A bed with a function to determine the risk of congestive heart failure

Abstract

A technique is disclosed for determining a probability measure of a risk of heart failure (e.g., congestive heart failure, i.e., CHF) for a user of a bed system. The method may include the steps of receiving, by a computer system, user data collected by a sensor of the bed system when the user lies on the bed system, determining sleep fragmentation information for a sleep session of the user based on an analysis of the user data, providing the sleep fragmentation information as an input to a model trained to determine a probability measure of a risk of heart failure for a user of the bed system based at least in part on user data associated with the user, receiving as an output from the model the probability measure of the risk of heart failure for the user of the bed system, and performing an action based on the probability measure.

Description

[Technical field]

[0001] [CROSS REFERENCE TO RELATED APPLICATIONS] This application claims priority to U.S. Provisional Patent Application No. 63 / 305,713, filed February 2, 2022. The disclosure of that prior application is incorporated by reference in its entirety into the disclosure of this application.

[0002] [Technical field] This specification describes bed related devices, systems and methods capable of predicting congestive heart failure and heart failure in a user on the bed.

[0003] [Background technology] Generally, a bed is a piece of furniture used as a place to sleep or relax. Many modern beds include a soft mattress on top of a bed frame. The mattress may contain springs, foam materials, and / or air chambers to support the weight of one or more occupants. Summary of the Invention

[0004] The present disclosure generally relates to systems, methods and techniques for predicting the risk of congestive heart failure (CHF) and / or other heart failure in users of a bed system, such as a smart bed. More specifically, the disclosed techniques provide for unobtrusively monitoring the cardiovascular activity of a user of the bed system and determining indicators of abnormal patterns in the user's sleep fragmentation and cardiovascular activity to generate an indicator (e.g., probability, likelihood, and / or measure) of heart failure risk. The indicator of heart failure risk may be generated over time (e.g., over multiple sleep sessions) and / or for any given night. Data collected about a user by the bed system may be provided as input to a model trained using machine learning techniques to determine a probability measure of CHF risk or other heart failure for the user. Various data may be collected and used to accurately assess the risk of developing CHF in various users of the bed system. For example, the probability measure of risk of CHF may be based on the user's sleep fragmentation information. The probability measure may also be based on other data including, but not limited to, detection of sleep and / or wakefulness, wakefulness after sleep onset, wakefulness before sleep onset, sleep efficiency, sleep duration, sleep episode length, time to sleep onset, wakefulness response index, heart rate (HR), respiratory rate (RR), daytime wakefulness, demographic factors, blood pressure, and / or interbeat interval (IBI). Thus, the techniques of the present disclosure may be used to quantify risk of CHF as a probability measure rather than merely identifying risk of CHF. A logistic regression model may be used to determine the probability measure.

[0005] CHF may describe a dysfunctional condition in which the user's heart does not pump blood normally. CHF may be classified as acute or chronic. Acute CHF may develop suddenly and may lead to a medical emergency. Chronic CHF, on the other hand, may develop gradually over time. CHF may be a marker of deteriorating cardiac health. For example, it may cause the heart to enlarge or beat faster. This may weaken the heart and reduce its pumping efficiency over time. Furthermore, due to the dysfunction of the heart muscle, a backlog of blood may occur that is returned to the heart, thereby causing congestion. Oxygen levels may drop within the cells of the user's body, fluid may build up in the lungs, and the user may experience heart palpitations and / or shortness of breath. CHF may have significant effects on the user's overall health, well-being, and organs.

[0006] The sleep fragmentation information can be used as an input to the model described herein to generate a risk probability measure of CHF in a user of the bed system. Sleep fragmentation can be defined as repeated short sleep interruptions during a sleep session (and / or sleep cycle), which can lead to excessive fatigue the next day. Sleep fragmentation can affect the user's natural sleep session and / or sleep stages and can be a symptom of a sleep disorder. Furthermore, sleep fragmentation can be quantified based on data that can be sensed by the bed system, such as wake after sleep onset, wake before sleep onset, wake response index, and wakefulness level.

[0007] One or more embodiments described herein may include a method for determining a probability measure of a risk of heart failure for a user of a bed system, comprising: receiving, by a computer system, user data collected by sensors of the bed system while the user lies on the bed system; determining, by the computer system, sleep fragmentation information for a sleep session of the user based on an analysis of the user data; providing, by the computer system, the sleep fragmentation information as an input to a model trained to determine a probability measure of a risk of heart failure for a user of the bed system based at least in part on user data associated with the user; receiving, by the computer system, the probability measure of the risk of heart failure for the user of the bed system as an output from the model; and performing, by the computer system, an action based on the probability measure.

[0008] In some implementations, the embodiments described herein may optionally include one or more of the following features: For example, the user data may include a ballistocardiogram (BCG) signal; The model may be trained by a training dataset of historical user data associated with the user; The model may be trained by a training dataset of historical user data associated with various users of the bed system; The model may be a logistic regression model; The model may be trained by the computer system using a feature vector with a numerical value assigned to it corresponding to the training user data; The training user data may include at least one of wake time after sleep onset, sleep efficiency, sleep duration, and sleep episode length; The training user data may also include at least one of wake time before sleep onset, amount of sleep interruption, heart rate, heart rate variability, respiration rate, and daytime alertness level; The training user data may also include demographic information, which may include at least one of weight, age, sex, and body mass index (BMI).

[0009] In some implementations, determining sleep fragmentation information for the user by the computer system may include identifying periods during the sleep session when the user is asleep and periods during the sleep session when the user is awake. In some implementations, the method may also include providing, by the computer system, at least one of (i) wake-after-sleep-onset data, (ii) sleep efficiency data, (iii) sleep duration data, and (iv) sleep episode length data as input to the model. The model may be trained by the computer system to correlate at least one of (i)-(iv) with a stage of sleep fragmentation for the user of the bed system. Additionally, the probability measure of risk of heart failure may be a numerical value.

[0010] In some implementations, the output from the model may be a value between 0 and 1, and the method may also include generating, by the computer system, a score for the risk of heart failure by multiplying the value by a numerical coefficient. In some cases, the probability measure of the risk of heart failure may be a categorical value, and the categorical value may be at least one of low risk, medium risk, and high risk.

[0011] In some implementations, performing an action based on the probability measure by the computer system may include outputting an alert to a user device of the user indicating that the user is at risk for heart failure based on a determination that the probability measure exceeds a threshold range. Further, performing an action based on the probability measure by the computer system may include sending a notification to a user device of a health care provider indicating that the user is at risk for heart failure based on a determination that the probability measure exceeds a threshold range. Further, the heart failure may be congestive heart failure.

[0012] One or more embodiments described herein may include a method for determining a probability measure of a risk of heart failure for a user of a bed system, the method comprising: receiving, by a computer system, user data collected by a sensor of the bed system while the user lies on the bed system; determining, by the computer system, sleep fragmentation information for the user's sleep session based on an analysis of the user data; determining, by the computer system, at least one of (i) wake-after-sleep-onset data, (ii) sleep efficiency data, (iii) sleep duration data, and (iv) sleep episode length data based on an analysis of the user data; providing, by the computer system, the sleep fragmentation information and at least one of (i) to (iv) as inputs to a model trained to determine a probability measure of a risk of heart failure for a user of a bed system based at least in part on user data associated with the user; receiving, by the computer system, the probability measure of the risk of heart failure for the user of the bed system as an output from the model; and performing, by the computer system, an action based on the probability measure.

[0013] One or more embodiments described herein may also include a system for determining a probability measure of a risk of heart failure for a user of a bed system, the system comprising a bed having at least one sensor and a computer system in communication with the at least one sensor of the bed, the computer system may be configured to receive user data from the at least one sensor collected by the at least one sensor when a user lies on the bed, determine sleep fragmentation information for a sleep session of the user based on an analysis of the user data, provide the sleep fragmentation information as an input to a model trained to determine a probability measure of a risk of heart failure for a user of a bed based at least in part on user data associated with the user, receive as an output from the model the probability measure of the risk of heart failure for the user of the bed, and perform an action based on the probability measure.

[0014] The system may optionally include one or more of the following features. For example, the bed may further include a controller and the computer system may be the controller. The computer system may also be separate from the bed.

[0015] One or more embodiments described herein may comprise a system for determining a probability measure of a risk of congestive heart failure for a user of a bed system, the system comprising: a bed having at least one sensor; and a computer system in communication with the at least one sensor of the bed, the computer system being capable of determining sleep fragmentation as a function of sensed data from the at least one sensor of a sleeper of the bed, calculating a probability measure of congestive heart failure as a function of the determined sleep fragmentation, and outputting a signal for performing an action based on the calculated probability measure of congestive heart failure.

[0016] One or more embodiments described herein may also comprise a system for determining a probability measure of a risk of heart failure for a user of a bed system, the system comprising: a bed having at least one sensor; and a computer system in communication with the at least one sensor of the bed, the computer system being configured to: receive user data from the at least one sensor collected by the at least one sensor when a user lies on the bed; determine sleep fragmentation information for a sleep session of the user based on an analysis of the user data; determine at least one of (i) wake-after-sleep-onset data, (ii) sleep efficiency data, (iii) sleep duration data, and (iv) sleep episode length data based on an analysis of the user data; provide the sleep fragmentation information and at least one of (i)-(iv) as inputs to a model trained to determine a probability measure of a risk of heart failure for a user of a bed based at least in part on user data associated with the user; receive as an output from the model the probability measure of the risk of heart failure for the user of the bed; and perform an action based on the probability measure.

[0017] One or more embodiments described herein may also comprise a system for determining a probability measure of risk of congestive heart failure for a user of a bed system, the system comprising: a bed having at least one sensor; and a computer system in communication with the at least one sensor of the bed, the computer system being configured to: determine post-sleep onset wakefulness data as a function of sensed data of a sleeper from the at least one sensor; determine sleep efficiency data as a function of the sensed data from the sleeper; determine sleep episode length data as a function of the sensed data from the sleeper; calculate a probability measure of risk of congestive heart failure as a function of the determined post-sleep onset wakefulness data, the determined sleep efficiency data, and the determined sleep episode length data; and output a signal for performing an action based on the calculated probability measure of congestive heart failure.

[0018] The devices, systems and techniques described herein may provide one or more of the following advantages. For example, the techniques of the present disclosure leverage the sleep and detection capabilities of a bed system (e.g., a smart bed with sensors) to quantify sleep fragmentation and determine a probability measurement of the bed system's user's CHF risk. By using existing technology, the bed system's user can gain greater insight into their health, more specifically their cardiac health, without having to invest in additional technology. For example, a sensor that can be used to measure pressure changes in the air chamber of the bed system can also be intended to measure pressure changes that are useful for determining sleep fragmentation information about the user.

[0019] Furthermore, the disclosed technology utilizes machine learning techniques to provide accurate evaluation of various data to determine probability measures of CHF risk in various users. The disclosed technology can utilize diverse and vast data (both historical and real-time data for specific users and / or user groups), including various data types and data from various sources, to more accurately determine probability measures. For example, system-based determinations and predictions (e.g., determined alertness levels) can be combined with raw sensor data (e.g., BCG signals) to accurately and quickly determine probability measures for many users with various health conditions, demographic factors, and other unique attributes. The machine learning model can also be continuously trained and improved as more data is collected by the bed system and more probability measure determinations are made for various users. As a result, CHF risk in various users can be accurately predicted and assessed over time or in real time.

[0020] As another example, the technology of the present disclosure provides for unobtrusive, automatic monitoring of a user's health while the user sleeps as well as while awake to gather further insights regarding the user's health over time and / or during sleep sessions. As a result, the user would not need to remember or keep records of changes in their health status or the need to schedule a doctor's appointment for a routine check-up or an unscheduled check-up if the user is concerned about their heart health. Furthermore, if a medical concern is discovered using the technology of the present disclosure, the user and associated health care providers may have access to months of user data before the medical concern appeared, was detected, or was otherwise evident. This wealth of user data may help the user and / or associated health care providers understand how the user's medical concern arose and / or how to address or prevent the medical concern from significantly impacting the user's life.

[0021] Furthermore, a user may visit a doctor to check their cardiac health at various times throughout their life (e.g., monthly, annually, every few months, years, etc.). Such brief visits may not accurately capture the user's cardiovascular activity, especially since the user is typically awake and the body is engaged in more tasks than when relaxed and resting. Thus, the techniques of the present disclosure may be used to measure and quantify cardiovascular activity over many sleep sessions in which the user's body is relaxed and resting to accurately determine the user's cardiac health. Data may be collected about the user over an extended period of time while the user's body is at rest. The ability to monitor a user's health while sleeping may provide more accurate health detection and reduce the need for the user to visit a doctor to check their health and cardiac metrics. This may also be beneficial in reducing the user's medical costs, especially when the user does not have insurance coverage for cardiac health testing and treatment. When multisensory data about a user is dynamically collected over time and fed into models trained to measure cardiac health, abnormalities in cardiovascular activity can be more accurately and quickly identified and assessed.

[0022] Similarly, the techniques of the present disclosure provide for determining a probability measure of CHF risk in a user on an extended, nightly basis, which may provide an accurate and frequent quantification of cardiac health in comparison to clinical polysomnography (PSG) recordings, which may only be performed once every few months or years due to cost and patient burden.

[0023] The details of one or more implementations are set forth in the accompanying drawings and the description below. Other features and advantages will become apparent from the description, drawings, and claims. [Brief description of the drawings]

[0024] [Figure 1] FIG. 1 is a conceptual diagram for determining a probability measure of congestive heart failure in a user of a bed system.

[0025] [Diagram 2] FIG. 2 is a conceptual diagram for training a model for determining a probability measure of congestive heart failure in a user of a bed system.

[0026] [Diagram 3] FIG. 3 is a flow chart of a process for determining a probability measure of congestive heart failure in a user of a bed system.

[0027] [Figure 4] FIG. 4 is a system diagram of one or more components that may be used to implement the techniques of this disclosure.

[0028] [Diagram 5] FIG. 5 is a schematic diagram illustrating an example of a computing device and a mobile computing device.

[0029] Like reference symbols in the various drawings indicate like elements. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0030] The present specification generally relates to a technique for estimating the risk of cardiac health deterioration, such as CHF, in a user of a bed system. More specifically, multi-sensory data collected by sensors (e.g., pressure sensors, temperature sensors, motion sensors, acoustic sensors, etc.) of a bed system (e.g., a smart bed) can be combined with data insights determined by a computing system regarding a user of the bed system to determine a probability measure of the risk of CHF in the user. The computing system can use a machine learning training model to determine the probability measure based on various data points. For example, the computing system can collect data measured by the bed system throughout a user's sleep session. Using this collected data, the computing system can estimate a coefficient that can quantify the CHF risk for a particular user of the bed system. This coefficient can be based, for example, on sleep fragmentation information determined by the computing system and further on data collected during a sleep session or over time (e.g., over several sleep sessions). This coefficient can then be provided by the computing system as an input to a machine learning and / or AI-based model for quantifying CHF risk. As described herein, the model may determine a probability measure of CHF risk in a particular user. Based on the probability measure determined using the model, the computing system may generate outputs, alerts, and / or notifications to be presented to a user device of a user of the bed system, a healthcare provider, or other relevant users.

[0031] Referring to the drawings, Figure 1 is a conceptual diagram for determining a probability measure of congestive heart failure in a user 106 of a bed system 100. The bed system 100, the computer system 102, the user device 112 and the data store 114 are in communication via the network(s) 108. The network(s) 108 may provide wired and / or wireless communication (e.g. Bluetooth, Internet, WIFI, local area network, etc.) between the components described herein.

[0032] Briefly, the bed system 100 may be a smart bed. The bed system 100 may include one or more sensors 105A-N. The sensors 105A-N may be arranged in various layouts and / or arrays throughout the bed system 100. For example, one or more of the sensors 105A-N may be configured in the base, side rails, headboard, and / or legs of the bed system 100 or may be otherwise integrated. The one or more sensors 105A-N may be, for example, load cells positioned in each of the legs of the bed system 100 and configured to measure weight and / or pressure changes (e.g., subtle weight changes) within the bed system 100. The one or more sensors 105A-N may also include pressure sensors integrated within the bed system 100, such as in the mattress and / or in a pump system fluidly connected to the air chamber of the mattress (if the mattress is an air mattress or other inflatable mattress). One or more of the sensors 105A-N may also include a temperature sensor. In some implementations, one or more of the sensors 105A-N may include an acoustic sensor configured to detect sounds within a predetermined decibel range for identifying snoring or other sleep-related activities. One or more other types of sensors may be used in the bed system 100. Furthermore, one or more of the sensors 105A-N may be part of a pad, mat, or other material that covers a portion of the bed system 100, such as a pad that covers a portion of a mattress.

[0033] The bed system 100 may include one or more air chambers capable of receiving a fluid, such as air or liquid, for inflation and / or deflation of the air chambers. A pressure sensor may be positioned in the air chamber. The pressure sensor may also be fluidly connected to the air chamber, for example along an air tube. The air chamber may be in fluid communication with a pump of the bed system 100. The pump may be in electrical communication with a remote control and / or a user device 112 via a control box. The control box may operate the pump to cause an increase and decrease in fluid pressure in the air chamber based on commands entered by a user 106, for example, using the remote control or the user device 112. Pressure changes and other data may be collected from sensors 105A-N, which may be located, positioned or otherwise integrated in the air chambers, pumps and / or other parts of the bed system 100.

[0034] In some implementations, information collected by the pressure sensor may be analyzed by the bed system 100 and / or the computer system 102 to determine various states of the user 106 lying on the bed system 100. As an example, the computer system 102 may use information collected by the pressure sensor to determine the heart rate or breathing rate of the user 106. The user may lie on one side of the bed system 100 that includes an air chamber, and the pressure sensor may monitor the pressure fluctuations in the chamber, which may be used by the computer system 102 to determine the heart rate and / or breathing rate of the user 106. More specifically, the computer system 102 may receive pressure data from the pressure sensor, which may be used by the computer system 102 to generate a time series of the cardiac and / or respiratory activity of the user 106. The time series may be considered as raw data. The computer system 102 may then analyze the time series to identify peaks indicative of changes in the heart rate. The computer system 102 may count peaks over time (e.g., over a predetermined amount of time) to determine the heart rate of the user 106. The computer system 102 may perform a similar process to determine the respiration rate of the user 106. The determined heart rate and determined respiration rate may be features. In this manner, the raw data of the time series may be converted into feature values ​​that may be used by a machine learning training model to determine a probability measure of CHF risk for the user 106. As will be further described with reference to FIG. 2, these feature values ​​may be used along with other feature values ​​to train a model using guided learning techniques. In some implementations, these feature values ​​may also be used along with unguided learning and / or classification techniques to train a model and then use the model during runtime.

[0035] As another example, additional processing may be performed using the collected data to determine the sleep state (e.g., awake, light sleep, deep sleep) of the user 106. For example, the computer system 102 may determine when the user 106 fell asleep and the various sleep states (e.g., sleep stages) of the user 106 while asleep. Based on the determined heart rate, respiration rate, and / or sleep state of the user 106, the computer system 102 may determine sleep fragmentation information and information regarding the objective sleep quality of the user 106.

[0036] Additional information associated with the user 106 of the bed system 100 that can be determined using information collected by the sensors 105A-N includes the movement of the user 106, the presence of the user 106 on the surface of the bed system 100, the weight of the user 106, cardiac arrhythmia of the user 106, snoring of the user 106 or other users on the bed system 100, and apnea of ​​the user 106. One or more other health conditions of the user 106 can also be determined based on information collected by the sensors 105A-N described herein. Taking the detection of the user's presence as an example, a pressure sensor can be used to detect the presence of the user 106 on the bed system 100 (e.g., via gross pressure change determination and / or via one or more of a respiratory rate signal, a heart rate signal and / or other vital signs). Detection of the presence of a user 106 on the bed system 100 may be beneficial for the computer system 102 to determine sleep fragmentation information and / or information regarding the overall sleep quality and health of the user 106.

[0037] The user device 112 may be any suitable type of mobile device of the user 106, including, but not limited to, a mobile phone, a smartphone, a laptop, a tablet, a wearable device, or other computing system. In some implementations, the user device 112 may collect user data that may be used in the processes described further below. The user device 112 may also present a mobile application from which the user 106 may view insights regarding his or her own sleep and health (e.g., over time, over a period of time, over the previous night or previous sleep session, over the current night or current sleep session, etc.) as generated / determined by the computer system 102 and based at least in part on data collected by the sensors 105A-N of the bed system 100.

[0038] The computer system 102 may be any suitable type of computing system, computer, network of computers, network of devices, computing device(s), cloud-based server and / or cloud-based system. In some implementations, the computer system 102 may be remote from the bed system 100 and configured to generate sleep and / or health insights for multiple users of many bed systems. In some implementations, the computer system 102 may be part of the bed system 100 and may be a controller configured to generate sleep and / or health insights for users 106 of the bed system 100.

[0039] The data store 114 may be a database, cloud-based storage, or other type of repository for storing information / data. The data store 114 may be part of the computer system 102. The data store 114 may also be separate from and communicatively coupled to the computer system 102. The data store 114 may store historical data about the user 106 and other users of other bed systems. The historical data may be used for model training and / or run-time determination of CHF risk in users of the bed system. The data store 114 may also store training data and machine learning training models used by the computer system 102 to generate probability measures of CHF risk in users of the bed system.

[0040] Still referring to FIG. 1, the bed system 100 may collect user data in block A. More specifically, the sensors 105A-N may collect data such as temperature, pressure, heart rate, breathing rate, movement, snoring, sleep apnea, presence in bed, etc. The sensors 105A-N may also collect ballistocardiogram (BCG) signals of the user 106 (e.g., load cells may collect fine-grained weight data indicative of the user's 106 cardiac data, and pressure sensors may collect pressure changes in the air chamber of the bed system 100 that may indicate the user's 106 biometric data). The user data may be collected continuously throughout the user's 106 sleep session (e.g., during the night). This data may be collected during one or more predefined time periods (e.g., when the user 106 is detected to enter the bed system 100, when the user 106 is expected to enter the bed system 100, during a particular portion of the user's 106 sleep session, etc.). In some implementations, the user device 112 may collect user data at Block A. For example, the user 106 may input data into a mobile application presented on the user device 112, such as the user's perceived alertness level(s), the user's perceived quality of sleep, the time the user goes to bed, the time the user wakes up or will wake up, and / or biometric data, which may be collected by the wearable device and communicated to the user device 112.

[0041] In block B, user data collected by the bed system 100 and / or the user device 112 may be transmitted to the computer system 102. The data may be transmitted in real time as detected. The data may also be transmitted in batches and / or at predetermined times during or after the user's 106 sleep session. In some implementations, the data may be transmitted at predetermined times, such as once a day, twice a day, once when the user 106 wakes up, once when the user 106 is detected to have left the bed system 100, or when the user 106 is not detected to be in the bed system 100. In some implementations, the data may be transmitted to the data store 114 for storage in block B. The data may later be retrieved by the computer system 102 for use in generating a probability measure of the user's 106 risk of CHF over time.

[0042] The computer system 102 may determine sleep fragmentation information for the user based on the user data (Block C). The sleep fragmentation information may be determined based on various data, such as data 110. The data 110 may be user data collected by the bed system 100 and / or the user device 112 in Block A. Part of the data 110 may be historical user data (e.g., data specific to the user 106 and / or data general to a population of users) stored in the data store 114. Furthermore, part of the data 110 may be determined / generated by the computer system 102 based on other data collected by the bed system 100 and / or the user device 112. As an example, heart rate, sleep interruptions and / or daytime alertness levels may be determined by the computer system 102 based on pressure signals detected by the sensors 105A-N. One or more other data may also be determined by the computer system 102 based on data collected by the sensors 105A-N of the bed system 100 and / or the user device 112.

[0043] The data 110 used in determining the sleep fragmentation information may include, but is not limited to, sleep detection, wake detection, wake after sleep onset, wake before sleep onset, sleep efficiency, sleep duration, sleep episode length, time to sleep onset, wake reactivity index / sleep interruption, heart rate, heart rate variability, respiration rate, daytime alertness level(s), demographic factors, blood pressure, and / or heart beat interval (IBI). Sleep duration and wake duration may be detected and used to determine the fragmentation of the user's 106 sleep session. The sleep duration metric may also be information of sleep fragmentation as it may indicate how long the user 106 actually sleeps and stays asleep during a sleep cycle together with the wake duration metric. A higher sleep duration metric may allow more sleep stages to occur and thus support the physical and mental development of the user 106. Furthermore, the sleep efficiency metric may also be information of sleep fragmentation as it may indicate the ratio of the length of time the user 106 actually sleeps to the length of time the user 106 is in bed. Higher sleep efficiency may allow the user 106 to experience restful sleep, which may enhance overall health and well-being. Sleep timing metrics may also be informative of sleep fragmentation, as consistent sleep timing may improve the user 106's mood and / or alertness throughout the day.

[0044] Additionally, pre-sleep arousal may be the same or similar measurement as time to sleep onset. The longer it takes the user 106 to fall asleep, the more likely the user 106 will experience sleep fragmentation. The arousal reactivity index may indicate the number of interruptions throughout the user's 106 sleep cycle and / or sleep session. The more interruptions, the more fragmented the user's 106 sleep may be. Heart rate and / or heart rate variability may indicate whether the user 106 is experiencing a sleep disorder. For example, a heart rate significantly higher than the user average may indicate the user 106 is experiencing light sleep (or, for example, increased sleep fragmentation). A higher respiration rate may also indicate interruptions to the user's 106 sleep. A low level of daytime arousal may indicate interruptions or sleep fragmentation of the user 106 sleep. Additionally, demographic factors such as weight, age, sex, and / or body mass index (BMI) may also indicate whether the user 106 is experiencing sleep interruptions / sleep fragmentation (e.g., compared to other users with similar demographic factors). Average sleep episode length data may indicate the user's dynamic transitions between sleep and wakefulness throughout a sleep session / night. Similarly, average sleep episode length data may also be used with the techniques of this disclosure. Sleep episode length data may include the time during which the user 106 experiences uninterrupted sleep. The sleep episode length may be considered as the portion of uninterrupted sleep within a sleep session. As an example, a user may sleep through sleep stages N1, N1, and N2 (where N2 is a deeper sleep state than N1, but not as deep as N3), then wake up, then sleep through sleep stages N1, N1, and N2, then experience a second, and possibly a third, wakefulness stage. Each sleep episode length through the sleep stages may vary in time length and may include, but is not limited to, 30 seconds, 1 minute, 1.5 minutes, 2 minutes, etc. On average, for example, a user may experience approximately 25 minutes of sleep without interruption. For example, a user who experiences approximately 5 minutes of sleep before an interruption may have sleep fragmentation. Similarly, if the blood pressure of the user 106 is detected not within an expected threshold range / amount, the user 106 may be experiencing sleep fragmentation and may have a problem with the user's cardiovascular system.

[0045] In block D, the computer system 102 may provide the sleep fragmentation information as input to a model. The model may be a model trained using machine learning techniques, as described with reference to FIG. 2. In some implementations, the sleep fragmentation information may be provided to a machine learning training algorithm. The model (and / or algorithm) may be trained to take into account all of the sleep fragmentation information described above to determine a probability measure of CHF risk for a user of the bed system.

[0046] Furthermore, the model can be trained as a population-based model to determine a probability measure of CHF risk in multiple users over time. The population-based model can be trained using data collected for a variety of different users. The population-based model can be applied to all users of the bed system and / or to a subset or group of users in the population (e.g., users with similar demographic factors such as weight, BMI, age, and / or gender). The population-based model can be used to determine a probability measure of CHF risk in a user 106 over a long / extended period of time. In some implementations, the population-based model can also be used to determine a probability measure of CHF risk in a user 106 during a particular sleep session (e.g., during a given night).

[0047] The model may also be trained as an individualized model for a particular user 106 and thus may be trained using user data collected for the particular user 106. The individualized model may be used to determine a probability measure of CHF risk for the user 106 over an extended period of time, during a particular sleep session (e.g., night by night) and / or over time.

[0048] The computer system 102 may receive an output from the model indicative of a probability measure of risk of heart failure (e.g., CHF) for the user 106 of the bed system 100 (Block E). As described herein, the output may include a numeric value indicative of the probability measure of CHF risk. The numeric value may be a dynamic value along a continuum, such as 0-100, where 0 may be the lowest probability measure of CHF risk and 100 may be the highest probability measure of CHF risk. In some implementations, the probability measure may be a string value, such as low risk, medium risk, and high risk.

[0049] The output may be provided at the end of a sleep session of the user 106. For example, the user 106 may go to bed and blocks A-G of FIG. 1 may be executed each night to determine the real-time CHF risk of the user 106. In some implementations, the output may also be provided after a predetermined number of sleep sessions and / or at predetermined intervals over an extended period of time (e.g., once a week, twice a month, once a month, etc.).

[0050] Optionally, the computer system 102 may output the probability measurement and / or perform the action(s) based on the probability measurement in block F. Optionally, the user device 112 may also perform block F. The computer system 102 may output the probability measurement and / or output a notification / message / alert regarding the probability measurement if the probability measurement meets or exceeds some threshold range or threshold. In some implementations, the computer system 102 may not output the probability measurement and / or the notification / message / alert unless the probability measurement meets or exceeds a threshold range or threshold. Thus, the computer system 102 may output the probability measurement (only) if the user is at high risk for CHF in real-time (e.g., acute CHF) or over a period of time (e.g., chronic CHF).

[0051] Optionally, the computer system 102 may output a monthly report indicating the average CHF risk of the user 106 to the user device 112 and / or to the computing device of the associated healthcare provider. The computer system 102 may also generate one or more other reports regarding the probability measures of CHF risk for the user 106, which may be transmitted to and presented on the user device 112 and / or to the computing device of the associated healthcare provider. Additionally, the computer system 102 may generate one or more behavioral interventions to help reduce sleep fragmentation for a particular user. The computer system 102 may generate an intervention to promote regularity and / or circadian alignment. Such an intervention may include a recommendation of a time window for the user to go to bed. As an example, the recommendation may include, but is not limited to, a recommendation that the user go to bed within a 30 minute time window, preferably between 10:00 pm and 11:00 pm each night. The computer system 102 may also generate interventions such as cognitive behavioral therapy programs. Therein, stimulation control and bed restraint may be recommended to the user for one or more time periods (e.g., 2 weeks, 3 weeks, 6 weeks, 8 weeks, etc.) One or more other interventions may also be possible.

[0052] Optionally, in block F, the computer system 102 may send the probability measurement and / or a notification / message / alert to a computing device of an associated healthcare provider, such as a cardiologist, the user's 106 doctor, a hospital, a nurse, etc. In some implementations, for example, the computer system 102 may determine whether the probability measurement should be shared with the associated healthcare provider. The computer system 102 may evaluate whether the probability measurement exceeds some threshold / range. Based on how much the probability measurement exceeds the threshold / range, the computer system 102 may determine whether to contact emergency healthcare services, such as an emergency room, hospital, ambulance, first responders, etc.

[0053] The computer system 102 may store the probability measure for the user 106 in the data store 114 (block G). The probability measure may then be used by the computer system 102 to quantify CHF risk for the user 106 over time. The probability measure may also be used by the computer system 102 to train a model and / or improve a model over time to more accurately determine the probability measure of CHF risk in the user 106 and / or other users.

[0054] 2 is a conceptual diagram for training a model for determining a probability measure of congestive heart failure in a user of a bed system. The training may be performed by the computer system 102 described with reference to FIG. 1. In some implementations, the training may be performed by another computing system. The training of the model may be performed at various times, for example, once a year, twice a year, once a month, etc. In some implementations, the model may be trained and improved continuously (or may be improved at one or more predetermined times).

[0055] The training techniques described herein may be used to train models that determine probability measures of CHF risk in a particular user. These techniques may also be used to train models that determine probability measures of CHF risk in multiple users or in a general population of users. In some implementations, these techniques may be used to train models that determine probability measures of CHF risk in a particular user or in a general population of users over time. These techniques may also be used to train models that determine probability measures of CHF risk in a particular user or in a general population of users for a given night (e.g., sleep session).

[0056] Although the training in FIG. 2 is described with respect to training one model, the training techniques described herein may also be performed to train two or more models. For example, the techniques may be used to train models for various groupings of users. The groupings may be defined, for example, based on demographic factors (e.g., age, sex, BMI, weight, geographic location, etc.) or non-demographic factors. The groupings may also vary based on the type of data collected and / or analyzed for the users. The groupings may also vary based on known health and / or sleep conditions of the users (e.g., sleep apnea, sleepwalking, asthma, snoring, etc.).

[0057] Further, the model may be trained using machine learning techniques. Machine learning techniques may include, but are not limited to, logistic regression, linear regression, deep neural networks, and decision trees. Logistic regression is a process that models the probability of a discrete outcome given one or more input variables. A logistic regression model may output binary values, such as 0 or 1, true or false, and / or yes or no. In such a scenario, the computer system 102 may normalize the output from the regression model by multiplying the output by a predetermined coefficient to generate a probability measure of CHF risk. Deep neural networks may be used to identify underlying relationships within a dataset. Such networks may adapt to changing inputs and generate the best, preferred, or optimal output. Exemplary deep neural networks that may be used with the training techniques of the present disclosure may include, but are not limited to, a multiplayer perception (MLP) model, a convolutional neural network (CNN) model, and / or a recurrent neural network (RNN) model. A decision tree is a form of supervised machine learning in which data is successively split according to one or more defined parameters. A tree may include decision nodes and leaves, where the leaves are decisions or final results and the decision nodes are where the data is split. One or more types of decision trees may be used, including but not limited to classification trees and regression trees. Although a variety of machine learning techniques may be used, the training technique of FIG. 2 is described with respect to logistic regression modeling, which involves the use of feature vectors.

[0058] Referring to Fig. 2, in block A, the computer system 102 may receive training data from the data store 114. In some implementations, the computer system 102 may receive user data collected by the bed system directly from the bed system, as described in blocks A-B of Fig. 1. The training data may include historical user data 200 and / or historical population data 202. The historical user data 200 and historical population data 202 may be collected over time by sensors of the bed system and / or the user device. Furthermore, the data 200, 202 may include probability measures of CHF risk previously determined by the computer system 102. The historical user data 200 and / or historical population data 202 may include any of the data described throughout this disclosure, including, but not limited to, sleep detection, wake detection, wake after sleep onset, wake before sleep onset, sleep efficiency, sleep duration, sleep episode length, time to sleep onset, wake response index / sleep interruptions, HR, HRV, RR, daytime alertness level(s), demographic factors, blood pressure, IBI, and / or other BCG signals.

[0059] In some implementations, some of the training data received in block A may be annotated and labeled (e.g., by a user and / or by a computing system). For example, low daytime alertness levels may be labeled as indicative of a certain level of sleep fragmentation. As another example, certain blood pressure data may be labeled as indicative of a certain level of sleep fragmentation based on the blood pressure data falling within and / or exceeding a threshold value. Additionally or alternatively, combinations of training data may be labeled and annotated as indicative of various levels of sleep fragmentation.

[0060] For training purposes, the computer system 102 may assign a numerical value to the feature vector corresponding to the training data in block B. The feature vector may be used to represent a numerical or symbolic characteristic or feature of a subject. In the context of CHF risk, each feature vector may represent a different metric used to quantify sleep fragmentation and therefore CHF risk. As an example, one feature vector may be assigned a value of the user's heart rate over time, another may be assigned a value of the time it takes the user to fall asleep, and yet another may be assigned a value of the number of interruptions the user experiences during a sleep session. As another example, the feature vector may be assigned values ​​including, but not limited to, total wakes after sleep onset (min), percentage of sleep efficiency, total sleep duration (min), total sleep episode length (min), and cardiovascular values ​​such as heart rate, such as beats per minute, and respiration rate, such as breaths per minute. In some implementations, the values ​​of the feature vector may be mapped to an n-dimensional space where each dimension may represent a different characteristic. Clusters of values ​​may be identified as indices of sleep fragmentation. A model may then be trained to correlate such indices of sleep fragmentation with CHF risk.

[0061] In block C, the computer system 102 may train a model to determine a probability measure of risk of heart failure (e.g., CHF risk) based on the feature vector. As previously described, the computer system 102 may train a model to correlate clusters of sleep fragmentation indicators (or feature vector values) with CHF risk.

[0062] The output from the logistic or linear regression model may include a binary value, such as 0 or 1, where 0 may indicate no CHF risk and 1 may indicate CHF risk. In some implementations, the model output may be a real value between 0 and 1 instead of a binary value. The computer system 102 may multiply the output by a numeric factor to generate an integer value that quantifies the CHF risk. For example, the computer system 102 may multiply the output by a factor of "100" to quantify the CHF risk as a probability measure. The computer system 102 may also multiply the output by one or more other factors to quantify the CHF risk as a probability measure.

[0063] The output from the deep neural network may include integer values, such as continuous values ​​or values ​​on a scale such as 0 to 100. The output from the decision tree may include binary values ​​(e.g., numbers, strings, logical values) and may be multiplied by a coefficient to quantify CHF risk as a probability measure. In some implementations, the output from the decision tree may also include a probability measure of CHF risk.

[0064] Next, in block D, the computer system 102 may output the model. Outputting the model may include storing the model in the data store 114 for future retrieval and use during runtime. Outputting the model may also include storing the model in local memory so that the model may be used during runtime for real-time CHF risk determination.

[0065] In some implementations, the model can be sent to a controller of a bed system, such as the bed system 100 of FIG. 1. The controller of the bed system can then be configured to collect user data collected by sensors of the bed system, provide the collected user data to the model to determine a probability measurement of CHF risk for the user of the bed system, and so on. Thus, the controller of the bed system, rather than a remote computer such as the computer system 102, can determine a probability measurement of CHF risk for the user. In some other implementations, the model can be sent to a device of the user. Then, instead of the computer system 102, the device of the user can receive the user data and provide the user data as an input to the model to determine a probability measurement of CHF risk for the user.

[0066] 3 is a flow chart of a process 300 for determining a probability measure of congestive heart failure in a user of a bed system. The process 300 may be performed during or after a sleep session of the user. The process 300 may be performed during and / or after each sleep session of the user. The process 300 may also be performed at predetermined times and / or over an extended period of time (e.g., every other sleep session, several times a week, once a day, etc.). In this way, the process 300 may be performed to determine a probability measure of a patient's CHF risk not only for a given / particular night, but also over an extended period of time.

[0067] Process 300 may be performed by computer system 102. Process 300 may also be performed by one or more other computing systems, devices, computers, networks, cloud-based systems, cloud-based services, bed controllers, and / or user devices. For purposes of explanation, process 300 is described in terms of a computer system.

[0068] Referring to the process 300, in block 302, the computer system may receive user data. As described herein, the computer system may receive user data collected by a sensor of the bed system when a user rests on the bed system. For further description of user data and the reception of the user data, please refer to blocks A-B of FIG. 1.

[0069] At block 304, the computer system may determine sleep fragmentation information for the user's sleep session based on an analysis of the data. This may include identifying periods during which the user is asleep during the sleep session at block 306. Additionally or alternatively, this may include identifying periods during which the user is awake during the sleep session at block 308. Thus, the computer system may determine the sleep fragmentation information as a function of sensed user data from at least one sensor of the user on the bed system.

[0070] The computer system may process user data, such as pressure data, to determine when the user enters the bed system and when the user falls asleep. Periods during which the user is asleep may be detected from pressure data indicative of changes in heart rate or other BCG signals. For example, if the user's heart rate drops below a threshold amount, the computer system may determine that the user has fallen asleep. Similarly, the computer system may process pressure data to determine when the user is likely snoring. This may be further used to identify periods during which the user is asleep. One or more other data received in block 302 may be processed to determine sleep fragmentation information regarding the user's sleep session. For further description regarding determining sleep fragmentation information, see block C of FIG. 1.

[0071] In block 310, the computer system may determine additional information regarding the user's sleep session. The additional information may include, but is not limited to, wake-after-sleep-onset data (block 312), sleep efficiency data (block 314), sleep duration data (block 316), and sleep episode length data (block 318). For example, the computer system may determine wake-after-sleep-onset data as a function of sensed user data of the user from at least one sensor of the bed (block 312). The computer system may also determine sleep efficiency data as a function of the sensed user data (block 314). The computer system may also determine sleep duration data as a function of the sensed user data (block 316). The computer system may determine sleep episode length data as a function of the sensed user data (block 318). In some implementations, the computer system may determine less than all of the data in blocks 312-318. As an example, the computer system may determine only wake-after-sleep-onset data, sleep efficiency data, and sleep episode length data. Any other combination of the data in blocks 312-318 may be determined and used by the computer system to determine a probability measure of CHF risk for the user.

[0072] One or more of blocks 312-318 may be determined based on processing the data received in block 302. Additionally or alternatively, one or more of blocks 312-318 may be determined based on the sleep fragmentation information determined in blocks 304-308. In some implementations, the data in blocks 312-318 may be used to determine the sleep fragmentation information in blocks 304-308. Thus, block 310 may be performed prior to block 304 and / or as part of determining the sleep fragmentation information in blocks 304-308.

[0073] Next, in block 320, the computer system may provide the sleep fragmentation information and / or the additional information as input to a model that determines a probability measure of a user's risk of CHF. The model may be trained to determine a probability measure of a risk of heart failure (e.g., risk of CHF) for a user of the bed system based at least in part on user data associated with the user. In some implementations, the model may be trained to correlate at least one of the additional data from blocks 310-318 with the sleep fragmentation information from blocks 304-308 to quantify a probability measure of CHF risk. The model may also calculate a probability measure as a function of the determined sleep fragmentation information. In some cases, the model may calculate a probability measure of CHF risk as a function of the determined wake-after-sleep-onset data, sleep efficiency data, and sleep episode length data. The model may also be trained to calculate a probability measure as a function of any combination of the additional data determined in blocks 312-318. The model may be trained using the techniques described in FIG. 2.

[0074] In some implementations, block 320 may include selecting, by the computer system, an appropriate model from a data store (e.g., data store 114) of existing models for determining a probability measure of CHF risk for a user. The computer system may select a model based on one or more factors, including, but not limited to, user information, demographic factors, whether the computer system is determining a probability measure for a current or given sleep session / night, whether the computer system is determining a long-term probability measure, etc. Thus, multiple models for determining a probability measure of CHF risk may be generated and trained and stored in a data store for run-time selection, retrieval, and use. In some implementations, a population-based model may be trained and used to determine a probability measure of CHF risk in a variety of different users, as described with reference to FIG. 1. See block D of FIG. 1 for further description regarding using a model to determine a probability measure of CHF risk.

[0075] In block 322, the computer system may receive a probability measure as a model output. As described with reference to FIGS. 1-2, the model output may be a numerical value indicating the probability that the user has a risk of CHF. The model output may also be a binary value, such as 0-1, particularly if the model is a logistic model or a linear regression model. The computer system may then normalize or adjust the binary value by multiplying it by a factor, such as "100", to quantify the CHF risk as a probability measure. In some implementations, the model output may be a logical value, such as true / false. The model output may also be a string value, such as at risk for CHF / not at risk for CHF. For example, the model output may include low risk, medium risk, and high risk. One or more other values ​​may be generated and output by the model. See block E of FIG. 1 and block C of FIG. 2 for further discussion of the model output.

[0076] At block 324, the computer system may selectively perform an action based on the probability measure. In other words, the computer system may output a signal to perform an action based on the calculated probability measure of CHF risk for the user. At block 326, the computer system may output an alert at the user device of the user. For example, if the probability measure exceeds some threshold (e.g., the probability measure is greater than 90 on a scale of 0 to 100), the computer system may generate an alert informing the user that they are at high risk for CHF. In some implementations, the computer system may only output an alert or some other notification / message at the user device of the user if the probability measure exceeds some threshold. As described with reference to FIG. 1, the alert or other notification / message may be presented in a mobile application at the user device.

[0077] The computer system may also output a notification at the user device of an associated healthcare provider in block 328. For example, if the probability measure exceeds some threshold, the computer system may generate the notification, which may also be an alert notifying the associated healthcare provider of the user's CHF risk. In some implementations, the computer system may optionally generate an alert or other form of contact to emergency medical services, such as an ambulance, first responders, and / or an emergency room, if the probability measure exceeds another threshold (e.g., the probability measure is greater than 95 on a scale of 0 to 100).

[0078] At block 324, the computer system may optionally perform one or more other actions. For example, the computer system may store the probability measures in a data store. The probability measures may then be used for future training and / or continuous improvement of one or more models. The computer system may also generate reports regarding the probability measures and / or the user's probability measures over one or more time periods. Such reports may be generated on a daily, weekly, monthly, yearly, etc. basis. Similarly, the probability measures may be used by the computer system to analyze the user's health and cardiovascular activity over time and generate various other reports and CHF risk assessments for the user. See blocks F-G of FIG. 1 for further explanation.

[0079] 4 is a system diagram of one or more hardware and software components that may be used to implement the techniques of the present disclosure. As described throughout this disclosure, the computer system 102, the user devices 112A-N, the bed system 100, and the data store 114 can communicate over the network(s) 108.

[0080] The bed system 100 may include a controller 402, sensors 105A-N, and a communication interface 406. For further description of the bed system 100 and the sensors 105A-N, please refer to FIG. 1. Furthermore, the controller 402 may be configured to control one or more components of the bed system 100. In some implementations, the controller 402 may include a processor capable of executing instructions that cause the controller 402 to perform one or more of the techniques described herein. For example, the controller 402 may perform one or more functions of the computer system 102, such as determining sleep fragmentation information for a user of the bed system 100 and determining a probability measure of CHF risk for the user. Additionally, the communication interface 406 may provide communication between the components of the bed system 100 and one or more other components described herein.

[0081] The computer system 102 may include a processor(s) 408, a training engine 410, a sleep fragmentation determiner 412, a CHF risk determiner 414, an output generator 416, and a communication interface 418. The processor(s) 408 may be configured to perform one or more of the operations described throughout this disclosure. The training engine 410 may be configured to train a model that uses machine learning techniques to determine a probability measure of CHF risk in a user of the bed system. For further description of training operations that may be performed by the training engine 410, see FIG. 2. Briefly, for example, the training engine 410 may retrieve (obtain) historical user data 200A-N and / or historical population data 202A-N from the data store 114. The training engine 410 may use the retrieved data to train a model, which may be stored as a model 420A-N in the data store 114.

[0082] During runtime, the sleep fragmentation determiner 412 may receive user data collected by the sensors 105A-N of the bed system 100 during the user's sleep session. The determiner 412 may process this information and determine sleep fragmentation information as described with reference to Figures 1 and 3.

[0083] The CHF risk determiner 414 may be configured to retrieve models 420A-N from the data store 114. User data collected by the sensors 105A-N and / or sleep fragmentation information determined by the sleep fragmentation determiner 412 may be provided as inputs to the retrieved models 420A-N in the CHF risk determiner 414 to generate an output indicative of a probability measure of CHF risk for a user of the bed system 100. The probability measure of CHF risk for the user may then be stored in the data store 114 as a probability measure 422A-N.

[0084] The output generator 416 may be configured to generate an output based on the probability measure of CHF risk for the user. For further description of outputs that may be generated, see Figures 1 and 3. The output may be sent to a user device 112A-N of a user of the bed system 100. As described herein, the output may also be transmitted to a user device 112A-N of a healthcare provider. Finally, the communication interface 418 may provide communication between the computer system 102 and one or more other components described herein.

[0085] 5 illustrates an example of a computing device 500 and an example of a mobile computing device that may be used to implement the techniques described herein. Computing device 500 is intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other suitable computers. Mobile computing device is intended to represent various forms of mobile devices, such as personal digital assistants, mobile phones, smart phones, and other similar computing devices. The components shown, their connections and relationships, and their functions are intended to be illustrative only and are not intended to limit the implementation of the invention described and / or claimed herein.

[0086] The computing device 500 includes a processor 502, a memory 504, a storage device 506, a high-speed interface 508 connecting to the memory 504 and multiple high-speed expansion ports 510, and a low-speed interface 512 connecting to the low-speed expansion port 514 and the storage device 506. Each of the processor 502, the memory 504, the storage device 506, the high-speed interface 508, the high-speed expansion port 510, and the low-speed interface 512 are interconnected using various buses and may be mounted on a common motherboard or in other manners as desired. The processor 502 may process instructions for execution within the computing device 500, including instructions stored in the memory 504 or on the storage device 506, and may display graphical information for a GUI on an external input / output device, such as a display 516 coupled to the high-speed interface 508. In other implementations, multiple processors and / or multiple buses may be used, along with multiple memories and types of memories, as desired. Additionally, multiple computing devices may be connected (eg, as a server bank, as a collection of blade servers, or as a multi-processor system) such that each computing device provides a portion of the required operations.

[0087] The memory 504 stores information within the computing device 500. In some implementations, the memory 504 is one or more volatile memory units. In some implementations, the memory 504 is one or more non-volatile memory units. The memory 504 may be another form of computer-readable medium, such as a magnetic disk or an optical disk.

[0088] The storage device 506 can provide mass storage for the computing device 500. In some implementations, the storage device 506 can be or include a computer-readable medium, such as a floppy disk drive, a hard disk drive, an optical disk drive, a tape drive, a flash memory, or other similar solid-state memory device, or an arrangement of devices, including a storage area network or other form of device. The computer program product can be tangibly embodied in an information carrier. The computer program product can also include instructions that, when executed, perform one or more methods, such as the methods described above. The computer program product can also be tangibly embodied in a computer-readable or machine-readable medium, such as the memory 504, the storage device 506, or a memory on the processor 502.

[0089] The high-speed interface 508 manages bandwidth-intensive operations for the computing device 500, and the low-speed interface 512 manages lower bandwidth-intensive operations. This allocation of functionality is merely exemplary. In some implementations, the high-speed interface 508 is coupled to the memory 504, the display 516 (e.g., via a graphics processor or accelerator), and a high-speed expansion port 510 that can accept various expansion cards (not shown). In such implementations, the low-speed interface 512 is coupled to the storage device 506 and the low-speed expansion port 514. The low-speed expansion port 514 may include various communication ports (e.g., USB, Bluetooth, Ethernet, wireless Ethernet) and may be coupled to one or more input / output devices, such as a keyboard, a pointing device, a scanner, or a network device, such as a switch or a router, for example, via a network adapter.

[0090] Computing device 500 may be implemented in a number of different forms, as shown in the figure. For example, it may be implemented as a standard server 520 or multiple times in a group of such servers. It may also be implemented in a personal computer, such as a laptop computer 522. It may also be implemented as part of a rack server system 524. Alternatively, components from computing device 500 may be combined with other components in a mobile device (not shown), such as mobile computing device 550. Each such device may include one or more of computing device 500 and mobile computing device 550, and the entire system may be made up of multiple computing devices in communication with each other.

[0091] The mobile computing device 550 includes, among other things, a processor 552, a memory 564, input / output devices such as a display 554, a communication interface 566, and a transceiver 568. The mobile computing device 550 may also be provided with a storage device such as a microdrive or other device to provide additional storage. Each of the processor 552, memory 564, display 554, communication interface 566, and transceiver 568 are interconnected using various buses, and some of the components may be mounted on a common motherboard or in other manners as desired.

[0092] The processor 552 may execute instructions within the mobile computing device 550, including instructions stored in the memory 564. The processor 552 may be implemented as a chipset of chips including separate analog and digital processors. The processor 552 may provide for coordination of other components of the mobile computing device 550, such as control of a user interface, applications run by the mobile computing device 550, and wireless communications by the mobile computing device 550.

[0093] The processor 552 may communicate with a user via a control interface 558 and a display interface 556 coupled to a display 554. The display 554 may be, for example, a TFT display (thin film transistor liquid crystal display), an OLED (organic light emitting diode) display, or other suitable display technology. The display interface 556 may have appropriate circuitry for driving the display 554 to present graphical and other information to the user. The control interface 558 may receive commands from the user and convert them for presentation to the processor 552. Additionally, an external interface 562 may provide communication with the processor 552 to enable short-range communication with other devices in the mobile computing device 550. The external interface 562 may provide, for example, wired communication in some implementations or wireless communication in other implementations, and multiple interfaces may be used.

[0094] The memory 564 stores information within the mobile computing device 550. The memory 564 may be implemented as one or more computer readable media, one or more volatile memory units, or one or more non-volatile memory units. An expansion memory 574 may also be provided and connected to the mobile computing device 550 via an expansion interface 572, which may include, for example, a SIMM (single in-line memory module) card interface. The expansion memory 574 may provide additional storage space for the mobile computing device 550 or may store applications or other information for the mobile computing device 550. In particular, the expansion memory 574 may include instructions for performing or supplementing the aforementioned processes, and may also include security information. Thus, for example, the expansion memory 574 may be provided as a security module for the mobile computing device 550 and may be programmed with instructions that allow secure use of the mobile computing device 550. Furthermore, secure applications may be provided via a SIMM card along with additional information, such as placing identification information on the SIMM card in an unhackable manner.

[0095] The memory may include, for example, flash memory and / or NVRAM memory (non-volatile random access memory), as described below. In some implementations, a computer program product is tangibly embodied in an information carrier. The computer program product includes instructions that, when executed, perform one or more methods, such as the methods described above. The computer program product may be a computer-readable or machine-readable medium, such as memory 564, expansion memory 574, or memory on processor 552. In some implementations, the computer program product may be received in a propagated signal, for example, via transceiver 568 or external interface 562.

[0096] The mobile computing device 550 may communicate wirelessly via a communication interface 566, which may include digital signal processing circuitry, if desired, and may provide communications under various modes or protocols, such as GSM voice (Global System for Mobile Communications), SMS (Short Message Service), EMS (Enhanced Messaging Service), MMS messaging (Multimedia Messaging Service), CDMA (Code Division Multiple Access), TDMA (Time Division Multiple Access), PDC (Personal Digital Cellular), WCDMA (Wideband Code Division Multiple Access), CDMA2000, or GPRS (General Packet Radio Service), among others. Such communications may occur, for example, via transceiver 568 using radio frequencies. Additionally, short-range communications may occur, such as using Bluetooth, WiFi, or other such transceivers (not shown). Additionally, a GPS (Global Positioning System) receiver module 570 may provide additional navigation and location related wireless data to the mobile computing device 550. It may be suitably used by applications running on the mobile computing device 550 .

[0097] Mobile computing device 550 may also communicate audibly using audio codec 560. Audio codec 560 may receive spoken information from a user and convert it into usable digital information. Similarly, audio codec 560 may generate sounds audible to the user, such as through a speaker in a handset of mobile computing device 550. Such sounds may include sounds from voice calls, recorded sounds (e.g., voice messages, music files, etc.), and may also include sounds generated by applications running on mobile computing device 550.

[0098] The mobile computing device 550 may be implemented in a number of different forms, as shown in the figure, for example, it may be implemented as a mobile phone 580, or it may be implemented as part of a smartphone 582, personal digital assistant, or other similar mobile device.

[0099] Various implementations of the systems and techniques described herein may be realized in digital electronic circuitry, integrated circuits, specially designed ASICs (application-specific integrated circuits), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include implementations in one or more computer programs executable and / or interpretable on a programmable system that includes at least one programmable processor, which may be coupled to receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device, for a special purpose or general purpose.

[0100] These computer programs (also referred to as programs, software, software applications, or code) include machine instructions for a programmable processor and may be implemented in a high level procedural and / or object-oriented programming language and / or in assembly / machine language. As used herein, the terms machine-readable medium and computer-readable medium refer to any computer program product, apparatus, and / or device (e.g., magnetic disks, optical disks, memory, programmable logic devices (PLDs)) used to provide machine instructions and / or data to a programmable processor. This includes machine-readable media that receive machine instructions as a machine-readable signal. The term machine-readable signal refers to any signal used to provide machine instructions and / or data to a programmable processor.

[0101] To provide interaction with a user, the systems and techniques described herein may be implemented on a computer that has a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user, and a keyboard and pointing device (e.g., a mouse or trackball) by which the user can provide input to the computer. Other types of devices may also be used to provide interaction with a user. For example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback). Input from the user may be received in any form, including acoustic, speech, or tactile input.

[0102] The systems and techniques described herein may be implemented within a computing system that includes back-end components (e.g., a data server), or includes middleware components (e.g., an application server), or includes front-end components (e.g., a client computer with a graphical user interface or web browser through which a user can interact with an implementation of the systems and techniques described herein), or includes any combination of such back-end, middleware, and / or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communications network). Examples of communications networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0103] A computing system may include clients and servers. Clients and servers are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other.

[0104] Although the specification contains many specific implementation details, these should not be construed as limitations on the scope of the disclosed technology or as it may be claimed. Rather, these should be construed as descriptions of features that may be specific to certain embodiments of the disclosed technology. Certain features described herein in the context of separate embodiments may also be implemented in combination, either partially or in whole, in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented in multiple embodiments individually or in any suitable subcombination. Furthermore, although features may be described herein as acting in a particular combination and / or may initially be claimed as such, in some cases one or more features from a claimed combination may be deleted from the combination, or the claimed combination may be directed to a subcombination or subcombination variation. Similarly, although operations may be described in a particular order, this should not be construed as requiring that such operations be performed in a particular order or sequential order to achieve desired results, or that all operations be performed. Specific embodiments of the subject matter have been described. Other embodiments are within the scope of the following claims.

Claims

1. 1. A method for determining a probability measure of risk of heart failure for a user of a bed system, comprising: receiving, by a computer system, user data collected by sensors of the bed system while a user lies on the bed system; determining, by the computer system, sleep fragmentation information for the user's sleep sessions based on the analysis of the user data; providing, by the computer system, the sleep fragmentation information as an input to a model trained to determine a probability measure of risk of heart failure for a user of the bed system based at least in part on user data associated with the user; receiving, by the computer system, as output from the model, the probability measure of the risk of heart failure for the user of the bed system; performing, by the computer system, an action based on the probability measure; A method comprising:

2. The user data includes a ballistocardiogram (BCG) signal.

2. The method of claim 1 .

3. The model is trained with a training dataset of historical user data associated with the user.

2. The method of claim 1 .

4. The model is trained with a training dataset of historical user data associated with various users of the bed system.

2. The method of claim 1 .

5. The model is a logistic regression model.

2. The method of claim 1 .

6. The model is trained by the computer system using feature vectors that are assigned values ​​corresponding to training user data.

2. The method of claim 1 .

7. The training user data includes at least one of: time to wake after sleep onset, sleep efficiency, sleep duration, and sleep episode length.

7. The method of claim 6.

8. The trained user data includes at least one of: time awake before sleep onset, amount of sleep interruption, heart rate, heart rate variability, respiration rate, and daytime alertness level.

7. The method of claim 6.

9. The training user data includes demographic information; The demographic information includes at least one of weight, age, sex, and body mass index (BMI).

7. The method of claim 6.

10. Determining sleep fragmentation information for the user by the computer system includes identifying periods during the sleep session when the user is asleep and periods during the sleep session when the user is awake.

2. The method of claim 1 .

11. Providing, by the computer system, as input to the model, at least one of: (i) wake-after-sleep-onset data; (ii) sleep efficiency data; (iii) sleep duration data; and (iv) sleep episode length data.

2. The method of claim 1 further comprising:

12. The model is trained by the computer system to correlate at least one of (i) to (iv) with a stage of sleep fragmentation of the user of the bed system.

12. The method of claim 11 .

13. The probability measure of risk of heart failure is a numerical value.

2. The method of claim 1 .

14. the output from the model is a value between 0 and 1; The method comprises: generating, by said computer system, a score for said risk of heart failure by multiplying said value by a numerical factor.

2. The method of claim 1 further comprising:

15. the probability measure of risk of heart failure is a categorical value; The categorical value is at least one of: low risk, medium risk, and high risk.

2. The method of claim 1 .

16. Performing an action based on the probability measure by the computer system includes outputting an alert to a user device of the user indicating that the user is at risk for heart failure based on a determination that the probability measure exceeds a threshold range.

2. The method of claim 1 .

17. Performing an action based on the probability measure by the computer system includes sending a notification to a health care provider user device indicating that the user is at risk for heart failure based on a determination that the probability measure exceeds a threshold range.

2. The method of claim 1 .

18. The heart failure is congestive heart failure.

2. The method of claim 1 .

19. 1. A method for determining a probability measure of risk of heart failure for a user of a bed system, comprising: receiving, by a computer system, user data collected by sensors of the bed system while a user lies on the bed system; determining, by the computer system, sleep fragmentation information for the user's sleep sessions based on the analysis of the user data; determining, by the computer system, based on analysis of the user data, at least one of: (i) wake after sleep onset data; (ii) sleep efficiency data; (iii) sleep duration data; and (iv) sleep episode length data; providing, by the computer system, the sleep fragmentation information and at least one of (i)-(iv) as inputs to a model trained to determine a probability measure of risk of heart failure for a user of the bed system based at least in part on user data associated with the user; receiving, by the computer system, as output from the model, the probability measure of the risk of heart failure for the user of the bed system; performing, by the computer system, an action based on the probability measure; A method comprising:

20. 1. A system for determining a probability measure of risk of heart failure for a user of a bed system, comprising: a bed having at least one sensor; a computer system in communication with the at least one sensor of the bed; Equipped with The computer system includes: receiving user data from the at least one sensor, the user data being collected by the at least one sensor while a user is lying on the bed; determining sleep fragmentation information for the user's sleep session based on the analysis of the user data; providing the sleep fragmentation information as an input to a model trained to determine a probability measure of risk of heart failure for a user of a bed based at least in part on user data associated with the user; receiving as output from the model the probability measure of the risk of heart failure for the user of the bed; Perform an action based on the probability measure It is configured as follows: A system characterized by:

21. The bed further includes a controller. The computer system is the controller.

21. The system of claim 20.

22. The computer system is remote from the bed.

21. The system of claim 20.

23. 1. A system for determining a probability measure of risk of congestive heart failure for a user of a bed system, comprising: a bed having at least one sensor; a computer system in communication with the at least one sensor of the bed; Equipped with The computer system includes: determining sleep fragmentation as a function of sensed data from the at least one sensor of a sleeper in the bed; calculating a probability measure of congestive heart failure as a function of said determined sleep fragmentation; and outputting a signal to perform an action based on the calculated probability measure of congestive heart failure. It is configured as follows: A system characterized by:

24. 1. A system for determining a probability measure of risk of heart failure for a user of a bed system, comprising: a bed having at least one sensor; a computer system in communication with the at least one sensor of the bed; Equipped with The computer system includes: receiving user data from the at least one sensor, the user data being collected by the at least one sensor while a user is lying on the bed; determining sleep fragmentation information for the user's sleep session based on the analysis of the user data; determining at least one of: (i) wake after sleep onset data; (ii) sleep efficiency data; (iii) sleep duration data; and (iv) sleep episode length data based on the analysis of the user data; providing the sleep fragmentation information and at least one of (i)-(iv) as inputs to a model trained to determine a probability measure of risk of heart failure for a user of a bed based at least in part on user data associated with the user; receiving as output from the model the probability measure of the risk of heart failure for the user of the bed; Perform an action based on the probability measure It is configured as follows: A system characterized by:

25. 1. A system for determining a probability measure of risk of congestive heart failure for a user of a bed system, comprising: a bed having at least one sensor; a computer system in communication with the at least one sensor of the bed; Equipped with The computer system includes: determining post-sleep onset arousal data as a function of the sleeper's sensed data from the at least one sensor; determining sleep efficiency data as a function of the sensed data from the sleeper; determining sleep episode length data as a function of the sensed data from the sleeper; calculating a probability measure of risk of congestive heart failure as a function of the determined sleep onset wake data, the determined sleep efficiency data, and the determined sleep episode length data; and outputting a signal to perform an action based on the calculated probability measure of congestive heart failure. It is configured as follows: A system characterized by:

Images