Patent Application
20220020492
The present invention relates to the field of sleep analysis, and in particular, to the field of sleep analysis performed by an individual without clinician input.
The use of (wearable) sleep trackers is being more popular and widespread, and are often used to perform sleep analysis or assessment without the input of a clinician/physician.
Sleep trackers often execute a software application, running on the processor, to perform a sleep analysis or assessment process, e.g. based upon biometric data gathered by the sleep tracker itself or an external sensor, such as movement data from an accelerometer (e.g. mounted on the individual's wrist) or heartrate data from a heartrate monitor.
However, despite their widespread use, the accuracy of sleep trackers has been brought into question. Studies evaluating the ability of wearable trackers to estimate sleep have generally found high sensitivity, comparatively low specificity, and invalidity around wake times during sleep. Additionally, even when users are presented with accurate sleep tracking data within a healthy normal range, users may lack the domain knowledge to appropriately interpret the information provided to them.
Moreover, a disconnect between sleep quality information provided by a sleep tracker and a subjective analysis of sleep quality performed by the individual has been observed. This is at least partly because an individual's subjective evaluation of their sleep quality is largely dependent on sleep onset latency (SOL) and the frequency and duration of wakefulness during the night, estimates of which vary from objective measures obtained using a sleep tracker.
This disconnect can cause anxiety and concern in an individual who uses a sleep tracker, and there is an ongoing desire to address this problem.
The invention is defined by the claims.
According to examples in accordance with an aspect of the invention, there is provided a computer-implemented method for generating a predictive indicator that indicates a likelihood of orthosomnia in an individual.
The computer-implemented method comprises: obtaining user interaction data, being data responsive to the individual's interactions, via a user interface for a processor, with a software application running on the processor for performing a sleep analysis or assessment process; and processing the obtained user interaction data to generate the predictive indicator that indicates a likelihood of orthosomnia in the individual.
Orthosomnia is a condition characterized by a preoccupation or obsession with improving or perfecting sleep quality data generated by a (computer-executed) sleep analysis or assessment process, and in particular to sensor-derived sleep quality data (being a prediction of sleep quality of the individual generated based on biometric/physiological data of the individual).
The present disclosure recognizes that a user interaction with a software application running or performing a sleep analysis/assessment process can be indicative of an individual's preoccupation with data generated by the software application. In particular, a manner in which a user interacts with the software application, or the information that is supplied to the software application by the user, is indicative of possible orthosomnia in the individual.
The present disclosure proposes an approach for identifying possible orthosomnia in the individual, by assessing/analyzing user interaction data responsive to the user's interaction with the software application. The user's interaction is via a user interface, which may be formed in the same device that comprises the processor that carries out or runs the software application (e.g. a mobile/cellular phone) or a separate device (e.g. the processor is a distributed processing system and the user interface is a mobile/cellular phone in communication with the distributed processing system).
In one example, the computer-implemented method comprises displaying the predictive indictor via the user interface. This provides an individual/clinician with additional information to aid them in performing a treatment/diagnostic process.
In some examples, the user interaction data comprises access data responsive to the individual accessing and/or opening the software application. This embodiment recognizes that the characteristics of the user accessing or opening the software application are indicative of whether or not individual has orthosomnia.
For instance, the access data may comprise one or more measures of accessing and/or opening frequency and/or duration of access. This embodiment recognizes a link between the amount of time that a user accesses the software application (and/or a frequency of access) and likelihood of orthosomnia. In particular, it is recognized that an individual who frequently accesses the software application, or accesses the software application for a long period of time, is more likely to have orthosomnia than an individual who less frequently, or for shorter periods of time, accesses the software application.
Frequent and/or long accesses thereby indicate a high likelihood of orthosomnia. Low and/or short accesses similarly indicate a low likelihood of orthosomnia.
The one or more measures may relate to a time or frequency of accessing particular parts, screens or displays of the software application, e.g. a part that displays sleep quality data. It is recognized that a preoccupation with certain elements of the software application (such as the quality data) may be indicative of potential orthosomnia.
In some examples, the step of processing the obtained user interaction data comprises: obtaining population access data, being data that is responsive to population trends of other individual's interactions with the software application running on the processor and/or one or more other versions of the software application running on one or more other processors; and comparing the access data to the population access data to generate the predictive indicator.
The population access data may thereby provide a benchmark, baseline or average expected level of accesses to the software application. In particular, a population trend of other individual's interactions with the software application (or other instances/versions of the software application) provides information establishing a baseline against which a likelihood of orthosomnia can be predicted. For instance, a strong (positive) deviation from a mean of a population could indicate the individual is accessing the software application more commonly than expected, indicating possible orthosomnia.
This approach thereby provides an accurate mechanism for assessing a likelihood of orthosomnia in the individual.
The access data may comprise time data indicating a time (of day) at which the individual interacts with the software application. The time data may, for example, comprise one or more time stamps or other similar data structures.
In some examples, this time data may be processed by itself to generate the predictive indicator. This embodiment recognizes that a time at which an individual interacts with the software application (as a whole, or with certain parts of the software application) may be indicative of possible orthosomnia. For instance, an individual is more likely to have orthosomnia if they interact with the software application during the night (e.g. as this would demonstrate an anxiety about their quality of sleep) or repeatedly throughout the day (e.g. as this would indicate a preoccupation with their purported quality of sleep).
In some examples, the step of processing the obtained user interaction data comprises: obtaining, from a biometric sensor, biometric data of the user, the biometric data being responsive to changes in one or more physiological parameters of the individual at or during a time at which the individual accesses and/or opens the software application, as indicated by the time data; and processing the biometric data to generate the predictive indicator.
This embodiment recognizes that a physiological response of the individual during an accessing of the software application is indicative of potential orthosomnia. For instance, if the individual reacts nervously or negatively during access of the software application, this could indicate potential orthosomnia.
The one or more physiological parameters of the individual comprise a physiological parameter responsive to an autonomic response of the individual, such as a heartrate, sweating, a temperature, a respiratory rate, skin color, eye movement and/or an eye dilation. These embodiments recognize that an involuntary response (autonomic response) may indicate a nervousness or concern with the information provided by the software application, and thereby indicates a likelihood of orthosomnia. The proposed approach thereby provides a user independent mechanism, i.e. an objective mechanism, for predicting a likelihood of orthosomnia.
In at least one embodiment, the step of processing the obtained user interaction data comprises: obtaining historic access data, being data responsive to the individual's historic interactions, via the user interface, with the software application; and comparing the access data to the historic access data to generate the predictive indicator.
A change in how a user accesses or opens the software application over time may indicate a change in potential orthosomnia. For instance, if a user is accessing the software application more frequently than in the past, then this may indicate that the probability of the user having orthosomnia is increasing.
In some examples, the step of obtaining the user interaction data comprises obtaining, via the user interface, user-derived sleep quality data representing a subjective perception of sleep quality of the individual.
The step of processing the obtained user interaction data may comprise: obtaining, from a sleep sensor, sleep sensor data responsive to one or more changes of physiological parameters of the individual during the individual's sleep; processing the sleep sensor data to generate sensor-derived sleep quality data, being data representing an objective measure of the sleep quality of the individual; and comparing the user-derived sleep quality data to the sensor-derived sleep quality data to generate the predictive indicator.
This embodiment recognizes that a difference between sensor-derived sleep quality data and user-derived sleep quality data indicates potential orthosomnia or an increased risk of orthosomnia. In particular, if there is a large difference between an individual's perception of their sleep, and the quality of sleep derived from sensor data, then there is an increased likelihood that the individual will become anxious about the quality data produced by the software application—which may distort their own perception of their sleep quality, leading to orthosomnia.
There is also proposed a computer-implemented method of controlling a software application running on a processor for performing a sleep analysis or assessment process.
The computer-implemented method comprises: generating a predictive indicator that indicates a likelihood of orthosomnia in an individual by performing any suitable herein described method; and controlling the software application to modify and/or supplement information presented to the individual, via the user interface, during the sleep analysis or assessment process based on the predictive indicator.
This approach recognizes that the effects of orthosomnia can be mitigated or reduced by controlling the information presented to the individual (at the user interface) by the software application. Through appropriate control of the software application, the individual can be provided with suitable information to assuage their concerns or anxieties and/or reduce the likelihood of orthosomnia in the individual.
The skilled person would appreciate that, in another aspect of the inventive concept, the method of generating the predictive indicator (for the method of controlling the software application) may be carried out by a process not disclosed in this specification. For instance, the predictive indicator may be an indicator received from a clinician (indicating a diagnosis of orthosomnia) and/or the individual (e.g. if they are concerned they may have orthosomnia).
In some examples, the software application is configured to generate and display, at the user interface, a sleep quality measure during the sleep analysis or assessment process.
The step of controlling the software application may comprise modifying the sleep quality measure in response to the predictive indicator indicating that a likeliness of orthosomnia in the individual falls within a first predetermined range.
The step of controlling the software application may comprise suppressing the display of the sleep quality measure in response to the predictive indicator indicating that a likeliness of orthosomnia falls within a second predetermined range, which may be different to the first predetermined range (if present).
The step of controlling the software application may comprise providing, at the display, supplementary information about sleep quality measures in response to the predictive indicator indicating that a likeliness of orthosomnia falls within a third predetermined range, which may be different to the first and/or second predetermined range (if present)—but may overlap the first and/or second predetermined range in other examples.
Any combination of these approaches could be used. These embodiments recognize that the information presented by the software application may affect a likelihood the individual having orthosomnia. By changing/suppressing/supplementing the information, a level of control can be exerted over the likelihood of the individual having orthosomnia, e.g. reducing a likelihood or intensity of the individual's orthosomnia.
In some embodiments, the step of obtaining the user interaction data comprises obtaining, via the user interface, user-derived sleep quality data representing a subjective perception of sleep quality of the individual.
The computer-implemented method may further comprise: obtaining, from a sleep sensor, sleep sensor data responsive to one or more changes of physiological parameters of the individual during the individual's sleep; and processing the sleep sensor data to generate sensor-derived sleep quality data, being data representing an objective measure of the sleep quality of the individual, wherein the supplementary information comprises information responsive to a difference between the user-derived sleep quality data and the sensor-derived sleep quality data.
It is recognized that explanative information that can explain or justify a difference between an individual's perception of their sleep, and a software applications prediction of the quality of their sleep, can help reduce the likelihood and/or intensity of orthosomnia in the individual. This is because the user is provided with information that aids them in their understanding of their clinical state, and therefore reduces the likelihood of the user developing, or furthering, orthosomnia.
In some examples, the computer-implemented method is performed by the software application. Of course, the computer-implemented method could be performed by a different software application (e.g. supplementary to the software application).
In some examples, the predictive indicator comprise a binary indicator indicating a prediction of whether or not the individual has orthosomnia. A binary indicator is a data point having only two possible value. In other examples, the predictive indicator comprises a categorical indicator (e.g. data having a discrete or finite number of possible values), a continuous indicator (e.g. data having a non-finite number of possible values—or the maximum number of possible values for a particular computing environment) or a numerical indicator (e.g. a probability).
There is also proposed a computer program product comprising computer program code means which, when executed on a computing device having a processing system, cause the processing system to perform all of the steps of any herein described method.
There is also proposed a processing system for generating a predictive indicator that indicates a likelihood of orthosomnia in an individual. The processing system is configured to obtain user interaction data, being data responsive to the individual's interactions, via a user interface for a processor, with a software application running on the processor for performing a sleep analysis or assessment process; and process the obtained user interaction data to generate the predictive indicator that indicates a likelihood of orthosomnia in the individual.
The skilled person would be readily capable of adapting any herein described processing system to carry out any herein described method and vice versa.
These and other aspects of the invention will be apparent from and elucidated with reference to the embodiment(s) described hereinafter.
For a better understanding of the invention, and to show more clearly how it may be carried into effect, reference will now be made, by way of example only, to the accompanying drawings, in which:
The invention will be described with reference to the Figures.
It should be understood that the detailed description and specific examples, while indicating exemplary embodiments of the apparatus, systems and methods, are intended for purposes of illustration only and are not intended to limit the scope of the invention. These and other features, aspects, and advantages of the apparatus, systems and methods of the present invention will become better understood from the following description, appended claims, and accompanying drawings. It should be understood that the Figures are merely schematic and are not drawn to scale. It should also be understood that the same reference numerals are used throughout the Figures to indicate the same or similar parts.
The invention provides a mechanism for identifying orthosomnia, being the preoccupation with sensor-derived sleep measures or automated sleep analysis processes, within an individual. Interactions between the individual and a software application performing a sleep analysis process are monitored and used to generate a predictive indicator that indicates a likelihood that the individual has orthosomnia. This predictive indicator may be used to control the software application, and in particular, to control the display of information provided by the software application.
Embodiments are based on the realization that interactions between the individual and the software application, and in particular the manner in which the individual interacts with the software application, is indicative of whether or not the individual has orthosomnia. In this way, a predictive indicator can be generated to aid in the treatment and/or diagnosis of the individual, i.e. act as an aid to a clinical decision making process.
This proposed concepts aid in the identification, management and relief of orthosomnia.
The increasing use of wearable sleep tracking devices allows individuals to monitor sleep patterns objectively at a consumer level. However, with access to this objective, albeit often inaccurate, data—“sensor-derived sleep quality data”—some individuals can (mistakenly, but unintentionally) identify and self-diagnose sleep disturbances, for which they proceed to seek medical treatment. Additionally, individuals may lack the knowledge to appropriately interpret sensor-derived sleep quality data as being within a healthy spectrum or as being indicative of a sleep disorder. For a subset of individuals, the use of sleep trackers has unintended, adverse effects, as individuals attempt to increase sleep scores or eliminate undesired disturbances in their data through methods which often reinforce poor sleep habits, including spending excessive time in bed (causing a decrease in sleep efficiency) in order to attempt to increase sleep duration. There is a recognized condition of orthosomnia, being the preoccupation with improving or perfecting sleep data. Orthosomnia can result in a harmful negative feedback loop as an individual's stress about their sleep causes their sleep quality to further deteriorate.
This condition was recognized by K. G. Baron et al in the journal article Baron, K. G., Abbott, S., Jao, N., Manalo, N. and Mullen, R., 2017. Orthosomnia: Are some patients taking the quantified self too far?. Journal of Clinical Sleep Medicine, 13(2), pp. 351-354.
In some cases of orthosomnia, individuals only identify perceived sleep issues after monitoring sensor-derived sleep quality data and finding instances of reported restlessness or low sleep efficiency, of which they were never previously aware, or due to lack of familiarity with a sleep scoring system. To individuals having orthosomnia, sensor-derived sleep quality data can feel more consistent with their experience of sleep than validated techniques, such as polysomnography or actigraphy. As a result, individual's perceptions can prove difficult to alter. This can negatively affect treatment because individuals can be reluctant to make changes that would decrease or affect their sensor-derived sleep quality data, such as refusing to try sleep restriction (or sleep consolidation therapy) because it would temporarily lower sleep duration. This is compounded by the individual's awareness of the lack of trust by clinician in sensor-derived sleep quality data, which affects the confidence that the individual may have in the clinician.
This invention is intended to provide approaches for predicting a likelihood of orthosomnia, and proposes a number of approaches for managing and relieving orthosomnia in an individual.
In particular examples, approaches facilitate improved understanding (by at least the individual) of the sensor-derived sleep quality data, placing value on their subjective sleep experience rather than solely relying on objective reports of potentially questionable accuracy.
The processing arrangement 100 comprises a user interface 110, e.g. comprising a display and an input interface. The user interface 110 is able to communicate with a processor 120. The user interface 110 and the processor 120 may, as illustrated, be housed in a same device 105 (e.g. a smartphone, a smartwatch, a laptop and so on). In other examples, the user interface 110 and the processor 120 are housed separately. For instance, the processor 120 may be a distributed or cloud-computing system.
The processor 120 is configured to run, operate or execute a software application running for performing a sleep analysis or assessment process. Typically, a sleep analysis process comprises obtaining biometric data of the individual, e.g. from a sensor 130, and processing the biometric data to generate sensor-derived sleep quality data. The sensor-derived sleep quality data represents an objective measure of the quality of the individual's sleep, such as a sensor-derived quality measure score (i.e. a “sleep quality measure”), a total sleep time, a number of sleep interruptions, length of deep sleep (e.g. non-REM sleep) and so on.
In some examples, processing the biometric data to generate the sensor-derived sleep quality data may involve smoothing areas known to be unreliable according to wearable data or areas known to be habitually relevant to the individual. For example, if it is known that the user typically reads for one hour before bed, their total sleep time (TST) is likely to be overestimated by the sensor-derived sleep quality data and should be adjusted.
In some examples, additional information may be used in the generation of the sensor-derived sleep quality data. The additional information may, for example, comprise user information, such as an age, gender, health conditions and diagnosed sleep disorders and/or typical/historic sleep patterns of the user. The additional information may comprise information about the sleep sensor and/or software application (e.g. to take account of any known measuring errors or inconsistencies) and/or information about the sleep environment (such as a noise level, light level, etc.).This additional information may be provided by a user at the user interface 110.
Approaches for generating or deriving quality data from biometric data of an individual (obtained at a sensor 130), and optionally additional information, are well known, and are not further described for the sake of conciseness. Approaches may comprise using a machine-learning algorithm (e.g. a neural network, a naïve Bayesian classifier or a support-vector machine) to generate the sensor-derived sleep quality data from the biometric data.
Some suitable examples of generating sensor-derived sleep quality data can be found in the US patents having publication numbers U.S. Ser. Nos. 10/624,574 and 10/524,674, or the US patent applications having publication numbers U.S. Ser. Nos. 16/495,687 and 16/589,181. Other suitable approaches will be well known.
The processor 120 may be adapted to control the user interface 110 to display a visual representation of the sensor-derived sleep quality data. In particular, the individual may be able to access or open the software application, and view the visual representation of the sensor-derived sleep quality data.
Thus, the software application is configured so that an individual is able to access and/or view the sensor-derived sleep quality data.
The sensor 130, i.e. a “sleep sensor”, is any sensor that is able to monitor one or more physiological parameters indicative of a change in a user's quality of sleep. For instance, the physiological parameter(s) may include one or more of: a heartrate; a movement; a respiratory rate; brain wave activity and so on. The sensor 130 may be formed in the same device 105 as the user interface 110 and/or the processor 120. Preferably, at least the sensor 130 is wearable by the user, for example, the whole device 105 may be wearable.
Other suitable configurations for a processing arrangement 100 will be readily understood by the skilled person, e.g. employing external servers, external processing systems or communications channels between different aspects of the processing arrangement.
The method 200 comprises a step 210 of obtaining user interaction data 215. The user interaction data is data responsive to the individual's interactions, via a user interface for a processor, with a software application running on the processor for performing a sleep analysis or assessment process.
The user interaction data may comprise access data, being data responsive to the user accessing and/or opening the software application. The access data may comprise one or more numerical values responsive to a user accessing and/or opening the software application, e.g. where a value represents a time of accessing/opening (e.g. a timestamp or the like), a frequency of accessing/opening (e.g. how many accesses/openings per predetermined time period, e.g. per day or in a last predetermined time window, e.g. a last hour or last two hours), a cumulative count of accessing/opening, a duration of an access (e.g. in seconds) and so on.
In at least one embodiment, the access data comprises time data indicating a time (of day) at which the individual interacts with the software application. This may be in the form of one or more time stamps, or other similar measures. In some examples, the access data comprises one or more measures of accessing and/or opening frequency and/or duration of access.
For the avoidance of doubt, the “access data” may comprise one or more measures of accessing and/or opening frequency and/or duration of access of one or more parts of the software application (e.g. accesses particular screens or data available with the software application.
In some examples, the user interaction data comprises user-derived sleep quality data representing a subjective perception of sleep quality of the individual, i.e. data that indicates how the individual feels about their sleep quality (preferably without being influenced by the sensor-derived sleep quality data). The user-derived sleep quality data may, for instance, comprise responses to a survey or a user-provided indication of perceived quality of sleep (e.g. a numeric indicator on a predetermined scale). The user-derived sleep quality data is obtained from a user input provided at the user interface.
In other examples, the user interaction data comprises one or more other metrics of an interaction between the individual and the software application. By way of example only, the user interaction data may comprise one or more values representing a number of times the individual views a particular piece of data (e.g. a particular screen or the sensor-derived sleep quality data, e.g. in the form of a sleep hypnogram) provided by the software application, a number of link (e.g. to articles) clicked within the software application and so on.
The method 200 then moves to a step 220 of processing the obtained user interaction data to generate the predictive indicator that indicates a likelihood of orthosomnia in the individual.
In the context of the present invention, a predictive indicator is any data that changes responsive to changes in a predicted likelihood that the individual has orthosomnia. The predictive indicator may comprise binary, categorical or numerical data. Binary data may indicate a prediction as to whether or not the individual has orthosomnia (e.g. “0” indicates predicted absence and “1” indicate predicted presence or vice versa). Categorical data may indicate a likelihood category (e.g. “Likely”, “Very Likely”, “Unlikely” and so on) that the individual has orthosomnia. Numerical data may indicate a numeric probability that the individual has orthosomnia, e.g. on a scale of 0 to 1 or 0 to 100.
The precise mechanism for generating the predictive indicator in step 120 may depend upon the content of the user interface data.
In some examples, where the user interaction data comprises access data, the access data may be compared to one or more thresholds to determine whether or not the individual has orthosomnia. The one or more thresholds may, for example, be derived from population data (e.g. indicating a population average and/or standard deviation for the access data).
By way of example, an individual may be considered to have orthosomnia if they access or open the software application for more than 1 minute a day and/or more than 5 days a week. As another example, an individual may be considered to have orthosomnia if they access or open the software application (e.g. more than a predetermined number of times, e.g. more than 3 times) during a predetermined time window (e.g. which may correspond to a time window when the individual is expected to be asleep).
As another example, one or more predetermined thresholds may be set based on population access data. Thus, step 220 may comprise obtaining population access data, being data that is responsive to population trends of other individual's interactions with the software application running on the processor and/or one or more other versions of the software application running on one or more other processors; and comparing the access data to the population access data to generate the predictive indicator.
The population access data may, for instance, define some population average and/or standard deviation.
An individual may be considered to have orthosomnia if their access data breaches some predetermined threshold with respect to population access data. For instance, if the access data comprises a value representing a frequency of access (and the population data indicates a population mean and/or standard deviation of frequency of access), the individual may be considered to have orthosomnia if their frequency of access is more than 2 or 3 standard deviations greater than the population mean average of frequency.
In some examples, a value for the predictive indicator may be set based on a difference between the access data and the population access data, e.g. a difference between the frequency of access and the average (population) frequency of access.
Other predetermined thresholds and/or approaches for setting the predetermined thresholds could be apparent to the skilled person. For instance, the predetermined thresholds may be set according to some clinical protocol and/or in response to one or more individuals/clinicians' input.
In some examples, a machine-learning algorithm could be used to generate the predictive indicator. The machine-learning algorithm may receive, as input, the user interaction data (e.g. the access data) and provide, as output, the predictive indicator. Examples of suitable machine-learning algorithms include a neural network, a naïve Bayesian classifier and a support-vector machine.
As one example, the predetermined thresholds may be set based on historic information or access data of the individual. For instance, a predetermined threshold may be based upon an average (mean) and/or standard deviation of values contained in historic access data, e.g. access data obtained over a previous or past week or month (although other time periods are also considered).
More generally, in some examples, the step of processing the obtained user interaction data comprises: obtaining historic access data, being data responsive to the individual's historic interactions, via the user interface, with the software application; and comparing the access data to the historic access data to generate the predictive indicator.
In some embodiments, the access data comprises time data indicating a time at which the individual interacts with the software application, e.g. in the form of one or more time stamps. In these examples, the step of processing the obtained user interaction data may comprise obtaining biometric data of the user, the biometric data being responsive to changes in one or more physiological parameters of the individual at or during a time at which the individual accesses the and/or opens the software application, as indicated by the time data.
The biometric data may then be processed to generate the predictive indicator, e.g. using predetermined thresholds, comparisons to population/historic data and/or machine-learning methods.
In this way, biometrics associated with the utilization of the software application are used to detect the probability of orthosomnia. This embodiment recognizes that a biological response of the individual can indicate whether or not the user has orthosomnia (e.g. based on an assessment as to whether they are nervous/stressed when accessing the software application).
In particular, biometric data may comprise physiological data (i.e. measurements) that represent an unconscious or autonomic response of the individual to information provided by the software application (e.g. at a user interface). Thus, the physiological data may be responsive to an autonomic response of the individual (during the time indicating by the time data).
Examples include a heartrate (or other heart characteristics, such as an acceleration of a heartrate), sweating, a temperature, a respiratory rate, skin color, eye movement and/or an eye dilation. Other suitable physiological parameters indicative of an autonomic response and/or nervousness/stress of an individual will be apparent to the skilled person.
The biometric data may be compared to one or more thresholds, e.g. set by a clinician or based on population data, to generate the predictive indicator. The threshold may be defined to identify an autonomic response indicating nervousness or stress when accessing the software application (herein identified as being a sign/symptom of orthosomnia). For instance, if the biometric data comprises a measured heartrate, the predictive indicator may indicate that the individual is likely to have orthosomnia if their measured heartrate exceeds some predetermined threshold when accessing the software application. As another example, if the biometric data comprises an eye dilation, the predictive indicator may indicate that the individual is likely to have orthosomnia if their measured eye dilation (e.g. detectable with a camera or the like) exceeds some predetermined threshold.
The biometric data may be compared to historic biometric data. For instance, if the biometric data comprises a measured respiratory rate, an increase in respiratory rate when the user access the software application may indicate a likelihood of orthosomnia (as this may indicate nervousness about the results of the sleep analysis/assessment).
The biometric data may be processed using a machine-learning method configured to generate the predictive indicator. Examples of machine-learning methods have been previously described.
In yet another embodiment, usable when the user interaction data comprises user-derived sleep quality data, the user-derived sleep quality data may be compared to sensor-derived sleep quality data to generate the predictive indicator. In this way, generating the predictive indicator may include an assessment of the mismatch between a subjective perception of sleep quality (the user-derived sleep quality data) and objective sleep tracker data (the sensor-derived sleep quality data).
The user-derived sleep quality data may, for example, comprise one or more a numerical measure or indicator of a perceived quality of sleep provided by the individual (e.g. on a scale of 0 to 10, 1 to 10, 0 to 100 or 1 to 100). The user-derived sleep quality data may comprise a plurality of such measures/indicators, e.g. representing perceived quality for different aspects of sleep (e.g. perceived sleep depth, sleep consistency, wakefulness and so on).
Thus, as illustrated, step 220 may comprise a step 221 of obtaining, from a sleep sensor, sleep sensor data 225 responsive to one or more changes of physiological parameters of the individual during the individual's sleep. The sleep sensor data may, for instance, comprise biometric data of the individual.
The sleep sensor may be external to the processing system performing the method 200, or may be housed in a same device as the processing system.
The step 220 may also comprise a step 222 of processing the sleep sensor data to generate sensor-derived sleep quality data, being data representing an objective measure of the sleep quality of the individual. Preferably, the sensor-derived sleep quality data provides sleep sensor derived predictions of the same values or measures (e.g. on a same scale) contained in the user-derived sleep quality data. Mechanisms for generating sensor-derived sleep quality data have been previously described with reference to
The step 220 may then perform a step 223 of comparing the user-derived quality input data to the sensor-derived sleep quality data to generate the predictive indicator.
The user-derived sleep quality data and the sensor-derived sleep quality data may be normalized with respect to one another. For instance, both types of quality data may include a measure of sleep quality on a same scale for the purposes of comparison. The predictive indicator may be responsive to a difference in the measure of sleep quality.
Step 223 may comprise determining a difference between the user-derived quality input data and the sensor-derived quality input data, and processing this difference to generate the predictive indicator. This may include processing the difference using predetermined thresholds, comparisons to population/historic data and/or machine-learning methods.
If possible, information used to determine the user-derived sleep quality data should be collected before the individual sees (e.g. at the user interface) any sensor-derived sleep quality data (e.g. graph, sleep score, etc.). This is because the sleep perception of indiviudals with orthosomnia may be affected by their objective sleep data, which would affect the reliabiltiy of the user-derived sleep quality data.
Proposed embodiments thereby facilitate generation of a predictive indicator responsive to user interaction data which represents information obtained from a user's interaction with a software application (for sleep analysis/assessment). The user interaction data may comprise information intentionally supplied by the user (e.g. usable to generate user-derived sleep quality data) and/or metadata concerning a user's interaction with the software application (e.g. information about a time, frequency and/or duration of access).
A combination of the above described approaches used for generating the predictive indicator may be employed.
In some embodiments, a combination of user interaction data and data derived from a sleep sensor (“sleep sensor data”) may be used to generate the predictive indicator. In other words, the predictive indicator may be responsive to a combination of user interaction data and sensor-derived data.
For instance, user interaction data may comprise a frequency with which the individual access the software application (or certain parts of the software application), and the sleep sensor data may comprise a predicted length of sleep (or time in bed)—e.g. by assessing movement of the individual. The predictive indicator may indicate likely orthosomnia if the user interaction data indicates an increased frequency of checking the software application (or certain parts thereof) and an increased level of sleep (e.g. attempts to obtain 9 or 10 hours of sleep). This indicates that the user is attempting to get more time asleep, and may therefore have orthosomnia.
Other suitable examples will be apparent to the skilled person, e.g. if the sleep sensor data indicates increased restlessness (during sleep) accompanied by frequent access of the software application during the night (e.g. more than 3 times during the night), indicating that the individual is stressed about their sleep during the night.
In some examples, the computer-implemented method further comprises a step (not shown) of displaying the predictive indictor via the user interface. This provides an individual/clinician with additional information to aid them in performing a treatment/diagnostic process.
This disclosure also proposes mechanisms to perform in response to the predictive indicator indicating that the user has, or is likely to have, orthosomnia. In particular, the disclosure proposes an approach for controlling the software application to modify and/or supplement information presented to the individual, via a user interface, during the sleep analysis or assessment process based on the predictive indicator.
The purpose of these mechanisms is the reduction of an individual's stress and anxiety related to sleep scores and the desire to achieve perceived perfect sleep.
Thus, there is proposed a computer-implemented method of controlling a software application running on a processor for performing a sleep analysis or assessment process.
Various approaches for controlling the software application are envisaged in the present disclosure if the individual is suspected of orthosomnia, e.g. if the predictive indicator indicators that the user is likely to have orthosomnia. These approaches are particularly useful if the software application is configured to generate and display, at the user interface, a sleep quality measure—which is derived from data obtained from a sleep sensor (i.e. is a sensor-derived sleep quality measure).
In one example, the method 300 comprises a step 310 of controlling the software application to modify the sleep quality measure (displayed at the user interface) in response to the predictive indicator indicating that a likeliness of orthosomnia in the individual falls within a first predetermined range.
Thus, instead of simply displaying a sleep quality measure as derived from sensor data, the sleep quality measure may be revised to take account of likely orthosomnia of the individual, e.g. making the sleep quality measure more positive (e.g. revising a sleep quality measure upwards) to reduce anxiety or stress on the individual. This step effectively aims to obfuscate the sleep quality measure based on the sleep tracker analysis algorithm, to reduce the impact of the sleep quality measure data on the individual.
In another example, the method 300 comprises a step 320 of suppressing the display of the sleep quality measure in response to the predictive indicator indicating that a likeliness of orthosomnia falls within a second predetermined range. Thus, a sleep quality measure may prevent a sleep quality measure from being generated and presented to the individual.
The display of the sleep quality measure may, for instance, be suppressed for a predetermined period of time (e.g. 3 days or a week) and/or during certain time period or times of the day (e.g. during a time during which the individual is supposed to be asleep).
This approach aims to address an individual's obsession with the sensor-derived sleep quality data, to reduce anxiety and/or stress about the sensor-derived sleep quality measure if they are predicted to have orthosomnia. In particular, this approach is intended to encourage the individual to avoid behavioral habits that may lead to, or reinforce existing, orthosomnia (i.e. frequent software application checking and the like).
In some examples, the step of suppressing the display of the sleep quality measure may comprise removing access to, or preventing the individual from accessing, the software application, e.g. for a predetermined period of time (e.g. 3 days or a week) and/or during certain time periods (e.g. during the night).
In some examples, the method 300 comprises a step 330 of (generating and) providing, at the display, supplementary information about sleep quality measures in response to the predictive indicator indicating that a likeliness of orthosomnia falls within a third predetermined range.
The purpose of the supplementary information may be to divert the individual's attention from a sleep quality measure and/or decatastrophize a poor sleep quality measure.
In particularly preferable examples, the supplementary information is responsive to a difference between user-derived sleep quality data and the sensor-derived sleep quality data. This approach can help emphasize potential discrepancies between the user's reported experiences and suggesting plausible explanations for such differences.
For instance, if a difference between user-derived sleep quality data and the sensor-derived sleep quality data is identified, the supplementary information may generate text indicating that the sensor-derived data may be inaccurate (e.g. as it differs significantly from the user-derived data). In particular, orthosomnia can relate to a scenario in which the user has a subjective feeling of high sleep quality, but the sensor-derived sleep quality data indicates that the user had a low sleep quality. The supplementary information may provide plausible explanations for this inaccuracy (e.g. incorrect placement of the sleep sensor, over or under-sensitivity of the sleep sensor and so on). This approach can help assuage an individual's concerns, whilst maintaining their confidence in the method of the present disclosure.
Accordingly, step 330 may comprise a step 331 of obtaining, from a sleep sensor, sleep sensor data 225 responsive to one or more changes of physiological parameters of the individual during the individual's sleep; and a step 332 of processing the sleep sensor data to generate sensor-derived sleep quality data, being data representing an objective measure of the sleep quality of the individual; and a step 333 of generating the supplementary information, which comprises information responsive to a difference between the user-derived sleep quality data 215 (e.g. obtained by process 200) and the sensor-derived sleep quality data. Steps 331 and 332 may be functionally equivalent to steps 221 and 222 (described with reference to
In some examples, the supplementary information may contain information specific to the software application performing the sleep analysis and/or assessment and/or the sleep analysis/assessment approach itself, such as a confidence level in the sleep quality measure (e.g. in terms of %). This aids the individual in gaining confidence in the system and algorithm.
In some examples, the supplementary information is responsive to historic and/or global sleep trends. This can, for instance, aid in contextualizing, decatastrophzing and/or otherwise justifying a poor night's sleep.
The supplementary information may thereby provide contextual information offering global normative information and/or information on the users' past history.
For instance, one example of supplementary information may be text indicating “You've had two nights of poor quality sleep this week. While we all want to sleep well every night, sleep specialists don't typically diagnose a problem unless you have difficulty falling or staying asleep typically three or more nights per week. Your history suggests that you typically have healthy sleep patterns and weeks like this aren't out of the ordinary. In fact, 95% of healthy normal sleepers have had weeks just like this or worse!”. It will be appreciated that the content of this exemplary text may be revised depending upon the specific circumstances for an individual—for instance, text may be omitted if irrelevant and/or values in the example text may be modified to reflect a true user's situation.
The supplementary information may also/otherwise provide text seeking to decatastrophize an individual for cognitive restructuring around the potential impact of a poor night of sleep includes offering user encouragement. Some example text may indicate “While last night's sleep wasn't the best, last time this happened you bounced back quickly.”. The skilled person will appreciate how the supplementary information may therefore depend upon historic data of the individual.
Supplementary information thereby provides the opportunity for the individual to reflect on the likely magnitude and type of impact of the poor night of sleep is on their day.
The approaches described above make use of predetermined ranges with respect to the predictive indicator. The nature of the predetermined ranges may depend upon the content of the predictive indicator, and can include binary ranges, categorical ranges and/or numeric ranges as appropriate. The predetermined ranges may be identical and/or may overlap one another.
The processor 460 may include a central processing unit (CPU), a digital signal processor (DSP), an ASIC, a controller, an FPGA, another hardware device, a firmware device, or any combination thereof configured to perform the operations described herein. The processor 460 may also be implemented as a combination of computing devices, e.g., a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration. In some embodiments, the processor is a distributed processing system, e.g. formed of a set of distributed processors.
The memory 464 may include a cache memory (e.g., a cache memory of the processor 460), random access memory (RAM), magnetoresistive RAM (MRAM), read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read only memory (EPROM), electrically erasable programmable read only memory (EEPROM), flash memory, solid state memory device, hard disk drives, other forms of volatile and non-volatile memory, or a combination of different types of memory. In an embodiment, the memory 464 includes a non-transitory computer-readable medium. The non-transitory computer-readable medium may store instructions. For example, the memory 464, or non-transitory computer-readable medium may have program code recorded thereon, the program code including instructions for causing the processing system 400, or one or more components of the processing system 400, particularly the processor 460, to perform the operations described herein. For example, the processing system 400 can execute operations of the method 700. Instructions 466 may also be referred to as code or program code. The terms “instructions” and “code” should be interpreted broadly to include any type of computer-readable statement(s). For example, the terms “instructions” and “code” may refer to one or more programs, routines, sub-routines, functions, procedures, etc. “Instructions” and “code” may include a single computer-readable statement or many computer-readable statements. The memory 464, with the code recorded thereon, may be referred to as a computer program product.
The communication module 468 can include any electronic circuitry and/or logic circuitry to facilitate direct or indirect communication of data between the processing system 400, the penetration device and/or the user interface (or other further device). In that regard, the communication module 468 can be an input/output (I/O) device. In some instances, the communication module 468 facilitates direct or indirect communication between various elements of the processing circuit 400 and/or the arrangement (
It will be understood that disclosed methods are preferably computer-implemented methods. As such, there is also proposed the concept of a computer program comprising computer program code for implementing any described method when said program is run on a processing system, such as a computer or a set of distributed processors.
Different portions, lines or blocks of code of a computer program according to an embodiment may be executed by a processing system or computer to perform any herein described method. In some alternative implementations, the functions noted in the block diagram(s) or flow chart(s) may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved.
The present disclosure proposes a computer program (product) comprising instructions which, when the program is executed by a computer or processing system, cause the computer or processing system to carry out (the steps of) any herein described method. The computer program (product) may be stored on a non-transitory computer readable medium.
Similarly, there is also proposed a computer-readable (storage) medium comprising instructions which, when executed by a computer or processing system, cause the computer or processing system to carry out (the steps of) any herein described method. There is also proposed computer-readable data carrier having stored thereon the computer program (product) previously described. There is also proposed a data carrier signal carrying the computer program (product) previously described.
Variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure and the appended claims. In the claims, the word “comprising” does not exclude other elements or steps, and the indefinite article “a” or “an” does not exclude a plurality. A single processor or other unit may fulfill the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. A computer program may be stored/distributed on a suitable medium, such as an optical storage medium or a solid-state medium supplied together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems. If the term “adapted to” is used in the claims or description, it is noted the term “adapted to” is intended to be equivalent to the term “configured to”. Any reference signs in the claims should not be construed as limiting the scope.
This application claims the benefit of U.S. Provisional Application No. 63/054,197, filed on 20 Jul. 2020. This application is hereby incorporated by reference herein.
| Number | Date | Country | |
|---|---|---|---|
| 63054197 | Jul 2020 | US |
