Fatigue monitoring and management system

Abstract

The system monitors the fatigue of the user. The system (100) may include one or more data sources, such as inconspicuous sleep sensors, configured to produce objective sleep measurements of a user. The system may also include a fatigue monitoring module that may be configured to generate a user fatigue state assessment based on data from the one or more data sources, such as in one or more processors.

Description

[0001] Divisional application statement

[0002] This application is a divisional application of a PCT international application with an international application date of October 6, 2014 and an international application number of PCT / US2014 / 059311, which entered the Chinese national phase on June 7, 2016, and a Chinese invention patent application with an invention name of “Fatigue Monitoring and Management System” and an application number of 201480066999.8.

[0003] 1 Cross-reference to related applications

[0004] This application claims the benefit of Australian Provisional Application No. 2013903881, filed on October 9, 2013, and Australian Provisional Application No. 2014902350, filed on June 19, 2014, the entire contents of which are incorporated herein by reference.

[0005] 2 STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

[0006] not applicable

[0007] 3 Names of parties involved in the joint research and development

[0008] not applicable

[0009] 4. Sequence Listing

[0010] not applicable 5. Background of the invention 5.1 Technical Field

[0012] The present invention relates to monitoring and management of fatigue. In particular, the present invention relates to systems and methods for monitoring and managing fatigue. 5.2 Background Technology

[0013] The body's respiratory system facilitates gas exchange.

[0014] The nose and mouth form the entrance to the patient's airway. The airway consists of a series of branching tubes that become narrower, shorter, and more numerous as they go deeper into the lungs. The main function of the lungs is gas exchange, allowing oxygen to pass from the air into the venous blood and expel carbon dioxide. The trachea is divided into the right and left main bronchi, which also eventually divide into terminal bronchioles. The bronchi make up the conducting airways and do not participate in gas exchange. Further division of the airway leads to the respiratory bronchi and ultimately the alveoli. The alveolar region of the lungs is where gas exchange occurs and is called the respiratory zone.

[0015] 5.2.1 Respiratory disorders

[0016] There is a range of breathing disorders.

[0017] Obstructive Sleep Apnea (OSA), a form of Sleep Disordered Breathing (SDB), is characterized by occlusion or obstruction of the upper airway during sleep. It is caused by a combination of an abnormally small upper airway and normal loss of muscle tone in the tongue area, soft palate, and posterior oropharyngeal wall during sleep. This condition causes affected patients to pause breathing for cycles of typically 30 to 120 seconds in duration, sometimes 200 to 300 times per night. This often leads to excessive daytime sleepiness and may lead to cardiovascular and brain disease. The syndrome is a common disorder, particularly in middle-aged obese men, although those affected may not be aware of the problem. See U.S. Patent 4,944,310 (Sullivan).

[0018] Cheyne-Stokes Respiration (CSR) is a disorder of a patient's respiratory control system in which a gradual decrease in breathing and a gradual increase in breathing alternate, resulting in repeated deoxygenation and reoxygenation of arterial blood. CSR can cause damage due to repeated hypoxia. In some patients, CSR is associated with repeated arousals from sleep, which cause severe sleep disruption, increased sympathetic nerve activity, and increased afterload. See U.S. Patent 6,532,959 (Berthon-Jones).

[0019] Obesity Hyperventilation Syndrome (OHS) is defined as the combination of severe obesity and chronic hypercapnia during wakefulness in the absence of other known causes of hypoventilation. Symptoms include dyspnea, morning headache, and excessive daytime sleepiness.

[0020] Chronic Obstructive Pulmonary Disease (COPD) encompasses a group of lower airway diseases that share certain common features. These include increased resistance to air flow, prolonged expiratory phase of breathing, and loss of normal elasticity of the lungs. Examples of COPD are emphysema and chronic bronchitis. COPD is caused by long-term smoking (the main risk factor), occupational exposure, air pollution, and genetic factors. Symptoms include: dyspnea on exertion, chronic cough, and sputum production.

[0021] Neuromuscular Disease (NMD) is a broad term that includes many diseases and ailments that compromise muscle function either directly through intrinsic muscle pathology or indirectly through neuropathology. Some NMD patients are characterized by progressive muscle damage leading to loss of mobility, wheelchair confinement, dysphagia, respiratory muscle weakness, and ultimately death from respiratory failure. Neuromuscular disorders can be divided into rapidly progressive and slowly progressive: (i) rapidly progressive disorders are characterized by muscle damage that worsens over months and leads to death within a few years (e.g., amyotrophic lateral sclerosis (ALS) and Duchenne muscular dystrophy (DMD) in adolescents); (ii) variable or slowly progressive disorders are characterized by muscle damage that worsens over many years with only a modest reduction in life expectancy (e.g., limb girdle, facioscapulobrachial, and myotonic muscular dystrophy). Symptoms of respiratory failure in NMD include becoming weak, having difficulty swallowing, dyspnea on exertion, dyspnea at rest, fatigue, drowsiness, morning headaches, difficulty concentrating, and mood changes.

[0022] Chest wall disorders are a group of thoracic deformities that result in an inefficient connection between the respiratory muscles and the thorax. The disorders are often characterized by restrictive defects and the potential for prolonged hypercapnic respiratory failure. Scoliosis and / or kyphoscoliosis may lead to severe respiratory failure. Symptoms of respiratory failure include: dyspnea on exertion, peripheral edema, orthopnea, recurrent chest infections, morning headaches, fatigue, poor sleep quality, and loss of appetite.

[0023] 5.2.2 Treatment

[0024] Continuous Positive Airway Pressure (CPAP) therapy has been used to treat obstructive sleep apnea (OSA). The hypothesis proposed for this is that CPAP acts as a gas splint and may prevent upper airway obstruction by pushing the soft palate and tongue toward or away from the posterior oropharyngeal wall.

[0025] An application that facilitates the delivery of air at a positive pressure to the entrance of a patient's airway through the use of a patient interface such as a nasal mask, full face mask, or nasal pillows.

[0026] Air is usually supplied to the patient's airway at positive pressure by a respiratory pressure therapy (RPT) device, such as a blower driven by an electric motor. The outlet of the RPT device is connected to the patient interface via a flexible conduit.

[0027] 5.2.3 Fatigue

[0028] The Fatigue Guidelines Summary (2013) published by the State of Queensland, Australia, defines fatigue as an impaired state, including physical and / or psychological factors, associated with reduced alertness and decreased performance. There are multiple contributing factors to fatigue, generally involving:

[0029] Poor sleep quality (eg, due to family / entertainment factors, stress, noise, respiratory diseases and other health problems). Poor sleep quality is a significant problem worldwide, affecting up to 60% of the adult population.

[0030] Prolonged wakefulness, defined as the time (in hours) since last sleep.

[0031] One of the main causes of fatigue in the U.S. is self-imposed wakefulness or sleep deprivation. In other words, poor “sleep hygiene” results in insomnia on Sunday nights because people intentionally limit their sleep and / or go to bed at different times during the weekdays and weekends. For example, people may skip sleep in an attempt to get more done and stay connected to their laptops, tablets, or smartphones “24 / 7” (NSF, 2013).

[0032] More generally, fatigue can be considered a complex problem involving multiple factors. Fatigue may be experienced after a short period of sleep loss (known as acute fatigue) due to a loss of sleep quality or quantity, or after a longer period of sleep loss (known as cumulative fatigue) due to the accumulation of sleep interruptions or loss of restorative sleep. For example, mental (cognitive) fatigue can manifest as: impaired alertness, coordination, and decision-making; emotional effects; and micro-sleeps during tasks. Restorative (deep slow wave) sleep is associated with physical rest, i.e., reducing physical fatigue. Cognitive (REM) sleep is associated with mental rest.

[0033] It is estimated that sleep-related fatigue costs businesses $150 billion per year in absenteeism and workplace accidents. It has been noted that error rates increase exponentially and linearly with increasing psychometric measures of fatigue (Dinges et al., 1997). Insomnia has been associated with 7% of all costly workplace accidents and errors and 24% of all workplace disasters (Shahly et al., 2012). Dawson and Reid (1997) equated performance impairment caused by fatigue to that caused by alcohol intoxication and showed that moderate levels of fatigue produced higher levels of impairment than prescribed levels of alcohol intoxication. Poor sleep quality and inadequate recovery lead to increased fatigue, decreased alertness, and impaired performance on various cognitive psychomotor tests (Harrington, 1978; Nicholson & D'Auria, 1999). Barker & Nussbaum's (2010) study of registered nurses reported that mental fatigue levels were higher than physical fatigue levels, and that acute fatigue levels were higher than chronic fatigue levels. All fatigue dimensions and states were negatively correlated with perceived performance. The Swedish Occupational Fatigue Inventory (SOFI) assesses perceived work-related fatigue across four dimensions: mental fatigue, physical exertion, physical discomfort, and lack of motivation. Total fatigue was measured using the mental fatigue dimension, mental fatigue was measured using the lack of motivation dimension, and physical fatigue was measured using the physical exertion and physical discomfort dimensions (Ahsberg, 2000).

[0034] 5.2.4 Existing methods

[0035] Existing fatigue management approaches rely primarily on reviewing rostered hours / shifts and potentially subjective descriptions of sleep events.

[0036] U.S. Patent Application No. 8075484B2 outlines "A system and method for assessing and alleviating fatigue, wherein an input device receives current work-rest pattern and / or sleep data from an individual. A data aggregation and processing platform combines the current work-rest pattern and / or sleep data with previous data related to the individual to produce fatigue assessment results, diagnostic assessment results, and corrective intervention results. At least one output display outputs the fatigue assessment results, diagnostic assessment results, and corrective intervention results to a user in a user-readable format. The user uses this information to modify the work-rest pattern to reduce or control future fatigue risks."

[0037] EP publication number 2278508A1 outlines "a sleep management method and system for improving a user's sleep quality, the system monitoring one or more parameters related to the user's sleep quality while the user is in bed, and receiving feedback from the user via a portable device such as a mobile phone on objective test data on cognitive and / or psychomotor performance during waking hours."

[0038] U.S. Application No. 20120316845A1 outlines: "Disclosed are distributed computing methods and systems in which intensive fatigue risk calculations are partitioned based on available computing resources, parameters of fatigue risk calculations, needs of time-sensitive users, etc. In the disclosed methods, a cost function is implemented for allocating access to available computing resources. Additional methods include partitioning computing tasks by user priority needs and by general mathematical characteristics of the computations themselves. Methods included herein include: computing only the maximum prediction likelihood rather than the full probability distribution, computing the prediction likelihood using Bayesian prediction techniques (rather than re-tabulating all data in its entirety), collating intermediate results of fatigue risk calculations, where serial results can be appropriately collated (e.g., serial time slices independent of the cumulative tasks involved), using simplified (e.g., linear, first-order) approximations of richer fatigue prediction models, assigning user-identified priorities to each computing task in a plurality of these requests, etc."

[0039] U.S. Application No. 20120329020A1 outlines: "A method for determining personalized educational information is provided, the personalized educational information relating to one or more fatigue-related individual characteristics of a subject. The method involves: receiving first input data indicating expression of one or more fatigue-related individual characteristics of the subject; estimating feature values ​​of the one or more fatigue-related individual characteristics, wherein estimating the feature values ​​comprises: using the first input data and a fatigue model that associates the fatigue level of the subject with a model parameter set to evaluate a value of the model parameter set; and evaluating one or more feature estimation functions using the estimated values ​​of the model parameter set; and determining personalized educational information about the one or more fatigue-related individual characteristics of the subject based on the estimated feature values."

[0040] PCT publication No. WO 2006000166 A1 outlines operator fatigue detection from operator muscle activity and evaluates fatigue assessment rules using data mining methods from operators with known fatigue levels.

[0041] PCT publication number 2000044580A1 outlines: "Apparatus for determining a state of alertness of a subject, such as a vehicle driver or the like. The apparatus includes means (1 to 11) for monitoring one or more physiological variables associated with the subject, such as EEG, EMG and EOG signals. The apparatus also includes means (13 to 16) for deriving from one or more variable data representing the physiological state of the subject corresponding to the one or more variables and means (17) for determining from the data when the state of alertness of the subject is below a predetermined threshold. The apparatus may include means for intervening in control of the vehicle if the driver's state of alertness is below the predetermined threshold."

[0042] CA application number 2578855A1 outlines a system including an employee module that provides employees with technology and tools to self-assess, regulate and manage personal sleep, stress and fatigue levels, and an enterprise module that provides guidelines for acceptable performance levels and assists employees in meeting the guidelines through management of personal sleep, stress and fatigue levels. An individual employee's fatigue level is recorded using an activity monitor-based personal monitoring component when working and when receiving a lot of sleep and not working, for comparison with company guidelines.

[0043] PCT publication WO 2010042615 A3 outlines: "Systems and methods are provided for optimizing sleep and post-sleep performance of an individual regardless of their environment and available sleep time. The systems and methods consider factors that determine the effects of a sleep segment on agility, cognitive function, and subjective fatigue after sleep: the duration and sleep architecture of the sleep segment, the circadian point at which the segment occurs, the accumulated amount of sleep debt preceding the segment, and the subject's sensitivity to sleep deprivation. The systems and methods include monitoring sleep architecture over an extended period of time, measuring the accumulated sleep debt and evaluating and / or tailoring the sleep architecture of each subsequent horizontal segment, determining a desired sleep state that the subject should enter, and generating sensory stimulation for guiding the subject into the desired sleep state."

[0044] PCT publication number WO 2009115073 A3 outlines methods and apparatus for maintaining, restoring or increasing the concentration and alertness of a fatigued person and for preventing accidents, impairment or incorrect decision making as a result of excessive fatigue, particularly during prolonged monotonous activities.

[0045] PCT publication No. WO 2005039415 A1 outlines: "It is possible to quantify the degree of fatigue and to display the degree of fatigue. A fatigue degree measuring device (1) includes: a biological signal peak detection member (23) for detecting the peak value of each cycle of an original waveform of biological signal data; a power value calculation member (24) for calculating the difference between the peak value on the upper limit side and the peak value on the lower limit side within a predetermined time range for the peak value obtained by the biological signal peak detection member (23) and setting the difference as a power value; and a power value gradient calculation member (25) for calculating the gradient of the power value. Absolute value processing is performed on a time series signal of the gradient of the power value and an integral value is calculated. This integral value is obtained as the degree of fatigue. Therefore, it is possible to quantify the degree of fatigue of a person."

[0046] PCT publication No. WO 2007138930 A1 outlines: "A fatigue estimation device has a body movement detection section (2) for continuously detecting a user's activity frequency as an activity level. The activity level detected by the body movement detection section (2) is output to a fatigue detection section (3) which estimates a user's fatigue level based on the activity level."

[0047] U.S. Application No. 20110178377A1 outlines: "Apparatus, systems and methods for measuring, aggregating and analyzing data collected using non-contact or minimal-contact sensors provide quality of life parameters for individual subjects, particularly in the context of controlled trials of interventions on human subjects (e.g., clinical trials of drugs, or evaluations of consumer products such as perfumes). In particular, non-contact or minimal-contact measurements of quality of life parameters such as sleep, stress, relaxation, drowsiness, temperature and emotional states of humans can be estimated, as well as automated sampling, storage and transmission to a remote data analysis center. An integral part of the system is to measure objective data with as little disruption as possible to the subject's normal behavior. The system can also support behavioral and pharmacological interventions aimed at improving quality of life."

[0048] U.S. Application No. 20100099954A1 outlines "a system and method for a user to monitor and / or modify his or her sleep. In one embodiment, a sleep coaching system includes a sensor for sensing physiological signals of a sleeping user, such as an EEG, a computer storage database for storing user and sleep-related data and suggestions, and a processor for generating a set of suggestions based on the user and sleep-related data to improve the user's sleep satisfaction. The suggestions that can be communicated to the user to improve the user's sleep satisfaction can include a sleep coaching plan, which can include one or more sleep coaching workshops that the user can attend."

[0049] Therefore, there is a need for systems that assist in monitoring and managing fatigue in a more objective and / or personalized manner. Summary of the invention

[0050] The present invention is directed to systems and methods for monitoring and managing fatigue.

[0051] A first aspect of the present invention is a fatigue monitoring and management system.

[0052] Another aspect of the invention relates to a method for monitoring and managing fatigue.

[0053] Aspects of the invention combine objective assessments of mental and physical "rest" derived from unobtrusive sleep sensors with physical activity and user data (both objective and subjective) to provide fatigue assessments.

[0054] One form of the invention includes a system for monitoring a user's fatigue state, the system comprising one or more data sources, such as objective measurements of sleep and SDB, subjective user data, objective fatigue measurements, and environmental data, and a monitoring module that analyzes the data to generate an assessment of the user's fatigue state.

[0055] The disclosed system can provide information to the user based on its assessed fatigue state (which can prompt the user to change their behavior or at least be aware of their fatigue level), or provide information to a third party in order to manage the user's activities (e.g., assigning a driver or pilot based on fatigue level). Thus, the system is configured to proactively manage the user's fatigue state.

[0056] In one example, the disclosed system acts as a personalized fatigue and sleep analyst that can provide a person with a personalized and scientifically valid "go to bed" time that will maximize sleep and minimize fatigue for better health and performance.

[0057] The following are aspects of the invention in various forms.

[0058] 1. Use the linear module to monitor fatigue (and daytime sleepiness levels) based on objective sleep parameters over the last 24, 48, 72 hours and longer periods of time, including light, deep and REM sleep, interruptions and bedtime and wake-up times

[0059] 2. A method that combines objective sleep measures with daily activity levels and caloric intake and body mass index (BMI) to monitor fatigue (chronic or acute) to form a quality of life measure.

[0060] 3. A method that combines objective sleep measurements and optionally activity and other lifestyle parameters with subjective user data collected via questionnaires and / or via games to estimate reaction time / vigilance as a proxy for fatigue.

[0061] 4. A method of combining AHI measurements of sleep periods estimated by an RPT device or RF sensor with a baseline AHI to allow a user to estimate their sleepiness / fatigue, where the AHI is seen to increase with sleepiness / fatigue (and other factors such as alcohol controlled for by a user questionnaire).

[0062] 5. A method for improving CPAP therapy compliance by using fatigue measurements to influence CPAP therapy compliance.

[0063] 6. A method of combining snoring levels and duration estimated by audio processing with the user's baseline snoring level, where it can be seen that snoring levels increase with sleepiness / fatigue (along with other factors such as alcohol controlled by the user questionnaire). The snoring measurement obtained from audio processing can be combined with sleep sensor data processing and other measures of sleep-disordered breathing to improve accuracy.

[0064] 7. Use of a default fatigue threshold (e.g., less than 5 hours of sleep causing acute fatigue), which is subsequently modified based on comparison of user demographics stored in a database with group values, and further modified based on data collected from users that changes over time.

[0065] 8. Cognitive refresh (estimated by REM sleep duration and optionally by a vigilance or attention game or questionnaire or prompt) is correlated to the user's state of mind (mood) via a questionnaire or prompt (e.g., via a smartphone app) and past values ​​to be able to predict state of mind based on future sleep parameters.

[0066] 9. Compare objective sleep and estimated fatigue parameters with standard demographic data to estimate "true sleep age", i.e., estimate the user's equivalent age based on his sleep and fatigue (e.g., a 20-year-old male with poor sleep hygiene might be classified as having a "true sleep age" of 45).

[0067] 10. Objective measurement of REM and deep sleep parameters from radio frequency motion sensors or mattress based pressure sensors, where deep sleep is associated with physical recovery and REM sleep is associated with mental (cognitive) recovery from fatigue.

[0068] Some versions of the invention may include a system for monitoring fatigue of a user. The system may include one or more data sources. The one or more data sources may include a sleep sensor, such as an unobtrusive sleep sensor, configured to generate an objective user sleep measurement. The system may also include a fatigue monitoring module, such as a processor module. The module may be configured to generate a user fatigue state assessment based on data from the one or more data sources.

[0069] In some cases, the one or more data sources may further include one or more of the following: an activity sensor configured to generate user physical activity data; an environmental sensor configured to generate environmental data related to ambient conditions in the user's sleeping position; a device configured to capture subjective user data related to the user's self-perceived fatigue state; a device configured to capture the user's daytime vital signs data; a device configured to capture objective measurements of the user's fatigue or sleepiness; a clock; and the user's work mode information.

[0070] In some cases, the subjective user data may include lifestyle parameters and perceived sleep quality, the lifestyle parameters including one or more of the following: caffeine intake, stress level, energy level, state of mind. Optionally, environmental data may include one or more of season, weather, and allergy information. The environmental data may include one or more of ambient temperature, ambient audio level, light level, air quality, and humidity. Objective fatigue or sleepiness measurements may be obtained from user testing. The objective fatigue or sleepiness measurements may be obtained by the user playing games. The fatigue monitoring module may generate a user fatigue state assessment based on a historical database, the historical database being configured to capture data from the one or more data sources within a predetermined time window. The fatigue monitoring module may be further configured to generate a user fatigue state assessment based on user baseline parameters derived from data trend analysis of the historical database. The fatigue monitoring module may be further configured to generate a user fatigue state assessment based on a group database, the group database including data from the one or more data sources of multiple users of the system. The fatigue monitoring module may be further configured to generate a user fatigue state assessment based on user baseline parameters derived from responses to a questionnaire.

[0071] In some cases, the sleep sensor may be further configured to provide a sleep disordered breathing measurement. The sleep disordered breathing measurement is a snoring measurement. Optionally, the sleep sensor data may be combined with audio data from an audio sensor to obtain a snoring measurement. The snoring measurement may be limited to intervals when the sleep sensor data indicates that the user is present and asleep. The sleep sensor may be a motion sensor, and obtaining a snoring measurement may include detecting snoring-like events in audio data that is contemporaneous or synchronized with high-frequency components in a respiratory motion signal from a motion sensor. In some cases, the sleep disordered breathing measurement may be an apnea index, a hypopnea index, and / or an apnea-hypopnea index. In some versions, the sleep disordered breathing measurement is an elevated respiratory rate.

[0072] In some versions of the system, the user fatigue state assessment may include an estimate of the user's current fatigue state. In some cases, the user fatigue state assessment may include or involve a prediction of the user's future fatigue state at a specific time. In some versions, the objective sleep measurement may include one or more of the following: heart rate, breathing rate, biomotion level, sleep statistics, skin galvanic response, and body temperature. In some versions, the sleep statistics may include one or more of the following: sleep duration, sleep quality, number of sleep interruptions, REM sleep duration, wakefulness after falling asleep, sleep inertia, and sleep latency.

[0073] Some versions of the system may include a third party information module configured to provide information to a third party related to the user's fatigue state assessment. In some cases, the system may include a user information module configured to provide information to a user related to the user's fatigue state assessment.

[0074] In some cases, the sleep sensor can be a sensor integrated with a respiratory pressure therapy device from which the user can receive or is receiving CPAP therapy. In some versions, the fatigue monitoring module can be a linear classifier configured to linearly combine data from the one or more data sources to produce a fatigue index. Optionally, the fatigue monitoring module can apply a set of rules to data from the one or more data sources to produce a fatigue index. In some cases, the fatigue index can be mapped to one of a set of fatigue states. The fatigue monitoring module can be implemented on a processing device associated with the user, which can be connected to the one or more data sources. The fatigue monitoring module can be implemented on a remote server, which is connected to the one or more data sources via a network.

[0075] Some versions of the present invention may include a method for monitoring fatigue of a user. The method may include generating a fatigue state assessment of the user, such as in one or more processors, based on data from one or more data sources. The data may include objective sleep measurements of the user, which are generated by an unobtrusive sleep sensor. Optionally, the method may further involve providing the fatigue state assessment to the user through an output device such as connected to the processor. In some cases, the method may further include generating or making a recommendation to the user based on the fatigue state assessment. In some cases, the recommendation may be an ideal time for the user to go to bed. In some cases, the recommendation may be an optimal time for the user to wake up. In some cases, the user may be undergoing CPAP treatment, and the recommendation may be a recommendation to improve or change the CPAP treatment. In some cases, the method may further include providing the fatigue state assessment to a third party. In some cases, the generation of the assessment may include applying a linear classifier to the one or more data sources. In some cases, the method may further involve applying a nonlinear transformation to one or more data sources, such as before generating the assessment.

[0076] Of course, parts of these aspects may form sub-aspects of the present invention. Likewise, each sub-aspect and / or each aspect may be combined in various ways and may also constitute additional aspects or additional sub-aspects of the present invention.

[0077] Other features of the present invention will be better understood from the information contained in the following detailed description, abstract, drawings and claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Various aspects of the invention will hereinafter be described, by way of example and not limitation, with reference to the accompanying drawings, in which like reference numerals refer to similar elements.

[0079] Figure 1 FIG. 1 is a block diagram of data flow for a fatigue monitoring and management system according to one form of the present invention.

[0080] Figure 2 A person sleeping in bed is shown being monitored by an example contactless sensor.

[0081] Figure 3 Contains example graphs of adenosine levels in a person's bloodstream over the course of a day.

[0082] Figure 4 For use as Figure 1 Block diagram of an example linear classifier for the fatigue monitoring module with data flow.

[0083] Figure 5 Plot assigned to an example representing sleep stages that vary according to age.

[0084] Figure 6 Contains a diagram of how sleep measurement and work pattern data can be displayed relative to an individual's fatigue index over several days.

[0085] Figure 7 Plot containing the predicted mean reaction times on the Psychomotor Vigilance Test (PVT) on training and test data versus the actual PVT mean reaction times (fatigue index). DETAILED DESCRIPTION

[0086] Before describing the present invention in further detail, it is to be understood that the present invention is not limited to the specific examples described below but can be varied. It is also to be understood that the terms used in the specification are only for describing the specific examples discussed herein and are not intended to be limiting.

[0087] 8.1 Fatigue Monitoring and Management System

[0088] Figure 1 A block diagram showing the data flow of a fatigue monitoring and management system 100 according to one form of the present invention. The main processing block is a fatigue monitoring module 110. The fatigue monitoring module 110 uses inputs from multiple data sources (data 115 to 170) to provide a user fatigue state assessment 180 for the fatigue monitoring and management system. The assessment 180 can use an estimate of the user's current fatigue state or a prediction of the user's future fatigue state at a specific point in time. The user information module 185 generates and provides information to the user based on the estimated or predicted fatigue state. The third-party information module 190 generates and provides information to a third party based on the estimated or predicted fatigue state. The information modules 185 and 190 can proactively manage user fatigue in advance, as described in more detail below.

[0089] The fatigue monitoring and management system 100 is based on personalized use of a fatigue module, whereby the system can "learn" user-specific sleep patterns and correlate the user-specific sleep patterns with subjective fatigue assessments (how the user "feels," e.g., according to the Karolinska Sleepiness Scale (KSS)), or objective fatigue measures (e.g., performance on the Psychomotor Vigilance Test, or PVT).

[0090] The fatigue monitoring and management system 100 can be optionally connected to an occupational health and safety management system (OH&S) implemented in a consumer setting to serve individual individuals, or implemented in a workplace as part of an enterprise workplace fatigue management approach. This is the system with group management capabilities, utilizing de-identified (anonymous to the company) data. The customized enterprise data represents a subset of the full group database (e.g., group data 150) that is used to provide standard data comparison to the fatigue monitoring module 110.

[0091] In a consumer embodiment, the fatigue monitoring module 110 may be implemented as software running on a processing device associated with the user, such as a personal computer or laptop, mobile phone, smartphone, or tablet computer, which is connected to one or more of the various data sources (data 115 to 170) in a wired or wireless manner, which may optionally include connecting to or communicating with one or more sensors to detect or generate signals from which data (discussed in more detail herein) may be derived. The sleepiness, sleep health, and cognitive performance questionnaires (used to obtain subjective user data 145) may be completed by the user on the same processing device. The user information module 185 may then provide information to the user via the same processing device.

[0092] Alternatively, the fatigue monitoring module 110 can be implemented "in the cloud", that is, on a remote server connected to various data sources (data 115 to 170) via a network. As in a workplace embodiment, this can aggregate fatigue status data from multiple individuals. However, consumer embodiments can also utilize a "cloud-based" fatigue monitoring module. The remote server can transmit the sleepiness, sleep health and cognitive performance questionnaire (used to obtain subjective user data 145) to each individual via the network, and each individual can complete the questionnaire using a network device such as a personal computer or laptop, mobile phone, smartphone, or tablet computer. The user information module 185 can then provide information to each individual via the same device. The third-party information module 190 can provide information to third parties such as corporate officials via the network.

[0093] The various data sources (data 115 to 170 ) used by the fatigue monitoring module 110 are described in detail below. In various forms of the invention, the fatigue monitoring module 110 uses some subset of the data sources (data 115 to 170 ) to generate an estimated or predicted fatigue state 180 .

[0094] Therefore, the methods described herein may be implemented in and / or by one or more processing devices. For example, as described herein, the data source and / or module embodiments of the present invention may be a processing device that may have one or more processors to implement a specific method or algorithm described in more detail herein. Such a device or apparatus may include an integrated chip, a memory and / or other control instructions, data or information storage media. For example, programmable instructions containing these methods may be encoded on an integrated chip in a memory of a device or apparatus to form an application specific integrated chip (ASIC). Such instructions may also or optionally be loaded as software or firmware using a suitable data storage medium. Such a processing device may optionally further communicate or be connected to input devices such as keyboards, touch screens, sensors, etc.

[0095] In the following, it is generally considered that people sleep at night and are awake during the day, so "daytime" is synonymous with "wakefulness period" and "night" is synonymous with "sleep period". However, the disclosed fatigue monitoring and management system 100 also contemplates use by shift workers whose sleep and wakefulness periods may not coincide with night and day, respectively. For such users, "daytime" can still be understood to mean "wakefulness period" and "night" can be understood to mean "sleep period", even though the true correspondence is more likely to be the opposite.

[0096] 8.1.1 Physical activity (data 115)

[0097] Increased fatigue is associated with lack of exercise and activity. Physical activity data 115 can be obtained from wearable activity sensors (e.g., pedometers, "step counters", three-axis accelerometer-based activity monitors, altimeters, and branded commercial activity sensors such as "Fitbit", "Jawbone Up", and "Actiwatch"). Alternatively, physical activity data 115 can be obtained from portable devices incorporating activity sensors, such as smartphones. Examples of physical activity data are the number of steps taken and exercise duration and intensity. Physical activity data 115 is used to construct a profile of the user's daily activity level.

[0098] 8.1.2 Daytime vital signs (data 118)

[0099] Fatigue may be related to stress levels during the day. Daytime dynamics of physiological vital signs such as heart rate, respiratory rate, and blood pressure can be collected by wearable devices such as a Polar heart rate chest strap, ECG / R-peak, or other data collection devices attached to the chest, or by earbuds (earphones) (e.g., made by LG or Blaupunkt) that record these signals using photoplethysmography (PPG). In one embodiment, blood pressure estimates from earbud sensors can be used to track blood pressure values ​​non-invasively during the day (e.g., when the user is listening to music / commuting / exercising, etc.).

[0100] The heart rate estimate may also be combined with the daytime physical activity data 115 to derive an energy expenditure estimate.

[0101] Detecting trend changes in heart rate can indicate increasing fatigue, such as in overtrained athletes.

[0102] Chronic stress may also manifest as daytime fatigue and may be monitored based on daytime physical activity data 115 , daytime vital sign data 118 , and objective sleep measures 120 (described below).

[0103] Other chronic symptoms, such as chronic disease progression, may manifest as changes in the daytime vital sign data 118, such as increased heart rate, decreased heart rate variability, increased respiratory rate, changes in blood pressure, and decreased energy expenditure.

[0104] Jackowska et al. (2012) found that sleep problems were associated with decreased daytime heart rate variability (HRV). Therefore, when looking at fatigue overall, the fatigue impact can be measured based on nighttime sleep patterns (e.g., fragmented / poor sleep) and then estimating the next day's HRV. As daytime impact increases, the cumulative fatigue impact may increase.

[0105] 8.1.3 Objective sleep measurement 120

[0106] A "sleep sensor" that monitors a user's sleep and breathing may be utilized to provide objective sleep measurements 120 such as daily and longitudinal trends of objective sleep quality and bio-motion levels in the bedroom and other sleeping positions. Advantageously, the objective sleep measurements 120 are extracted from a sleep sensor that is minimized or least noticeable to the user. Unobtrusive sleep sensor technology does not require wearing, i.e., the sensor is "set and forget," thereby maximizing usage life. An example of an unobtrusive sleep sensor is a contactless sensor. Figure 2 Monitoring of a sleeping person 1000 in bed is shown by an example non-contact sensor 7000, which may be a radio frequency Doppler motion sensor (SleepMinder). Another example of an unobtrusive sleep sensor is a mattress based capacitive or resistive mat (including piezoelectric) pressure sensor.

[0107] In embodiments where the user receives CPAP therapy for obstructive sleep apnea from a RPT device through a patient interface, such as a mask, the sleep sensor may be a sensor integrated with the RPT device or the patient interface, such as a pressure or flow rate sensor.

[0108] Objective sleep measurements 120 extracted from sleep sensor signals may include sleep statistics such as sleep duration, sleep quality (actual amount of sleep during the expected sleep period), and the number of interruptions to sleep. Bedtime and wake-up time may also be extracted from sleep sensor signals. REM sleep periods, as well as deep sleep (slow wave sleep) periods may be extracted. As described above, REM sleep duration indicates cognitive rest, while deep sleep duration indicates physical rest. Relative ratios of deep sleep, light sleep, and REM sleep durations may also be extracted. Another sleep statistic is sleep inertia, which represents the expected level of sleepiness immediately following awakening, especially when the user awakens directly from deep sleep. Wake-up after falling asleep (WASO) provides a total of fragmented sleep, which may be used in combination with the number of interruptions detected. The time spent sleeping (sleep latency) and the sleep stage upon awakening may also be extracted.

[0109] In the case of a sleep sensor with the ability to measure heart rate, such as an RF Doppler motion sensor that monitors a ballistocardiogram or a mattress-based pressure sensor, a power spectrum analysis can be performed on the signal to show variability in intracardiac intervals that can predict fatigue. In particular, the ratio of HF (high frequency) to LF (low frequency) power in the heart rate spectrum can be used to estimate parasympathetic activity, where a decrease in the parasympathetic component (of the autonomic nervous system) indicates increased stress levels and increased fatigue. In particular, sympathetic nervous system (fight or flight response) activity in chronic fatigue can indicate just when the body should be recovering, i.e., during sleep. Heart rate data (as well as HRV) can also be obtained from contact sensors, such as a wearable watch-type device with an optical sensor, a pulse oximeter, a chest strap such as a Polar heart rate monitor, or a headband with EEG electrodes, etc.

[0110] The galvanic skin response (GSR, also known as galvanic skin response) can be recorded by a wearable device (e.g., a Basis watch, or other commercially available GSR measurement devices). The GSR signal can be used as a surrogate measure of the "fight or flight" activity of the sympathetic nervous system. Combining the extracted GSR and HRV signals can also produce an estimate of the ratio between sympathetic nervous system activity and parasympathetic nervous system activity.

[0111] HRV and GSR data may also be collected while the user is awake (eg, via chest electrodes or a wristwatch-like device), in which case the HRV and GSR data form part of the daytime vital signs data 118 .

[0112] For example, PCT Publication No. WO 2010 / 091168, which is incorporated herein by reference in its entirety, outlines methods for extracting objective sleep measurements 120 such as biological motion level, breathing rate, heart rate, and sleep statistics from non-contact radio frequency Doppler motion sensor signals.

[0113] A specific example of a sleep sensor is based on a rubber (or other conductive material with a plastic or rubber covering) impregnated with graphene (a crystalline form of carbon) piezoresistive cross-mattress strip with electronic buffers, amplifiers and digitizing circuitry.

[0114] In cases where the objective sleep measurement 120 appears to be worse than the subjective user data 145 , comparisons may be made to other life parameters as well as the SDB measurement 125 .

[0115] Another objective measure 120 of actual sleep quality that may affect fatigue levels is body surface temperature (calibrated to ambient temperature). For example, such a temperature monitoring device may be embodied as a BodyMedia Sensewear device that incorporates two sensors to measure ambient and skin issues, thereby correcting for internal body temperature. Body temperature is important for assessing sleep quality because characteristic temperature patterns run throughout a normal sleep night. Fatigue promoters (such as jet lag) also often cause deviations from this pattern because the sleep temperature profile is "over-dominated" by basic physiological changes in body temperature. Therefore, an objective sleep measurement 120 that may play a role may be an estimated body temperature throughout the night, and the variation of this temperature profile from the expected profile of the user.

[0116] 8.1.4SDB Measurement 125

[0117] In addition to poor sleep hygiene (such as self-imposed sleep restriction), a major contributing factor to fatigue is the presence of sleep-disordered breathing (SDB). Examples of SDB measurements 125 are the apnea-hypopnea index (AHI) and the incidence and severity of periodic limb movements (PLM). For example, PCT Publication No. WO 2010 / 091168 outlines a method for extracting SDB measures such as AHI from non-contact radio frequency Doppler motion sensor signals, the contents of which are incorporated herein by reference.

[0118] Another SDB measurement 125 is elevated breathing rate, for example, an average breathing rate of 20 to 24 breaths per minute during sleep (depending on age and underlying conditions).

[0119] In addition, snoring is a SDB measure that can be quantified based on short-time Fourier transform (STFT) analysis of sampled audio waveforms obtained from audio sensors and estimates of their normalized sub-band power levels 125. Mel-frequency cepstral coefficients (MFCCs) can be used to distinguish snoring from speech, and spectral flux (analyzing the changes between power spectrum estimates) can be used to detect the onset of snoring. RMS (root mean square, or quadratic mean) can be used in combination with STFT power levels and running median filtering to distinguish trigger sound levels from background noise.

[0120] Snoring (and snoring stops associated with apnea or hypopnea) can be combined with bio-motion levels and extracted sleep statistics in the following manner. The smartphone can be placed on a nightstand next to the bed where the subject normally sleeps. The Universal Audio Recorder (Voice Memo) smartphone app can be used to save audio files from the built-in audio sensor (microphone) to memory. To minimize temporary storage space, the audio should be sampled at 11.025kHz, mono (single channel). In addition, 22.5kHz or 44.1kHz (or other sampling rates) can be used. Preferably, the sample should be sampled at the highest rate available, and the sampled signal can be subsequently downsampled. (Some domestic smartphone microphones set the sampling rate cap at 8kHz). It is also preferred that the microphone faces the user. For example, the Samsung Galaxy S3 has two microphones, and the "recording" microphone is a pinhole on the lower right edge. The Apple iPhone 5 also has 2 microphones, with the recording microphone on the bottom edge of the left hand. Note that some iPod versions require an accessory headset / microphone to be connected to be able to record (e.g., versions older than the Gen 4 iPod Touch).

[0121] For ease of processing, the audio data can be read from the input file in blocks. The block size is configurable and represents a trade-off between memory I / O speed and available memory. Ten to twenty minute frequency bands are example block sizes. Signal preprocessing can involve applying a digital wideband filter, keeping frequency content in the range of 100 Hz to 4,000 Hz. This can be achieved by using a direct form FIR filter using a Kaiser window or by other means. The signal can also be re-sampled at 8000 Hz (if storage space constraints require this). Optionally, the signal can be compressed and expanded, for example by applying a μ-law compressor or similar method. This can also be managed by other signal processing devices, without necessarily requiring compression and expansion.

[0122] In the frequency domain, features may include subband power (in Hz) in the following subbands: 0.01-500, 500-1000, 1000-1500, 1500-2000, 2000-2500. For higher sampling rates, higher frequency bands may be considered. Frequency bands may also be divided into smaller or larger frequency segments. For example, a specific subband below 800 Hz may be considered, i.e., 0-800 Hz (de-averaged). Other spectral measures are spectral centroid, and harmonic spectra of audio processing steps such as "pitch"-products, spectral flux, spectral spread, spectral autocorrelation coefficient, and / or spectral kurtosis. For example, the ratio of power greater than 800 Hz to power less than 800 Hz may distinguish simple snoring from SDB (wherein the specific separation frequency is chosen empirically based on the snoring population).

[0123] Visually, it is possible to track the "lines" of snoring frequencies in the spectrogram (based on the STFT). Therefore, it is possible to use an automated algorithmic approximation to detect and extract these "lines" based on techniques used in military sonar and the like, including hidden Markov models and image tracking techniques.

[0124] In the time domain, features can include zero-crossing rate, autocorrelation coefficient, and running amplitude measurements. Other methods are to calculate short-term energy and short-term zero-crossing rate.

[0125] A common problem with detecting snoring from audio signals is corruption from background noise - that is, noise sources not originating from the user being monitored. These sources can include white noise or other colored noise-like sources (e.g., fans, air conditioners, road noise), speech, music from TV, radio, smartphones (e.g., a user watching a movie on a tablet), another person talking in the room, or even another person snoring (like the person in the bed), etc. This problem can be solved by prior audio processing, and also by combining audio processing with sleep sensor data processing (including absence / presence, wakefulness, and sleep stage detection).

[0126] If the user plays audio / media on the same smartphone (or on a connected device or network device) that implements snoring detection, it is possible to exclude non-snoring components, i.e., exclude those sounds that do not originate from snoring. In order to play audio / media - or other voice sources, such as conversations in the room - a speech recognition system (e.g., based on a sound synthesizer) can be used to scan the recorded audio sources. For this application, a speech recognition feature is used to prevent snoring detection - assuming that snoring only occurs when there are other voice sources. This can cause problems when the user is asleep and the radio or TV is on, but this problem can be solved by using sleep stages detected by a separate sleep sensor (i.e., specifically limiting snoring detection to the intervals of (a) when the user is present and (b) when the user is asleep).

[0127] By combining audio processing and sleep sensor data processing (including absence / presence and sleep segmentation), it is possible to limit snoring detection to when the user closest to the audio sensor is asleep. Therefore, even if the bed partner snores, but the monitored user is awake, these snoring events will be excluded. This also has the effect of excluding snoring when the user gets up (e.g., to go to the bathroom) because the user is detected as being awake and out of range (absent) of the sleep sensor.

[0128] There are still confounding factors: both bedmates may be asleep and both snoring (or, only the distant (unmonitored) user is snoring, and both are asleep). In this case (or other cases), access to the respiratory motion signal of the motion sensor can be combined with snoring-like events detected from the audio signal. For example, audio snoring events can be associated with high-frequency components superimposed on the respiratory motion signal - for example, high-frequency components during the inspiration phase. In other snoring types, snoring can occur in both the inspiration and expiration phases. In the respiratory motion data, these snoring patterns can depend on the positioning of the motion sensor. Detection of apnea and hypopnea events from the respiratory motion signal can also be associated with changes in the detected snoring pattern. An additional advantage of obtaining the snoring audio signal and the vibrations from the motion sensor at the same time is that, since snoring is more common during inspiration, snoring can be used to decide whether the motion signal represents inspiration or expiration. Additional clues can be obtained by evaluating the ratio of inspiration to expiration (normal subjects are usually 1:2).

[0129] Snoring events may also be associated with sleep stages, such as occurrence during non-REM sleep versus REM sleep. While traditional snoring may be more prevalent during deep and light sleep, it may be less commonly seen during REM sleep. REM sleep may contain different audio events, such as those of resuming breathing from apnea (for example).

[0130] Another means of detecting snoring from a particular user - or even accurately assessing their breathing rate - is to compare the breathing rate estimates extracted from the audio signal and from the motion signal. When these breathing rate estimates are closely correlated, it is likely that the most recent user was detected. In cases where the breathing rate estimates all maintain high quality, but are numerically discrete, another user may have been detected. In cases where the audio signal contains significant snoring (or the audio quality is very poor), the breathing rate estimate cannot be derived from the audio signal, and the signals cannot be correlated in this way.

[0131] 8.1.5 Objective fatigue measurements 130

[0132] An important aspect of the disclosed fatigue monitoring and management system 100 is that it can incorporate acceptable objective sleepiness or fatigue measurements 130, such as those obtained from user testing. Generally, tests used to obtain objective fatigue measurements 130 address cognition, executive function, stress and energy levels, all of which are affected by fatigue. Formal user tests that can be used to obtain objective fatigue measurements 130 include the Mental Vigilance Test (PVT), Osler (Oxford University Sleep Resistance Test), and the Maintenance of Wakefulness Test. The PVT assesses reaction time and hand / eye coordination. The PVT test is a reasonably widely accepted type of attention measurement that has been previously shown to be a reasonable indicator of sleepiness (a known correlate of fatigue). In one example of a PVT, light is randomly presented to a user via a handheld device, where the user must press a button to respond to the device. The metrics provided by the PVT include average or median reaction time and number of misses. In another example of a PVT, the user must follow a target and maintain its position. Sustained attention response testing is another method of measuring sleepiness.

[0133] In one embodiment, PVT may be performed four times a day (eg, 2 hours after waking up, 1 hour after lunch, 30 minutes after dinner, and 1 hour before going to bed).

[0134] Other types of tests induce mental fatigue and assess its level. For example, the computerized Kraepelin test is one way to induce mental fatigue load. In this test, the user is asked to repeat mental arithmetic and selection tests over and over for a long period of time (e.g., 1-2 hrs). The task itself has been shown to induce fatigue, and fatigue can be assessed by observing the speed of calculations (plus changes in speed) and the accuracy of the answers.

[0135] Another way to quantify fatigue during waking hours is to use EEG fatigue measures. Several EEG fatigue predictors have been evaluated, such as those described by Trejo et al.

[0136] In one embodiment of the disclosed fatigue monitoring system (e.g., system 100), a user wears an EEG headband during a training phase of the system to provide objective fatigue measurements 130. These measurements are used to train a model related to the measured objective and subjective user data and the actual fatigue experienced by the user.

[0137] Such a test may be incorporated into the fatigue monitoring and management system 100 to provide an objective sleepiness or fatigue measurement 130 (and thus serve as an internal "calibration" point for fatigue).

[0138] Objective fatigue measure 130 can be a "hard" measure from more formal tests of sustained attention as described above, or a "soft" measure derived from a game. As an alternative to the latter, attention tests can be "gamified" to maintain engagement. Gamification is the use of game thinking and game mechanics in a non-game context to engage users and solve problems (Zichermann & Cunningham, 2011). Specifically, gamification of attention tests can achieve social aspects and goal-seeking aspects. Social aspects allow group segment comparisons, such as "my sleep" versus the sleep of other employees, people of my age, top performers, etc. These comparisons can also be shared on social media sites such as Twitter and Facebook. Goal-seeking aspects generate "well done" information in the app through point accumulation, or increase the concept of goals and rewards through tangible prizes brought by corporate sponsors of the system. Data from games (such as cognitive, vigilance, and executive function estimates) can provide neurobehavioral performance estimates. These involve memory, mistakes made (e.g., missing game "goals"), and mental and physical reaction times.

[0139] 8.1.6 Working mode (data 135)

[0140] If work mode information 135 is provided, the fatigue monitoring and management system 100 can obtain more detailed information, which can be as simple as "working five days a week, 8:30 a.m.-5 p.m." or as complex as "detailed shift schedule." The work mode information 135 can indicate the user's possible sleep pattern and improve the accuracy of the fatigue state estimation 180.

[0141] In one example, the ratio of daytime sleep to nighttime sleep can be extracted as part of the objective sleep measurement 120. The ratio can vary based on naps, shift work, jet lag, and other factors. Specifically, if the fatigue monitoring and management system is provided with work pattern information 135 of upcoming work time (e.g., shift change) or travel across multiple time zones, it can allow the user to "expect" the presence of sleep because the user is prepared for the upcoming change (e.g., if the user deliberately changes his sleep habits in anticipation of changes in sleep time).

[0142] 8.1.7 Historical Measurement 140

[0143] The fatigue monitoring module 110 can utilize historical information (e.g., historical measurements 140) from other data sources (data 115 to 170). For example, useful information about the user's possible fatigue state in the last 1-7 days can be provided. For example, for a typical sleep time observed to be

[0144] For a 7.5 hour user, a gradual shift to 6 hours per night may indicate cumulative fatigue. A sudden shift to 5 hours per night or less (i.e., within a short period of time) may indicate acute fatigue.

[0145] The history of each parameter from other data sources may be stored in the database of historical measurements 140 within a predetermined time window, for example, 7 days.

[0146] 8.1.8 Subjective User Data 145

[0147] Subjective user data 145 represents user input data, such as responses to questionnaires. One form of subjective user data 145 relates to a user's subjective or self-perceived sleepiness. For example, the Epworth sleepiness scale (ESS) and the Stanford sleepiness scale (SSS) are both questionnaire-based and subjectively quantify sleepiness. The Stop-Bang assessment questionnaire can provide an SDB risk estimate.

[0148] As an indicator of subjective sleepiness, the Karolinska Sleepiness Scale (KSS) has been developed and is widely used. The KSS asks a person to rate their sleepiness at that point in time based on the following five-point scale:

[0149] 1 = Extremely alert

[0150] 3 = Alert

[0151] 5 = Neither alert nor sleepy

[0152] 7 = Sleepy, but not difficult to stay awake

[0153] 9 = Extremely sleepy, trying hard to avoid sleep.

[0154] Subjective user data 145 may be captured by a user device such as a smartphone or tablet. In one example, a smartphone "app" provides a "tired button" that the user activates when the user feels tired / fatigued (e.g., when yawning) and optionally allows the user to select their perceived severity of fatigue via an associated counter. In another example, referred to as a Visual Analog Scale (VAS), the app provides a sleepiness continuum along which the user moves a "slider" (e.g., in a graphical user interface (GUI)) to indicate their subjective sleepiness level, rather than requiring the user to rate themselves against a verbal statement.

[0155] In one embodiment, the VAS may be performed four times per day (eg, automatically triggered by a processor) (eg, 2 hours after waking up, 1 hour after lunch, 30 minutes after dinner, and 1 hour before going to bed).

[0156] Other forms of subjective user data 145 related to perceived fatigue are lifestyle parameters such as caffeine intake, stress level, energy level, state of mind (mood), perceived sleep quality and menstrual cycle, and the likelihood of associated fatigue. Each such lifestyle parameter can be captured by the user device described above and used as input to the fatigue monitoring module 110.

[0157] 8.19. Group data 150

[0158] Aspects of the disclosed fatigue monitoring and management system 100 are that analysis can be used to compare a user's records to larger group records to establish norms and simultaneously determine situations where the user can fit within the group. For example, some individuals are more susceptible to short-term sleep loss than others, and this can be learned by monitoring changes in personnel records over time. At a higher level, certain regions may present different fatigue characteristics, for example, involving local industries, short-term effects due to time changes, regional eating habits and BMI, socioeconomic levels, seasonal effects such as daytime duration, prevalence of diabetes, and the like.

[0159] To this end, the group data 150 includes parameters from multiple users of the fatigue monitoring and management system 100. The group data 150 can make a preliminary assessment of fatigue status based on demographic, psychological, behavioral and geographic parameters using data from similar users. In order to enable comparison with normative measurements, the group data 150 can be de-identified to protect user privacy.

[0160] 8.1.10 Baseline (Data 155)

[0161] Baseline parameters 155 represent normal values ​​for a user, and may be derived from trend analysis of historical measurements 140 (initially over a short time frame, but extending to an overview of weekly, monthly, or yearly data).

[0162] A baseline may also be established by responses to questionnaires such as the Pittsburgh Sleep Questionnaire (PSQ) and the Quality of Life (QoL) Survey. The questionnaires may be administered and responses captured 155 via a user device such as a smartphone or tablet.

[0163] 8.1.11 Environment (Data 160)

[0164] The user's local environment may play a role in their fatigue level, particularly if the user is sensitive to pollen counts or light levels (e.g., seasonal affective disorder). Large-scale environmental data 160 such as season, weather, and allergy information may be collected based on the user's sleeping location and cross-checked with appropriate online databases (either by providing details in a distribution map, or obtained via a positioning device, such as a GPS device).

[0165] More specific environmental data 160 includes ambient temperature of the user's sleeping location (related to comfortable sleeping temperature), ambient audio level (to detect loud noises), light level (to identify sunrise time, cross-check with season information and location, and control room lights), air quality, and humidity. This data can be obtained from a specially calibrated environmental sensor such as Onset HOBO that records temperature values, or through a smartphone temperature (e.g., ambient audio) sensor that is in the same room as the user.

[0166] 8.1.12 Daylight hours (data 170)

[0167] In fatigue assessment, the daytime data 170 obtained from the clock can be used as a key lookup measurement. The daytime data 170 can capture basic diurnal or circadian fatigue variations, for example, using a sinusoidal model.

[0168] 8.1.13 Other data input

[0169] Some embodiments of the fatigue monitoring and management system 100 incorporate physiological measures of acceptable sleepiness and alertness, such as adenosine and melatonin levels in the blood, as potential calibration points for the user's overall sleepiness level. Adenosine levels indicate sleepiness because adenosine slows neural activity. Figure 3 Graph 3000 containing adenosine levels in a person's blood over the course of a day. In the first segment 3010, the person is asleep and the level is low. At time 3020, the level begins to rise just before the person wakes up, and continues to rise during the wake period until it exceeds a "severe sleepiness level" 3030, at which point the person is sleepy. At time 3040, the person falls asleep, after which the level begins to fall. Note that caffeine can cause a short-term increase in wakefulness levels by preferentially binding to adenosine receptors in the brain and thereby mitigating the neurodepressant effects of adenosine.

[0170] The fatigue monitoring and management system 100 may also collect information from the user regarding symptoms / disease states of fatigue, including anxiety, headaches and nasal congestion, asthma, anemia (eg, related to menstruation), depression, arthritis, diabetes, and sleep apnea.

[0171] Another type of data that the fatigue monitoring module 110 can use is about bedding, such as the type of sheets, when they were last changed, the age of the mattress and comforter, and the presence of dust allergies. This data can be combined with its estimated fatigue pattern, sleep trends (particularly the amount of movement), the room's environmental data 160 (e.g., temperature), and the group data 150.

[0172] In a driving embodiment of the fatigue monitoring and management system 100, information related to the user's driving can be obtained through an accelerometer and a positioning device (e.g., a GPS-based device) to distinguish driving from movement (i.e., distinguishing driving from other types of movement). Such data can be used to guide a map database including a local road network, particularly a map database of so-called "accident hotspots" for fatigue-related accidents, such as areas or highways where the incidence of "falling asleep at the driver" accidents or disasters is known to be elevated.

[0173] 8.1.14 Fatigue Monitoring Module 110

[0174] As mentioned above, the fatigue monitoring module 110 takes input from one or more data sources (data 115-170) to generate a user fatigue state assessment 180. The fatigue state assessment 180 may take the form of an assessment of the user's current fatigue state, or a prediction of the user's future fatigue state.

[0175] The input data may be subjected to a non-linear transformation, such as a logarithm, before being used, for example, by the fatigue monitoring module 110. The input data may be normalized, for example, by mean subtraction, to remove "static bias" between individuals.

[0176] Figure 4 FIG. 4 is a block diagram of a linear classifier 4000 that can be used to implement a fatigue monitoring module 110 according to one form of the present invention. Figure 4 , the linear classifier 4000 uses three data parameters x, y and z extracted from the input data source (data 115 to 170) and combines them with coefficients α, β and γ respectively to generate a numerical fatigue index f. In one example, a fatigue index value of 1 indicates a high level of fatigue and a value of 0 indicates a low level of fatigue. In other embodiments, the linear classifier 4000 linearly combines more or less than 3 input parameters to generate the fatigue index f.

[0177] Another input to the linear classifier 4000 may be the time t corresponding to the fatigue index f. If the time t is current, the fatigue index f is an assessment of the current fatigue state. If the time t is a future time, the fatigue index f represents a prediction of the fatigue state at time f. The coefficients α, β, and γ typically vary with time, so the time t can be used to obtain the correct values ​​of the coefficients α, β, and γ to assess fatigue.

[0178] In one example, the input parameters input to the linear classifier 4000 include the following parameter group: {wake-up time, sleep duration last night, number of interruptions last night, running average of sleep duration in the last 5 nights, age, gender, deep sleep duration last night, room temperature last night}.

[0179] For the linear classifier 4000 to be useful, it requires appropriate values ​​of weighting coefficients α, β, and γ, and an "optimal" set of input parameters to characterize a given individual (e.g., for some individuals, the most important input parameter is wake time, for other individuals it may be last night's sleep duration). Multiple linear regression is a supervised approach to establish an "optimal" set of input parameters and corresponding coefficients by a training set. In the multiple linear regression method, the fatigue monitoring and management system 100 learns the "optimal" set of input parameters and corresponding coefficients to assess or predict a person's fatigue state. In the absence of any prior learned data, a standard linear module can be used to preset the linear classifier 4000.

[0180] The training data may be obtained from an individual user, in which case the classifier is specific to that user, or from multiple individuals, in which case the classifier is more general. Such a general classifier may include submodules based on training data from different regions. The general classifier is able to perform fatigue assessment based on data from previously unknown individuals.

[0181] Where training data is available, this would typically involve obtaining objective or subjective fatigue measures from users during the day (using, for example, a psychomotor vigilance test, a visual analogue scale, or an EEG-based fatigue test). A linear model can then be fit to a useful output fatigue index.

[0182] There are many alternatives to linear regression, such as logistic regression and nonlinear regression, which the fatigue monitoring module 110 may use to learn the “best” set of input parameters and corresponding coefficients to estimate the fatigue index f.

[0183] In other forms of the disclosed fatigue monitoring and management system 100 , the fatigue monitoring module 110 may utilize different methods to generate the fatigue state assessment 180 , such as a non-linear classifier, a support vector machine, or a neural network.

[0184] Another embodiment of the fatigue monitoring module 110 uses a rule-based approach to generate a fatigue index. An example of such an embodiment includes the following rule set with any one or more or all of the following rules:

[0185] If sleep time < 6, increase fatigue index [Note: sometimes very short sleep can affect the user's day, especially if the user has been doing well up to that point; but if it persists, then they may have chronic fatigue].

[0186] If the time to fall asleep is >10, increase the fatigue index.

[0187] If bedtime has varied by more than 2.5 hours over the past three nights, increase the fatigue score.

[0188] Deep sleep <7%, increased fatigue index.

[0189] If the decrease in deep sleep vs. the average of the previous three nights is >8%, increase the fatigue index. (If the downward trend continues on the second or subsequent nights, further increase the fatigue index. If the subsequent increase in deep sleep is >5%, the user is recovering and the fatigue index is reduced.)

[0190] If REM is very short (<5% to 7%), increase the fatigue index.

[0191] If sleep efficiency is <75%-80%, increase fatigue index (may need to check for outliers - if motion sensors are picking up background motion in the bedroom, and the user didn't bother entering bedtime into the app).

[0192] Otherwise, reduce the fatigue index.

[0193] Some other rules to consider are:

[0194] Reduced weekend fatigue index - especially if combined with longer sleep duration and reduced REM.

[0195] If the user takes a long time to fall asleep (high sleep latency), increase the fatigue index (but sometimes this feature can be misinterpreted).

[0196] If the PSQI / QOL score is poor, the user may suffer from sleep disorders, so the fatigue index is increased.

[0197] If activity data indicates that the user is often active after periods of sitting, increase the fatigue index.

[0198] The fatigue assessment 180 may be a fatigue index value f. Alternatively, the fatigue monitoring module 110 may map the calculated fatigue index to one of a set of fatigue states using one or more thresholds. Such fatigue states may represent a more easily interpreted fatigue assessment 180. In one embodiment of such a mapping, the possible set of fatigue states and corresponding thresholds are:

[0199] “All good” - therefore no disturbing fatigue state was detected (f<0.5).

[0200] "At risk" - Thus, the user's fatigue risk (0.5 < f < 0.8) is elevated. In this state, an indication of how this fatigue risk has arisen, as well as the severity of the fatigue risk, is provided. For example, the user may exhibit early-stage sleep deprivation and can correct it by changing behavior before experiencing a chronic fatigue state.

[0201] "Acute fatigue" - Thus, relatively short-term (about 1 - 2 days) sleep restriction or very poor-quality sleep results in acute fatigue of the listless type (0.8 < f < 1). Based on the work pattern information 135 or other data, this can be regarded as a high-risk state (e.g., the user is performing safety-critical duties, or the user reports planning to drive a car for a long time, etc.). This may initially manifest as physical fatigue, followed by an increase in mental fatigue (associated with related cognitive impairments).

[0202] "Chronic fatigue" (0.5 < f < 0.8 over several days) may be due to long-term sleep restriction or generally poor sleep hygiene (including poor diet, and / or a mixture of caffeinated / energy drinks to stay awake, followed by falling asleep with alcohol and sleeping pills) / insomnia. A potential SDB condition may also be the root cause.

[0203] In other embodiments, a large number of thresholds may be used corresponding to a wider range of possible fatigue states.

[0204] 8.1.15 User Information Module 185

[0205] The user information module 185 generates and provides a report for the user that includes a fatigue state assessment and, optionally, sleep statistics that form part of an objective sleep measurement.

[0206] In a consumer embodiment of the fatigue monitoring and management system 100, the user information module 185 can provide the user with personalized insights regarding the estimated fatigue level, risk factors, and improvement strategies.

[0207] Figure 6 A schematic diagram 6000 showing how the user information module 185 presents the objective sleep measurement 120, physical activity data 115, and work pattern data 135 relative to the user fatigue index f over several days. In Figure 6 it, the two measurements shown in the lower diagram 6020 are the sleep onset time and the actual shift time. The upper diagram 6010 includes the fatigue index f, which is calculated using the Figure 4 linear classifier 4000 based in part on these measured sleep and shift times. In Figure 6In the second day 6030 and the third day 6040, the fatigue index f of the user is higher because the user has experienced a shorter sleep period in the previous nights 6050 and 6060, and the shift time is longer than the usual shift time.

[0208] The user information module 185 may sound an alert if the user has not given herself or himself enough opportunity to sleep (i.e., the user has simply not allowed enough sleep time in their daily routine - they may or may not be aware of this). However, in a consumer embodiment of the fatigue monitoring and management system 100, the alert may not be prescriptive because, for example, travel or social activities may cause a short-term or temporary adverse change in fatigue status. In the case of a workplace embodiment, the alert may be configured to be more prescriptive, particularly if the user is in a safety-critical role (e.g., mining, train driver, crane operator, etc.).

[0209] The fatigue assessment 180 can be used to provide advice to the user. For example, for a user who has poor sleep at night and collects objective sleep measurements 120 through the unobtrusive sleep sensor mentioned above. During the day, fatigue-related data such as objective fatigue measurements 130, environmental data 160, physical activity data 115, location data, and dietary data are captured from the user. The fatigue monitoring module 110 analyzes the data to generate a fatigue state assessment 180. The user information module 185 estimates and recommends the "ideal sleep time" for that day based on the estimate. Therefore, the user will know in advance what time the time is and can choose to follow or ignore the advice. The attachment or compliance data itself becomes an input for further analysis. An alarm can be issued before the recommended ideal sleep time (i.e., a "sleep time" reminder, for example, transmitted via the user's smart phone). Over time, the fatigue monitoring and management system develops into an individual personalized "ideal sleep time" recommendation system.

[0210] The user information module 185 can also recommend to the user the time to wake up the next day based on the predicted fatigue state in various situations to achieve optimal awake alertness, or (b) optimal all-day alertness. The user can be warned at the recommended time by the clock program on the user's smartphone.

[0211] One benefit of estimating fatigue state 180 with objective sleep measurement 120 is that the user may actually sleep better than he or she feels he or she sleeps. Communicating this fact to the user through user information module 185 can be used to break the vicious cycle: the user believes they are going to have a bad day (psychological or illusion of fatigue), followed by excessive use of stimulants (such as coffee or other caffeinated beverages), followed by the use of alcohol or sleeping pills before going to bed. More generally, if the system observes such a pattern of stimulants, user information module 185 can recommend reducing (and possibly eliminating) caffeine and similar alcohol intake (and if applicable, smoking or "dipping") management for a period of time.

[0212] In the above-mentioned bedtime application, the user information module 185 may remind the user to change the bedding and / or environmental conditions in his or her sleeping position.

[0213] In the above driving embodiment, the user information module 185 can provide specific personalized suggestions to the user. This can potentially improve safety. For example, the user information module 185 can recommend that the user seek alternative means of transportation, such as public transportation, carpooling (car sharing), etc. If the estimated or predicted fatigue state 180 indicates an increased risk of falling asleep while driving. Over time, fatigue is monitored, and geographic areas and their fatigue levels can be evaluated and associated with the possibility of commuting accidents. The user information module 185 can also recommend nap strategies based on the time of day, and provide appropriate warnings to avoid oversleeping (excessive napping).

[0214] In the travel embodiment, the user information module 185 may recommend that the user increase exposure to sunlight during daylight hours (if possible), or utilize fluorescent (full spectrum) lamps / light sources to prepare the user for the upcoming change in time zone.

[0215] In a gaming embodiment, the user information module 185 may remind the user to go to bed.

[0216] In a learning embodiment, the user information module 185 may act as an advice engine for healthy sleep habits to improve energy levels during study and testing times.

[0217] A benefit of CPAP therapy is reduced fatigue (e.g., reduced daytime sleepiness). An increased fatigue index in a CPAP patient may indicate less than optimal compliance with the prescribed CPAP therapy. Therefore, in a CPAP therapy embodiment of the fatigue monitoring and management system 100, the user information module 185 may provide suggestions to improve the patient's CPAP therapy compliance. Specifically, the education delivered by the user information module 185 (improved understanding of the association) may demonstrate the benefits of improved compliance to patients who are irregularly using their CPAP therapy. A high fatigue index is associated with low treatment compliance, and therefore, the user is motivated to improve his fatigue index as well as his compliance.

[0218] In the case where a CPAP patient suffers from insomnia, the patient may choose not to use their therapy. In this scenario, by delivering personalized sleep hygiene improvement suggestions, and optionally providing a relaxing breathing program (e.g., guided breathing delivered via an audio program executed on a personal computing device), the fatigue monitoring and management system 100 provides a means to enjoy better sleep, and therefore improve CPAP therapy compliance.

[0219] This fatigue assessment 180 can be used to infer whether a particular sequence of music used at bedtime is conducive to promoting sleep and affecting fatigue levels the next day (e.g., identifying "fatigue-busting" tracks).

[0220] This fatigue assessment 180 can be used to infer whether a particular breathing exercise promotes sleep or alertness (e.g., alertness breathing exercises use one or more self-adjusting lighting levels and dedicated audio sequences obtained via a light or display device to reduce breathing rate and adjust inhalation / exhalation times, and optionally biofeedback from a non-contact sleep sensor)

[0221] 8.1.16 Third-party information module 190

[0222] In a workplace embodiment of the fatigue monitoring and management system 100, the third-party information module 190 provides a report including information about sleep and fatigue of multiple employees to a third party, such as a corporate employer's website health consultant. The report is customizable and includes one or more of the following features:

[0223] Reports per employee (or employee group) per day

[0224] Fatigue risk metric based on factors including sleep history, time of day, and workload

[0225] Sleep mode change indication

[0226] Recent sleep history that triggers patterns of poor sleep quality or quantity

[0227] Actual sleep time - sleep duration, sleep quality, and sleep timing

[0228] Long-term trends – including estimates of long-term sleep debt, comparisons of individuals to groups (i.e., comparisons to the mean)

[0229] Balance of restorative (deep) and cognitive (REM) sleep

[0230] Sleep Disordered Breathing - Risk Assessment

[0231] Sleep inertia (waking up from deep or light sleep)

[0232] Other key measurements (therapy adherence, medication tracking, etc.)

[0233] In the bedtime embodiment, the third party information module 190 may act as a search tool for bedding manufacturers to explore users' average bedtimes and relative comfort levels (entered through subjective user data 145 and through objective sleep measurements 120).

[0234] 8.1.17 Sample Use Cases

[0235] A sample use case of the fatigue monitoring and management system 100 for a single user is as follows. The fatigue monitoring module 110 is implemented as a software module on a cloud server, communicating with a software application (app) on a smartphone, wirelessly connected via Bluetooth to a contactless sleep sensor and a wearable activity sensor for daily tracking.

[0236] Night #1: The user monitors their sleep using a contactless sleep sensor connected to the user's smartphone or other local memory storage device. The device transmits objective sleep measurements 120 such as bedtime, sleep latency (time to fall asleep), number of interruptions, wake times, and sleep score to the fatigue monitoring module 110. Snoring levels are also recorded during the night (either via the smartphone's microphone or via a microphone built into the contactless sleep sensor).

[0237] Day #1: In the morning, the user is presented with a Stop-Bang questionnaire because their snoring level intensity and number of sleep interruptions are considered to be of interest (above a threshold) to the fatigue monitoring module 110. The user is also prompted to enter their height and weight, and optional caloric expenditure data. A wearable activity sensor wirelessly connected to the user's smartphone captures the user's physical activity data 115 for subsequent transmission to the fatigue monitoring module 110. The user's exercise level is compared to typical healthy values. If "fatigue events" (e.g., yawning, closing eyes, etc.) are experienced during the day, the user is asked to record these events through an app running on their smartphone. Other subjective user data 145 collected by the smartphone includes whether the user commutes (and if so, whether the user drives), and whether the user naps during the day.

[0238] Night #2: The user receives first feedback from the user information module 185 via the smartphone, including some tips to improve sleep hygiene if necessary. This forms the basis for the user-specific fatigue calibration.

[0239] During the next week, ongoing comments are provided by the user information module 185.

[0240] Day #7: One week later (7 days, since it is ideal to capture a full work week if possible), fatigue state estimate 180 is presented to the user via user information module 185 or to the OH&S program manager via third party information module 190. Fatigue state assessment 180 may be utilized before the 7 day period has passed, especially if SDB patterns are detected.

[0241] The user's fatigue level is compared to group data 150, and these parameters are assumed to be the user's "true sleep age." Researchers in the field of sleep medicine have mapped the distribution of possible sleep stages as a function of age. Figure 5 Included is a graph 5000 (from Shambroom and Fabregas) representing an example distribution of sleep stages as a function of age. By comparing a user's actual sleep stage distribution to the population distribution, the user's sleep age can be determined.

[0242] Based on the fatigue index, as well as other lifestyle parameters that may be captured as subjective user data 145 by the fatigue monitoring and management system 100, a quality of life estimate may be derived by the fatigue monitoring module 110. Such parameters include caffeine intake, perceived stress and energy levels, and "state of mind" (mood) estimates.

[0243] 8.1.18 Example Performance

[0244] One embodiment of the fatigue monitoring and management system 100 utilizes data from 20 users collected over a 7-day period from the following data sources:

[0245] Pittsburgh Sleep Questionnaire and Quality of Life Survey (baseline data 155)

[0246] SleepMinder non-contact motion sensor (subjective sleep measurement 120 and SDB measurement 125)

[0247] HOBO Temperature Logger (Environmental Data 160)

[0248] FitBit activity sensor (physical activity data 115)

[0249] PVT (objective fatigue measure 130) performed four times daily via smartphone

[0250] VAS performed four times daily via smartphone (subjective user data 145)

[0251] Bedtime / wake-up time (manually entered into smartphone app)

[0252] Ambient audio recorded by the smartphone app (environmental data 160 ).

[0253] The linear classifier 4000 is used to implement the fatigue monitoring module 110. The linear classifier 4000 is trained to use 90% of the collected data, taking all sources except the PVT average reaction time score as input. The average PVT reaction time score for the first PVT of the day is calculated as a representative of the output fatigue index value. Next, the trained linear classifier 4000 is applied to the test data to predict the PVT average reaction time, which test data includes the remaining 10% of the data. Figure 7 Included are plots of predicted PVT mean reaction time (predicted fatigue index) versus actual PVT mean reaction time (actual fatigue index) on the training data (circles) and test data (crosses). The RMS prediction error for the test data is 49.3 milliseconds, while the RMS prediction error for the training data is 52.1 milliseconds, indicating that the linear classifier 4000 can generalize to unknown users. 2 is 0.70.

[0254] 8.2 Glossary

[0255] For purposes of this disclosure, in some forms of the invention, one or more of the following definitions may be employed. In other forms of the invention, alternative definitions may be employed.

[0256] Air: In some forms of the invention, the air provided to the patient may be atmospheric air, in other forms of the invention, the atmospheric air may be supplemented with oxygen.

[0257] Continuous Positive Airway Pressure (CPAP): The application of a breathable gas or air to the entrance of the airway at a continuous positive pressure relative to atmosphere, preferably substantially constant throughout a patient's breathing cycle. In some forms of CPAP therapy, the pressure will be varied during different breathing cycles of the patient, e.g., increasing the pressure in response to detecting an indication of partial upper airway obstruction and decreasing the pressure when no indication of partial upper airway obstruction is present.

[0258] Apnea: Apnea may be described as a pausing of breath that occurs when respiratory flow drops below a predetermined threshold for a sustained period of time (e.g., 10 seconds). Obstructive apnea may be described as an apnea that occurs when some obstruction to the airway does not allow air to flow regardless of the patient's effort. Central apnea may be described as an apnea that occurs when apnea is detected due to a reduction in respiratory effort or an absence of respiratory effort.

[0259] Respiratory Rate: The rate at which a patient breathes spontaneously, usually measured in breaths per minute.

[0260] Hypopnea: Preferably, hypopnea can be seen as a decrease in respiratory flow, but not a cessation of respiratory flow. In one form, when respiratory flow drops below a threshold for a sustained period of time, it is called hypopnea.

[0261] Patency (airway): The degree of openness of the airway, or how open the airway is. An open airway is wide open. Airway patency can be quantified, such as 1 for open and 0 for closed.

[0262] Respiratory flow, airflow, patient airflow, respiratory airflow (Qr): These synonyms can be understood to refer to the estimated respiratory airflow of the PAP device, as opposed to "true respiratory flow" or "true respiratory airflow", which is the actual respiratory flow experienced by the patient, usually expressed in liters per minute.

[0263] Upper airway obstruction (UAO): This includes partial and total upper airway obstruction. This may be associated with a flow-limited state, in which the flow level increases only slightly or may even decrease as the pressure difference across the upper airway increases (Starling obstruction behavior).

[0264] 8.3 Remarks

[0265] A portion of this invention contains material that is subject to copyright protection. The copyright owner has no objection to the facsimile reproduction by anyone of the patent text or patent disclosure, as it appears in the Patent and Trademark Office patent files or records, but otherwise reserves all copyright rights whatsoever.

[0266] Unless otherwise expressly stated herein, where a range of values ​​is provided, it is understood that each intervening value, tenth of the unit of the lower limit, between the upper and lower limits of the range, and any other intervening value stated or in the stated range is encompassed by the technology. The upper and lower limits of such intervening ranges, which may be independently included in the intervening ranges, are also encompassed by the technology, subject to any specifically excluded limits in the stated range. Where the stated range includes one or both limits, ranges excluding either or both limits are also encompassed by the technology.

[0267] Furthermore, where values ​​are described herein as implemented as part of the technology, it is understood that such values ​​may be approximate unless otherwise noted, and that such values ​​may be used in any suitable significant digits, within such ranges as may be allowed or required for actual technology implementations.

[0268] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art. Although any methods and materials similar or equivalent to the methods and materials described herein may also be used in practice or testing of the present invention, a limited number of exemplary methods and materials are described herein.

[0269] When a particular material is considered to be preferably used to form a component, an obvious alternative material with similar properties may be used as a substitute. In addition, unless otherwise specified, any and all components described herein are understood to be capable of being manufactured, and may be manufactured together or separately.

[0270] It must be noted that as used herein and in the appended claims, the singular forms "a," "an," and "the" include their plural equivalents unless the content clearly dictates otherwise.

[0271] All publications mentioned herein are incorporated by reference to the disclosure and description of the methods and / or materials that are the themes of these publications. The publications discussed herein are provided only for their disclosure prior to the filing date of the present application. No content herein is deemed to admit that the present invention does not enjoy the right prior to such publications due to prior inventions. In addition, the publication date provided may differ from the actual publication date, which may need to be independently confirmed.

[0272] In addition, when interpreting this disclosure, all terms should be interpreted in the broadest reasonable manner and consistent with the context. In particular, the term "comprising" should be interpreted as referring to elements, components or steps in a non-exclusive manner, indicating that the mentioned elements, components or steps may exist, be used or be combined with other elements, components or steps not expressly mentioned.

[0273] The subject headings used in the detailed description are only for the convenience of readers and should not be understood as limiting any subject matter in the present disclosure or claims. The subject headings should not be used to interpret the scope or limitation of the claims.

[0274] Although specific embodiments are described in the present technology, it should be understood that these embodiments are only used to illustrate the principles and applications of the present technology. In some embodiments, the terms and symbols refer only to specific details that are not necessary for the implementation of the present technology. For example, although the terms "first" and "second" may be used, unless otherwise specified, they are not used to illustrate any order, but can be used to distinguish different elements. In addition, although the process steps in the method can be described or illustrated in sequence, this order is not required. Those skilled in the art can appreciate that the order can be improved, and / or its various aspects can be executed simultaneously or even synchronously.

[0275] Therefore, it should be understood that many improvements may be made to the embodiments described, and other devices may be designed without departing from the spirit and scope of the present invention.

[0276] 8.4 List of reference numerals

[0277]

[0278]

[0279] 9. Citations

[0280] E., 2000. Dimensions of fatigue in different working populations. Scandinavian Journal of Psychology, 41: 231–241.

[0281] http: / / onlinelibrary.wiley.com / doi / 10.1111 / 1467-9450.00192 / pdf

[0282] Belenky,G.,et al.,2003.Patterns of performance degradationandrestoration during sleep restriction and subsequent recovery:a sleep dose-response study.J Sleep Res,vol.12no

[0283] 1.http: / / onlinelibrary.wiley.com / doi / 10.1046 / j.1365-2869.2003.00337.x / pdf

[0284] Dawson and Reid,1997.Fatigue,alcohol andperformanceimpairment.Nature,388:235.

[0285] Dinges DF,Pack F,Williams K,Gillen KA,Powell JW,Ott GE,etal.Cumulative sleepiness,mood disturbance,and psychomotorvigilanceperformance decrements during a week of sleep restricted to 4–5hourspernight.Sleep 1997;20(4):267–77.

[0286] Harrington,J.,1978.Shiftwork and Health:A Critical Review oftheLiterature.Report to the Medical Advisory Service,UK Health andSafetyExecutive.

[0287] Institute of Medicine(US)Committee on Sleep Medicine and Research;Colten HR,Altevogt BM,editors.Sleep Disorders and Sleep Deprivation:AnUnmetPublic Health Problem.

[0288] Washington(DC):National Academies Press(US);2006.4,FunctionalandEconomic Impact of Sleep Loss and Sleep-Related Disorders.Available from:

[0289] ]http: / / www.ncbi.nlm.nih.gov / books / NBK19958 / 63.

[0290] Jackowska et al.,2012.Sleep problems and heart rate variabilityoverthe working day,Journal of Sleep Research,Volume 21,Issue 4,February2012.

[0291] Nicholson PJ,D'Auria DA,1999.Shift work,health,the workingtimeregulations and health assessments.Occup Med(Lond).1999Apr;49(3):127-37.

[0292] http: / / occmed.oxfordjournals.org / content / 49 / 3 / 127.full.pdf

[0293] NSF(National Sleep Foundation),2013.

[0294] http: / / www.sleepfoundation.org / article / sleep-related-problems / excessive-sleepiness-and-sleep

[0295] Sasaki,T.,2007.Overtime,job stressors,sleep / rest,and fatigueofJapanese workers in a

[0296] company.https: / / www.jstage.jst.go.jp / article / indhealth / 45 / 2 / 45_2_237 / _pdf

[0297] Shahly V,Berglund PA,Coulouvrat C,et al.The Associations ofInsomniaWith Costly Workplace Accidents and Errors:Results From the AmericaInsomniaSurvey.Arch Gen Psychiatry.

[0298] 2012:69(10):1054-10

[0299] Shambroom JR,Fabregas SE.Age Related Changes in ObjectivelyMeasuredSleep Observed in a Large Population in the Home,available throughmyzeo.com.

[0300] State of Queensland,Department of Natural Resources and Mines,2013.Guidance Note for Fatigue Risk Management.

[0301] Trejo et al.EEG-Based Estimation of Mental Fatigue.

[0302] http: / / aiolos.um.savba.sk / ~roman / Papers / hci07_1.pdf

[0303] Zichermann,G.and Cunningham,C.(August 2011).IntroductiontoGamification by Design:Implementing Game Mechanics in Web and Mobile Apps(1sted.).Sebastopol,California:O'Reilly Media.

Claims

1. A system having one or more processors for monitoring a user's treatment compliance, the system comprising: a fatigue monitoring module configured to generate an assessment of the user's fatigue state based on data from one or more data sources, the one or more data sources including one or more sensors; as well as A user information module is configured to provide an improvement strategy to the user based on the fatigue level determined by the fatigue monitoring module, the improvement strategy being intended to improve the user's treatment compliance level related to treatment.

2. The system according to claim 1, wherein: The therapy compliance includes respiratory therapy compliance.

3. The system according to claim 1 or 2, wherein: The treatment compliance includes continuous positive airway pressure (CPAP) treatment compliance.

4. The system according to any one of claims 1 to 3, wherein: The user information module is configured to associate a high fatigue index with low treatment compliance.

5. The system according to any one of claims 1 to 4, wherein: The fatigue level includes a fatigue index, and the user information module is configured to generate a suggestion of suboptimal compliance with prescribed CPAP therapy based on an increase in the fatigue index.

6. The system according to any one of claims 1 to 5, wherein: The user information module is configured to generate recommendations for improving the user's CPAP therapy compliance level.

7. The system according to any one of claims 1 to 6, wherein: The user information module is configured to generate recommendations for improving or changing CPAP therapy.

8. The system according to any one of claims 1 to 7, wherein: The one or more sensors include a sleep sensor.

9. The system according to claim 8, wherein: The sleep sensor is a sensor integrated with a respiratory pressure therapy device.

10. The system according to claim 9, wherein: The respiratory pressure therapy device includes a motor driven blower.

11. The system according to any one of claims 1 to 10, wherein: The one or more processors are configured to provide a guided breathing procedure to the user based on a fatigue status indication to improve the therapy compliance.

12. The system according to claim 11, wherein: The guided breathing process includes an audio program executable on a personal computing device.

13. The system according to claim 11 or 12, wherein: The system includes a light or display device configured for self-adjusting illumination during breathing exercises to reduce breathing rate.

14. The system according to any one of claims 1 to 13, wherein: The fatigue monitoring module is implemented at a remote server, and the remote server is connected to the one or more data sources via a network.

15. The system according to any one of claims 1 to 14, wherein: The one or more sensors include a wearable activity sensor including an accelerometer.

16. The system according to any one of claims 1 to 15, wherein: The one or more sensors include a heart rate sensor.

17. A system according to any one of claims 1 to 16, wherein: The one or more sensors include a wearable photoplethysmographic sensor.

18. A system according to any one of claims 1 to 17, wherein: The one or more sensors include an earbud sensor.

19. The system according to any one of claims 1 to 18, wherein: The one or more sensors include a contactless motion sensor.

20. The system of claim 19, wherein: The non-contact motion sensor includes a radio frequency Doppler motion sensor.

21. A respiratory therapy device for providing therapy to a patient, the respiratory therapy device comprising: a motor-driven blower configured to provide positive airway pressure therapy through a patient interface; a sleep sensor for determining sleep measurements; Processing device; as well as monitor; Wherein the processing device is configured to generate a fatigue index and generate a graphical display on the display, the graphical display presenting a relationship between the fatigue index and the sleep measurement over multiple days.

22. The respiratory therapy apparatus of claim 21, wherein: The graphical display also presents the relationship between the fatigue index and the work measure.

23. The respiratory therapy apparatus of claim 22, wherein: The graphical display presents changes in the fatigue index over a period of time of the work measurement.

24. A respiratory therapy apparatus according to any one of claims 21 to 23, wherein: The fatigue index is determined based on the work measure and the sleep measure.

25. A device for monitoring sleep, comprising: One or more data sources, including: an unobtrusive sleep sensor configured to generate an objective sleep measurement of the user; and a fatigue monitoring module of the processor configured to generate an assessment of a fatigue state of the user based on data from the one or more data sources; and A processing device is configured to generate an output of the sleep age of the user based on the output of the fatigue monitoring module.

26. The apparatus of claim 25, wherein: The processing device generates the sleep age based on a comparison of objective sleep and estimated fatigue parameters with standard demographic data.

Images