the beginning of Part 4. During the time period of Parts 1 and 2, the dotted line in the curve diagram (E) will not change due to changes in the person's activity level.

本发明已结合各种实施例作为示例并结合实现方式进行描述。但是，根据对附图、本发明和独立权利要求的研究，本领域技术人在实施所要求保护的主题时，能够理解和实现其它变型。在权利要求书以及说明书中，词语“包括”不排除其它元件或步骤，且不定冠词“一个”不排除多个。单个元件或其他单元可以满足权利要求书中描述的若干实体或项目的功能。在互不相同的从属权利要求中列举某些措施并不指示这些措施的组合不能用于有利的实现方式中。The invention has been described in conjunction with various embodiments as examples and in conjunction with implementations. However, from a study of the drawings, the invention and the independent claims, other variations will be understood and implemented by those skilled in the art when implementing the claimed subject matter. In the claims and in the specification, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" does not exclude a plurality. A single element or other unit may fulfil the functions of several entities or items described in the claims. The enumeration of certain measures in mutually different dependent claims does not indicate that a combination of these measures cannot be used in an advantageous implementation.