AN APPARATUS AND METHOD OF CONTROLLING A PAIN ALLEVIATING SLEEP ASSISTANCE DEVICE

Abstract

According to an aspect, there is provided an apparatus (10) to control a sleep assistance device (30) for alleviating a user's pain during sleep. The apparatus (10) comprises: a receiver (11) configured to receive sensor data (14) associated with the user; a processor (12) configured to analyze the sensor data (14) to predict a pain measure of the user, and to generate a control signal (15) for the sleep assistance device (30) in accordance with the predicted pain measure; and a transmitter (13) configured to transmit the control signal (15) to the sleep assistance device (30) for alleviating pain.

Claims

1. An apparatus to control a sleep assistance device for alleviating a user's pain during sleep, the apparatus comprising: a receiver configured to receive sensor data associated with the user; a processor configured to: analyze the sensor data to predict a pain measure of the user; analyze the sensor data to predict a sleep disturbance event in a predetermined time period; and generate a control signal for the sleep assistance device in accordance with the predicted pain measure and the predicted sleep disturbance event so as to minimize the possibility of occurrence of the predicted sleep disturbance event; and a transmitter configured to transmit the control signal to the sleep assistance device for alleviating pain.

2. The apparatus according to claim 1, wherein the processor is configured to: compare the sensor data with known data; and predict the pain measure of the user based on the comparison of the sensor data and the known data.

3. The apparatus according to claim 2, wherein the processor is configured to: analyze the sensor data to determine a stage of sleep of the user; predict the pain measure of the user based on the sensor data and the stage of sleep of the user.

4. The apparatus according to claim 1, wherein the processor is configured to: analyze the sensor data to determine a stage of sleep of the user; compare the sensor data with expected data for the stage of sleep; and predict the sleep disturbance event based on the comparison of the sensor data and the expected data.

5. The apparatus according to claim 1, wherein the processor is configured to: store the sensor data in a memory associated with the apparatus; train a classifier using the sensor data stored in the memory; and predict the sleep disturbance event in accordance with a likelihood score of the classifier.

6. The apparatus according to claim 1, wherein during a sleep onset period prior to the user falling asleep, the receiver is configured to receive initialization information; the initialization information comprises data indicative of one or more of: a position of the user, and physiological signals of the user; and the processor is configured to predict the pain measure of the user based on the sensor data and the initialization information.

7. The apparatus according to claim 1, wherein the receiver is configured to receive user data; the processor is configured to predict the pain measure of the user based on the sensor data and the user data; and the user data comprises one or more of: the age of the user; medical history of the user; medication prescribed to the user; medication taken by the user; a medical diagnosis of the user; lifestyle information of the user; sleeping preferences of the user; sleep environment of the user; a location of the user; the gender of the user; the ethnicity of the user; physical activity of the user; information on the user's preferred operating settings of the sleep assistance device; subjective input of the user; and an occupation of the user.

8. The apparatus according to claim 1, wherein the sleep assistance device is one of a plurality of sleep assistance devices; and the processor is configured to select the sleep assistance device from among the plurality of sleep assistance devices by analyzing the sensor data.

9. The apparatus according to claim 1, wherein the processor is configured to determine a setting instruction for the sleep assistance device in accordance with the predicted pain measure of the user; and the control signal comprises the setting instruction for the sleep assistance device.

10. The apparatus according to claim 1, wherein the receiver is configured to receive feedback data from the sleep assistance device, the feedback data comprising one or more of: present setting instruction of the sleep assistance device; a duration of operation of the sleep assistance device; status information of the sleep assistance device; and operating restrictions of the sleep assistance device; and the processor is configured to generate the control signal in accordance with the feedback data.

11. The apparatus according to claim 10, wherein the processor is configured to: store the feedback information in a memory associated with the apparatus in accordance with the sensor data received with the feedback information; analyze the stored feedback information and sensor data to determine an optimum mode of operation of the sleep assistance device for the user; and generate the control signal in accordance with the optimum mode of operation of the sleep assistance device.

12. The apparatus according to claim 1, wherein the sensor data is one or more of: electroencephalogram, EEG, data; movement data; image data; respiration data; heart data; electrodermal data; blood pressure data; audio data; and temperature data.

13. A method of controlling a sleep assistance device for alleviating a user's pain during sleep, the method comprising: receiving sensor data associated with the user; analyzing the sensor data to predict a pain measure of the user; analyzing the sensor data to predict a sleep disturbance in a predetermined time period; generating a control signal for the sleep assistance device in accordance with the predicted pain measure and the predicted sleep disturbance event so as to minimize the possibility of occurrence of the predicted sleep disturbance event; and transmitting the control signal to the sleep assistance device for alleviating pain.

14. A computer program which, when executed on a computing system, carries out a method of controlling a sleep assistance device for alleviating a user's pain during sleep, the method comprising: receiving sensor data associated with the user; analyzing the sensor data to predict a pain measure of the user; analyzing the sensor data to predict a sleep disturbance in a predetermined time period; generating a control signal for the sleep assistance device in accordance with the predicted pain measure and the predicted sleep disturbance event so as to minimize the possibility of occurrence of the predicted sleep disturbance event; and transmitting the control signal to the sleep assistance device for alleviating pain.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Exemplary embodiments will now be described, by way of example only, with reference to the following drawings, in which:

[0068] FIG. 1 is a block diagram of main apparatus components according to a general embodiment of an aspect of the invention;

[0069] FIG. 2 is a flowchart of a method according to a general embodiment of an aspect of the invention;

[0070] FIG. 3 is a block diagram of a system according to an embodiment of an aspect of the invention;

[0071] FIG. 4 is a flow diagram of a control method according to an embodiment of an aspect of the invention;

[0072] FIG. 5 is a graph showing the expected physiological readings of an individual in the different stages of sleep;

[0073] FIG. 6 is a graph of the stages of sleep against time in a standard sleep cycle; and

[0074] FIG. 7 is a hardware diagram illustrating hardware that may be used to implement invention embodiments.

DETAILED DESCRIPTION OF THE EMBODIMENTS

[0075] Embodiments of the present disclosure and the various features and advantageous details thereof are explained more fully with reference to the non-limiting examples that are described and/or illustrated in the drawings and detailed in the following description. It should be noted that the features illustrated in the drawings are not necessarily drawn to scale, and features of one embodiment may be employed with other embodiments as the skilled artisan would recognize, even if not explicitly stated herein. Descriptions of well-known components and processing techniques may be omitted so as to not unnecessarily obscure the embodiments of the present disclosure. The examples used herein are intended merely to facilitate an understanding of ways in which the embodiments of the present may be practiced and to further enable those of skill in the art to practice the same. Accordingly, the examples herein should not be construed as limiting the scope of the embodiments of the present disclosure, which is defined solely by the appended claims and applicable law.

[0076] It is understood that the embodiments of the present disclosure are not limited to the particular methodology, protocols, devices, apparatus, materials, applications, etc., described herein, as these may vary. It is also to be understood that the terminology used herein is used for the purpose of describing particular embodiments only, and is not intended to be limiting in scope of the embodiments as claimed. It must be noted that as used herein and in the appended claims, the singular forms a, an, and the include plural reference unless the context clearly dictates otherwise.

[0077] Unless defined otherwise, all technical and scientific terms used herein have the same meanings as commonly understood by one of ordinary skill in the art to which the embodiments of the present disclosure belong. Preferred methods, devices, and materials are described, although any methods and materials similar or equivalent to those described herein may be used in the practice or testing of the embodiments.

[0078] Embodiments of aspects may provide an apparatus, method, computer program and system of controlling a sleep assistance device for alleviating a user's pain during sleep.

[0079] As discussed above, many individuals, such as elderly people or people with medical conditions, suffer from chronic pain which often results in increased sleep fragmentation or disturbance, thereby directly influencing their sleep quality. It is therefore desirable to reduce an individual's pain so as to minimize sleep disturbances and improve sleep quality. In other words, it is desirable to avoid sleep disturbance events, such as those caused by pain, so that a person's sleep is not interrupted.

[0080] A person's perception of pain may be dependent on their sleep stage (such as, for example, their electroencephalogram, EEG, activity), their sleeping position and/or their movements. Pain in relation to sleep disturbance therefore varies depending on the stage of sleep. That is, a level of pain that may cause a disturbance to the person when they are in a light sleep state may not be sufficient to disturb the person when they are in a deep sleep state. Sleep assistance devices may be used to alleviate pain and reduce sleep disturbances. Optimum settings of sleep assistance devices may vary depending on the stage of sleep of the user. For example, a high or intense setting of a sleep assistance device may actually disturb the user when they are in a light sleep state. It is therefore desirable to control these devices in accordance with the stage of sleep of the user so that pain may be effectively reduced or managed and sleep disturbance events may be avoided or minimized.

[0081] FIG. 1 shows a block diagram of information flow into main apparatus components in apparatus 10. The apparatus 10 comprises a receiver 11, a processor 12 and a transmitter 13. Sensor data 14 associated with a user is received by the receiver 11. The sensor data 14 may therefore be data relating to the user and the user's condition. The processor 12 analyzes the received sensor data 14 to predict a pain measure of the user. In accordance with the predicted pain measure, the processor 12 generates a control signal 15 for a sleep assistance device, which is a device for alleviating or reducing pain or discomfort of the user. The control signal 15 is transmitted by the transmitter 13 to the sleep assistance device, so that the sleep assistance device may be controlled in accordance with the predicted pain measure of the user.

[0082] The apparatus may be a suitable processing device that is capable of being communicably connected to other devices. The device may, for example, be a smart phone associated with the user, or may be a smart home device, which is able to store and process data from multiple sleep monitoring devices. Additionally, the apparatus may be provided as part of a cloud computing network such that some of the processing is performed in the cloud. The apparatus may also be one of the monitoring/sensor devices that measures the sensor data. In this case, the device may still be connected to other sensor devices and/or the cloud. That is, the sensor device may receive data from the other devices, perform the analysis of the data, generate control signals for the sleep assistance device(s) and transmit the control signals to the devices. Alternatively, the apparatus may be a sleep assistance device which is connected to and receives data from the other devices.

[0083] FIG. 2 shows a flow chart representing the method according to a general embodiment of an aspect of the invention. Firstly, in step S21, sensor data is received. The sensor data is associated with a user, such that the received sensor data may relate to the user or the user's condition. The received sensor data is then analyzed to predict a pain measure of the user at step S22. Next, at step 23, the predicted pain measure is used to generate a control signal for a sleep assistance device. That is, a control signal for the sleep assistance device is generated in accordance with the predicted pain measure of the user, such that the sleep assistance device may be controlled based on the predicted pain of the user. Finally, in step S24, the generated control signal is transmitted to the sleep assistance device, which is a device for alleviating pain or discomfort of the user.

[0084] FIG. 3 shows a block diagram of a system according to an embodiment of an aspect of the invention. The system 100 comprises an apparatus 10, a sensor device 20, a sleep assistance device 30 and a memory 40. The sensor device 20 acquires sensor data 14 associated with a user during sleep and the sleep assistance device 30 is for alleviating pain or discomfort of the user. The apparatus 10 receives sensor data 14 from the sensor device 20, analyzes the sensor data 14 to predict a pain measure of the user, generates a control signal 15 for the sleep assistance device 30 in accordance with the predicted pain measure and transmits the control signal 15 to the sleep assistance device 30. The sleep assistance device may send feedback data 16 to the apparatus 10. The apparatus 10 may send data to and receive data from the memory 40.

[0085] Embodiments of the present invention may therefore enable a sleep assistance device to be controlled in accordance with sensor data and a predicted pain measure of the user. The sleep assistance device may therefore be controlled in a manner that is suited to the condition of the user and the effectiveness of the sleep assistance device may be maximized. By maximizing the effectiveness of the sleep assistance device, the pain relief provided to the user may also be maximized. The pain of the user may therefore be relieved or reduced and sleep disturbances or interruptions may be avoided. That is, the sleep assistance device may be controlled in accordance with the predicted pain measure of the user so that the stimulation of the sleep assistance device may be appropriately delivered to the user. The effectiveness of the sleep assistance device may therefore be increased and disturbance to the user's sleep may be minimized since pain may be relived prior to it disturbing or waking the user.

[0086] Embodiments of the present invention may therefore prevent disturbances to the user's sleep, by predicting such disturbances and taking necessary measures to eliminate or minimize the factors (either directly, or indirectly) that contribute to the sleep disturbance. It may be considered that two types of predictions are provided: [0087] (i) Sleep disturbance (for example, wake up) prediction [0088] (ii) Increased pain (perception) prediction

[0089] The sleep disturbance prediction may be direct and indirect. [0090] Direct: the sleep disturbance may be estimated directly from EEG and movement data. This may be achieved, for example, by training machine learning algorithms. [0091] Indirect: First pain increase (pain measure) may be estimated (for example, from EEG, movement, timing of medications, etc.), and based on the pain estimate and EEG and movement data, sleep disturbance may be estimated.

[0092] Pain detection and localization may be achieved by training a neural network classifier from brain signals (EEG data). According to an embodiment of an aspect, EEG data is used to train a neural network classifier. Additionally, motion data may be used if available. Other data may also be used such as, for example, medication intake information (time and type of medication). If pain medications are used, accurate predictions of the time periods when pain is expected may be determined using the pharmacokinetics and pharmacodynamics of the corresponding medications, well before physiological signals start to change. That is, if the time and type of medication is known, the time when the effect of the medication is likely to wear off may be predicted and taken into account in the pain measure and/or sleep disturbance event prediction.

[0093] Other data that may be used in the pain prediction and sleep disturbance prediction may include information on the user and their sleeping environment. For example, the data may include information on: the age of the user; the medical history of the user; medication prescribed to the user; a medical diagnosis of the user; lifestyle information of the user; sleeping preferences of the user; sleep environment of the user; a location of the user; the gender of the user; the ethnicity of the user; physical activity of the user; and/or an occupation of the user.

[0094] In addition to the EEG and motion data, other physiological signals may also be used for predicting wake-up and pain increase. Physiological signals such as respiration (rate and depth), heart signals (heart rate and heart rate variability), electrodermal activity (skin conductance, and skin temperature) and sound signals (sound of breathing, sound of motion, coughing, and speech-like signals in sleep) may be used. All of these may be monitored in real-time, or close to real-time.

[0095] The sensor data may therefore be acquired from any monitoring device or sensor which is capable of detecting/measuring appropriate metrics of the user that provide information on the user's sleep. The sensor data may be; electroencephalogram, EEG, data; movement data; image data; respiration data; heart data; electrodermal data; blood pressure data; audio data; and temperature data. If the sensor data is image data, an image (or sequence of images) of the user's face may be analyzed to detect facial expression changes that indicate that the user is in pain.

[0096] There is well-established literature concerning different sleep stages and the corresponding EEG signals features during these sleep stages. According to an embodiment of an aspect, one way to predict the disruption of sleep-patterns is by detecting more than expected variations in the EEG signals. For instance, in stage 1 sleep EEG signals are expected to be mainly in F1 range (frequency) and with A1 (amplitude). If the current monitoring shows larger variations than these expected ranges (F1 and A1), this can be considered as a potential trigger for sleep disturbance. Consequently, the system can be activated to prevent further deviations, and bring the signals to the levels expected of them for a given sleep stage. In other words, this can be considered as a feedback system, where the effect and adaptation of sleep assistance device(s) (for example, position change and/or pain relief) is guided by the changes in the EEG signals.

[0097] The stages of sleep may be categorized as follows: [0098] Wakefulness: High frequency, low-voltage, which is usually beta-band corresponding to 13-24 Hz. [0099] Falling asleep and Stage 1: Alpha band (8-12 Hz), emerging theta-wave (4-7 Hz). Breathing and heart rate decline. [0100] Stage 2: Sleep spindles. [0101] Stage 3 and Stage 4: Delta waves (4 Hz or less). [0102] REM (Stage 5): EEG somewhat similar to wakefulness, irregular breathing and heart rate, rapid eye movements. Although EEG patterns are somewhat similar to wakefulness, rapid eye-movements and time of occurrence (i.e. preceding sleep stages) can be used to differentiate this from wakefulness. It is difficult to wake a person during this stage.

[0103] Another implementation method for prediction may be to have personalized definitions of wakefulness, sleep-disruption and pain-perception. In other words, user data may be used define these stages. Instead of using general definitions of these, having a personalized definition may improve the prediction accuracy. For example, if person A is known to mostly wake-up during stage 1, this sleep stage can also be defined as wake-up period, and since the sleep stages have a well-defined pattern, pain reduction methods can be activated during the REM-stage. Alternatively or additionally, if it is known that the likelihood for the user waking up is much higher when they are in the supine position, the pain relief device may be activated when the person changes to prone position.

[0104] As another example, pain definition for person A may be lack of movement of left leg. If this is triggered, pain relief techniques may be activated. Moreover, other physiological signals may also be used for description of pain during sleep. In particular, changes in the respiration, heart rate, skin temperature, and/or blood pressure may be indicative of painful periods.

[0105] The acquired sensor data relating to the user or the user's condition may therefore be used to control a sleep assistance device, which is a device for alleviating or reducing a user's pain or discomfort. The device may be controlled in accordance with a predicted pain measure or a future sleep disturbance event, so as to reduce the likelihood of sleep disturbance occurring.

[0106] Upon predicting a sleep disturbance, the desire is to prevent the disturbance from occurring. However, it may not be immediately clear what the best way to avoid the disturbance is (for example, which sleep assistance device from multiple devices to select, or the settings of the devices). Embodiments of the present invention may be considered as a control apparatus in which one or more sleep assistance devices are continuously tuned based on the feedback from the user (as determined from the sensor data), with minimal intervention during sleep desired.

[0107] The sleep assistance device may, for example, be a pain relief device or a position change device. Additionally or alternatively, the sleep assistance device may be a sound generation device, a temperature changing device, an air quality and air circulation changing device, and/or a medication providing device. A pain relief device may apply a stimulation to the individual to reduce pain. The stimulation may, for example, be an electrical stimulation, light stimulation and/or heat stimulation. The stimulation may increase blood flow to an area causing pain so as to reduce the pain. An example of such a pain relief device is the Philips BlueTouch pain relief patch. The BlueTouch device uses blue LED light to reduce pain by providing soothing warmth to an area of pain of a user, such as, for example, back pain. The blue LED light stimulates the production of nitric oxide (NO) in the body, which promotes the circulation. As a consequence, the supply of oxygen and nutrients to the muscle is improved and, at the same time, pain transmission is reduced. The muscles may relax and pain may be relieved. NO is said to have antioxidant, cytoprotective and anti-inflammatory characteristics. It is thus able to protect muscles and nerves against damage and prevent further injuries. The Philips BlueTouch device may therefore reduce pain in a user, which, when applied during sleep, may prevent or reduce sleep disturbances. Although the Philips BlueTouch device is discussed here, any other device directed to pain relief may be used as a sleep assistance device. Such devices include, but are not limited to, nerve stimulation devices (for example, transcutaneous electrical nerve stimulation), spinal cord stimulation, implantable devices, medication release devices, and/or massaging devices.

[0108] Due to their nature, pain relief devices, such as the Philips BlueTouch device, may have limited application times. For example, the Philips BlueTouch device has a recommended application time of 15 to 30 minutes, depending on the mode of the device. The operation time of the device may therefore be taken into account when generating the control signal. For example, if the sleep assistance device is a pain relief device that has a maximum application time of 30 minutes, the control signal may be generated to ensure that the pain relief device is not used for longer than 30 minutes. If the pain relief device has reached its maximum application time, the processor may select another sleep assistance device from among a plurality of sleep assistance devices for pain relief.

[0109] The pain relief device may also be a medication dispenser. For example, Parkinson's disease patients have such a device that gradually releases medication. Such a device may therefore be controlled in accordance with the predicted pain measure of the user and/or a predicted sleep disturbance event. For example, the device may release medication depending on a predicted sleep disturbance event if needed.

[0110] A position change device is a device that causes the user to change position when sleeping, preferably with minimal disturbance to the user (for example, using haptic and/or audio stimulation). The position change device may cause the user to move to a position that may reduce pain or be less likely to result in pain or discomfort. For example, the position may be suited to improved circulation, or may be a position that does not aggravate an injury of the user.

[0111] An example of a position change device is the Philips SmartSleep Snoring Relief Band. The Snoring Relief Band is a device that is worn around the chest of the user and uses vibrations to cause the user to move position while sleeping. For example, the device delivers gentle vibrations to cause the user to move from a position in which they are lying on their back to a position in which they are lying on their side. The Snoring Relief Band is used to cause the user to move to a position in which they are less likely to snore. However, the technology of the device may also be used to prompt the user to move to a position that is likely to relieve pain or is less likely to cause pain or discomfort. Although the Philips SmartSleep Snoring Relief Band is discussed here, any other device directed to causing the user to move may be used as a sleep assistance device. Such devices include, but are not limited to, a smart watch that can trigger user movement by haptic or auditory stimuli, and/or a movement device integrated in bed, etc.

[0112] The sensor data may be acquired from any device or sensor which is capable of detecting appropriate metrics of the user that provide information on the user's sleep. For example, the sensor data may be electroencephalogram (EEG) data acquired from an EEG sensor or the sensor data may be movement data acquired from an accelerometer. EEG data provides information on the electrical activity of the user's brain and it is possible to determine information about the user's sleep from this information. For example, it is possible to determine the stage of sleep of the user. Furthermore, it is also possible to determine whether the user feels pain or discomfort and whether a sleep disturbance is likely.

[0113] Movement data may be acquired from a device such as an accelerometer to monitor movement of the user. The user is more likely to move in certain stages of sleep compared with other stages of sleep and so the stage of sleep may be estimated from the movement data. Furthermore, an increase in movement may be indicative of pain or discomfort of the user. Increased restlessness may also indicate an approaching sleep disturbance event. A sleep disturbance may therefore be predicted from the movement data.

[0114] The sensor data may comprise multiple types of data received from one or more sensor devices. For example, the sensor data may comprise both EEG data and movement data. The use of multiple data types may increase the accuracy of the sleep disturbance prediction. The data may be monitored in real-time so that a pain measure of the user may be assessed in real-time and appropriate control of one or more sleep assistance devices may be applied. Similarly, the real-time monitoring of the sensor data may enable prediction of a sleep disturbance event in a subsequent time period. Appropriate control of one or more sleep assistance devices may be applied in accordance with a predicted sleep disturbance event so as to reduce the likelihood of the sleep disturbance event occurring, thereby improving the user's sleep quality. That is, a future sleep disturbance event may be predicted using, for example, the interaction of the EEG, movement, and position data to minimize sleep disturbances. The control and usage of one or more sleep assistance devices occurs based on the predicted sleep disturbance (i.e. before disturbance happens) so as to provide intervention and decrease the likelihood of the disturbance happening.

[0115] An example of a sensor device for acquiring the sensor data and providing it to the apparatus is the Philips SmartSleep Deep Sleep Head Band. The Deep Sleep Head Band is worn by the user during sleep and collects EEG data using sensitive sensors. The Deep Sleep Head Band detects deep sleep in real-time based on the EEG data. The Deep Sleep Head Band can also provide acoustic stimulation to the user to enhance slow waves (slow wave activity) during sleep, thereby improving the quality of sleep. Thus, a sensor device for acquiring sensor data and a sleep assistance device may be the same device.

[0116] Monitoring and analysis of the sensor data, such as, for example, the user movement data and EEG data, in real-time therefore allows the user's pain level to be monitored and potential sleep disturbances to be predicted, and appropriate action may be taken to reduce the chance of the sleep disturbance happening. Additional data, such as further information on the user or their lifestyle (for example, age, gender, medical conditions, circadian rhythm, etc.) and/or information on the user's sleep environment (for example, temperature, noise, whether they share a bed, etc.) may also be used in the prediction of the pain level and sleep disturbance events. Further information on the user or their sleep environment may improve the accuracy of the analysis of the sensor data since such additional factors may have an effect on the user's sleep and sleep cycle. The additional information may also be used in the generation of the control signal to the sleep assistance device so that the control of the sleep assistance device may be further improved.

[0117] One or more sleep assistance devices may be activated or, if they are already active, their settings may be adjusted in accordance with the predicted pain measure and/or the predicted sleep disturbance event. The settings of the device(s) may be tailored according to the analyzed sensor data and the pain level and/or predicted sleep disturbance event. The sleep assistance devices may therefore be controlled to reduce the pain level of the user and/or to avoid the sleep disturbance event. Multiple sleep assistance devices may be used in cooperation to provide the most suitable pain relief to the user. The control of multiple devices together may be provided by the apparatus based on the analysis of the sleep data, as well as the analysis of any additional data.

[0118] If multiple sleep assistance devices are available, the apparatus may select which of the devices to control based on the received sensor data. Additional data, such as the user or environment information, may also be used in the determination. One or more appropriate devices from among a number of available devices may therefore be selected that are best suited to the user at that moment so as to effectively relieve pain and/or avoid a predicted sleep disturbance event. The settings of a selected device may be controlled with respect to the settings of other selected devices. The sleep assistance devices may be a pain relief device and a position change device which may be considered together to determine which device should and can be activated to provide the most appropriate relief to the user and the device settings to achieve the relief.

[0119] The sleep assistance devices may have operation restrictions or rules, which may be dependent on the use of other devices. For example, it may be preferable that two particular devices are not used at the same time, or it may be preferable that one device is used before another device. Considering the example of the pain relief device and the position change device being used in collaboration, it may be preferable that both devices are not activated simultaneously so as to increase the coverage during the night (because, for example, the pain relief device can be used only for limited time) and to minimize unintended disturbances (for example, the position change device can disturb the user, or an individual that the user shares a bed with). In some circumstances (for example, close to morning) both devices may be used simultaneously. The interaction between the pain relief device and the position change device may be explored and optimized to improve sleep quality and reduce sleep disturbances.

[0120] FIG. 4 shows a flow diagram of a control method according to an embodiment of an aspect of the invention. Movement data and EEG data is collected at step S41 and step S42, respectively. This data is used in real-time sleep data analysis at step S43. Next, at step S44 a pain measure is predicted and a sleep disturbance (such as, for example, a wake up event) is predicted using the analyzed sleep data and, from the predictions, it is determined whether to activate a sleep assistance device. Thus, at step S45, it is determined whether to activate the position change device, with the position change device settings adjusted at step S46 if the position change device is activated. If it is determined not to use the position change device, then a determination of whether to use the pain relief device is performed at step S47. The pain relief device setting are adjusted at step S48 if the pain relief device is activated. The adjustments of the position change device settings and the pain relief device settings are therefore made in accordance with the sleep data analysis and the resulting predicted pain measure and sleep disturbance prediction. As can be in FIG. 4, the movement data may also be utilized in the adjustment of the positon change device settings and the EEG data may similarly be utilized in the adjustment of the pain relief device settings. The settings of the position change device and/or the pain relief device may also be fed back to be used in the sleep data analysis (not shown).

[0121] Thus, future sleep disturbance events may be predicted ahead of time so that action may be taken to attempt to avoid the sleep disturbance events. The sleep disturbance events may be associated with the pain measure of the user.

[0122] In the example shown in FIG. 4, it is preferable for the position change device to be used ahead of the pain relief device. However, the pain relief device may be used ahead of the position change device and/or the devices may be used at the same time. The use of the devices may depend on a number of factors including the sensor data, the user data and/or the environment data.

[0123] In certain circumstances, the pain relief device may have a set usability time, which cannot be exceeded. Many different pain relief options may be available to use. The position changing device may not generally have such a usability limit. However, in a case in which the user shares a bed with another individual, a limit on the position change device may also be imposed so as to reduce the risk of disturbing the other individual. The apparatus may determine which device to use based on the real-time analysis of, for example, EEG and movement data to prevent or minimize a predicted sleep disturbance.

[0124] In an embodiment of an aspect, the first option is always to use position changing device. The sleep data continues to be monitored after the position change device has been activated/controlled and if it is determined (for example, from movement and EEG data) that the user is still restless after the position change, the data is re-evaluated and a new device usage decision may be made. The second option includes using position change or pain relief device, but it is again preferred to use the position change device ahead of the pain relief device. This can be implemented by having a stricter threshold for pain relief device usage (or a lower weight in case of classifier-like implementation). If there is a need for a new re-evaluation, both devices are equally likely to be selected. If the pain relief device has been used and has reached its usage limit, the position changing device is used.

[0125] The decision on which pain relief device settings to use may be driven by the nature of the discomfort or pain, as well as the allowed usage time limit, and the position and estimated wake-up time of the user. Other sleep characteristics (for example, time in bed, duration of deep sleep, etc.) or lifestyle characteristics (for example, stress, activity, nutrition) characteristics may also be used. Multiple devices may be used together. For example, if a pain relief device provides localized stimulation to a specific area of the user, the position change device may be used prior to the pain relief device to move the user to an optimal position for the pain relief device. The pain relief device may be provided as part of a bed, rather than as a wearable sensor, and so it may be necessary to cause the user to move to the pain relief device using the position change device prior to using the pain relief device.

[0126] Information from the sleep assistance devices may be fed back to the apparatus to be used in the determination of the sleep disturbance event and/or the generation of the control signal for one or more sleep assistance devices, which may be the same or different devices from those providing feedback. Alternatively or additionally, the generated control signal may be an updated control signal for the sleep assistance devices providing the feedback. The kind of information that is provided by the sleep assistance devices may depend on the intelligence and features of the devices. In the simplest case, information on the usage time, the usage duration and the intensity settings may be shared. For sleep assistance devices that are able to provide localized and targeted solutions, then the location of pain relief, and the user position change information may also be shared.

[0127] The apparatus may learn the best position for sleep and the best position to provide pain relief support, as well as the sleep assistance device settings for the user over time. The best position for sleep and the best position for providing pain relief may be different. Therefore, according to an embodiment of an aspect, the position change device may bring the user to the best position for application of the pain relief device, instead of aiming to minimize the user pain by position change. That is, multiple devices can be controlled in cooperation.

[0128] An initialization of the apparatus may be performed daily prior to the user falling to sleep. In the initialization, the user's position is measured before falling asleep since it is assumed that the user will consciously try to assume a position in which they are comfortable and do not feel pain or discomfort, or feel the minimum amount of pain or discomfort. It may be preferable to perform the initialization daily because the nature of pain may evolve over time and so the preferred position of the user may change accordingly. Of course, the initialization can be performed at more or less frequent intervals, for example, once every three days, weekly or fortnightly. The decision of the frequency of the initializations may be learned by the apparatus from the user data over time. For example, if the position at which the user feels minimal pain or discomfort varies most nights, then it may be beneficial to perform the initialization frequently, for example, nightly so that the apparatus can acquire the optimum sleep position for the night. Conversely, if the user has consistently felt a minimal amount of pain in a given position for a number of nights, the apparatus may use this position as the preferred position of the user and initialization may not be necessary. Performing regular (for example, daily) initializations will enable the apparatus to track the pain or comfort level change of the user over time. This may be calculated as a function of sleep duration, and device usage time (for example, sleep duration/(pain relief device usage time+position change device usage time)). This is just one particular way for calculating the score. In general, different methods may be devices from using sleep signals (e.g. EEG or actigraphy) and sleep assistance devices parameters. For this particular equation, if it is observed that the user is able to have longer durations of sleep without using any external help, this is positive and desired indication.

[0129] A number of methods may be used for predicting a sleep disturbance event.

[0130] Method 1: Sleep Stage Specific Sleep Disturbance Prediction

[0131] The main idea of the first method is that every sleep stage is defined by specific brain wave (EEG) features which are well established. Moreover, the brain wave features may be personalized to the user by monitoring the user and their EEG data for a number of nights.

[0132] The first method comprises the following steps: [0133] Step 1: Detect current sleep stage: use historical data, or look up table to quantitatively describe the sleep stage, for example, in terms of EEG signal frequency range and amplitude, user movement, heart rate, heart rate variability, respiration rate, etc. features. [0134] Step 2: Continuously check if the calculated features (for example, frequency, amplitude) are within the expected range. If not (for a specified amount of time), predict a sleep disturbance event.

[0135] The first method therefore compares the sleep data with known data and predicts a sleep disturbance event if the sleep data deviates from the expected data since a deviation from the expected pattern may be indicative of approaching disturbance. That is, since the sleep stages come in order and every sleep stage has relatively well defined features, these features may be monitored to determine disturbance. In this case, the disturbance will be predicted based on the trend of the features. In other words, if, for example, the trend indicates a change which is not in line with what is expected, then this can be considered as a disturbance indicator.

[0136] For example, during deep sleep, if the frequency of the EEG waves is increasing, or if the amplitude of the delta waves is decreasing, these can be taken as disturbance indicators. Note, however, that a decrease in frequency, and an increase in delta amplitude may not be considered as a disturbance because it is in line with the deep sleep changes.

[0137] Method 2: Using Machine Learning to Predict Sleep Disturbance in a Linear Manner

[0138] While the first method uses EEG as the main data, the second method uses additional parameters in the prediction, such as, for example, movement, respirations, heart rate, skin temperature, blood pressure, and some other physiological parameters. The machine learning method may be the preferred method for pain prediction.

[0139] The second method comprises the following steps: [0140] Step 1: Collect user data for a period of time, such as, for example, one week, and use a short time window (for example, 30 seconds) to train a binary classifier (awake or not-awake), for linear prediction (for example, predict the following window features, from the previous windows). [0141] Step 2: Observe the output of the classifier to predict a disturbance (for example, a disturbance is predicted if the likelihood of awake exceeds 90%). [0142] Classifiers such as convolutional neural network (CNN), support vector machine (SVM), random forest and K-means clustering may be used.

[0143] Thus, the approach of the second method is to train a binary predictor that is able to assign a likelihood score for the next time period, such as, for example, 30 seconds or 60 seconds based on the past samples. The aim here is to determine the likelihood of the user waking, and to take action (i.e. control sleep assistance devices) in accordance with the likelihood.

[0144] If, during application of a pain relief device, the likelihood for the user waking up is higher than a threshold, use of the device should be stopped so that the user is not disturbed. If, during sleep, the likelihood of the user waking up is higher than a threshold and pain is determined/predicted (from the changes in the physiological signals), then one or more sleep assistance devices, such as pain relief devices, may be activated and the settings of the pain relief devices set accordingly.

[0145] Method 3: Use Machine Learning to Predict Sleep Activity in a Global Manner, for the Whole Duration of the Night

[0146] The third method is similar to the second method yet longer term predictions (for example, predictions for the next hour, rather than the next minute) of the sleep stages and wake up periods are performed.

[0147] The third method comprises the following steps: [0148] Step 1: Collect user data for a longer time period (for example, one month). Train an algorithm (for example, CNN) to predict the wake up times throughout the night, as a function of going to bed time, and first sleep cycle (onset-I-II-III-IV-REM). [0149] Step 2: Only use the devices outside of the predicted wake-up periods.

[0150] The third method enables the application of pain relief to be activated during deep sleep stages (stages 3 and 4) prior to the predicted wake-up period. The idea here is to find the optimal time to apply the pain relief by better aligning the application to be during certain sleep stages.

[0151] Method 4: A Combination of the First to Third Methods

[0152] An alternative approach may be to use the first, second and third methods in combination. For example, the methods may be applied sequentially, such that method 3 is applied first, followed by method 2 and then method 1.

[0153] FIG. 5 shows a graph of the expected physiological readings of a user in the different stages of sleep, which may be used in the sleep disturbance prediction methods, as discussed above. The graph shows the expected physiological data against sleep time and the stages of sleep, and includes heart rate and respiration. FIG. 6 shows a graph of sleep stages against time in a standard sleep cycle. Table 1 below lists the expected EEG data (frequency and amplitude) for each stage of sleep.

TABLE-US-00001 TABLE 1 Expected EEG data for sleep stages Stage of Sleep Expected Data Awake/REM: EEG freq. 15-50 Hz Beta activity EEG amplitude: 30 V Stage I: Theta EEG freq. 4-8 Hz waves EEG amplitude: 50-100 V Stage II: Spindles EEG freq. 10-15 Hz EEG amplitude: 50-150 V Stage III EEG freq. 2-4 Hz EEG amplitude: 100-150 V Stage IV: Delta EEG freq. 0.5-2 Hz waves EEG amplitude: 100-200 V

[0154] Embodiments of aspects may therefore provide an apparatus and method for alleviating a user's pain during sleep in which a control signal may be generated and transmitted to a sleep assistance device in accordance with a predicted pain measure of the user. Sleep disturbances and/or interruptions may therefore be reduced or avoided and the quality of sleep of the individual may be improved. Furthermore, the operation of the sleep assistance device may be controlled, such that the effectiveness of the device may be maximized and the pain or discomfort of the individual may be reduced.

[0155] FIG. 7 is a block diagram of a computing device, such as a server incorporating resources suitable for sensor data processing, which may embody the present invention, and which may be used to implement some or all of the steps of a method embodying the present invention, and perform some or all of the tasks of an apparatus of an embodiment. For example, the computing device of FIG. 7 may be used to implement all, or only some, of steps S21 to S24 of the method illustrated in FIG. 2, to perform all, or only some, of the tasks of the apparatus shown in FIG. 1 to perform all, or only some, of the tasks of receiver 11, processor 12 and/or transmitter 13. The computing device comprises a processor 993, and memory 994, and to implement all, or only some, of steps S41 to S48 of the process illustrated in FIG. 4. Optionally, the computing device also includes a network interface 997 for communication with other computing devices, for example with other computing devices of invention embodiments.

[0156] For example, an embodiment may be composed of a network of such computing devices. Optionally, the computing device may also include one or more input mechanisms 996 such as a keyboard and mouse for the user to input any of, for example, user information or preference settings of the sleep assistance device, and a display unit 995 such as one or more monitors. The display unit may show a representation of data stored by the computing device for instance, representations of the determined pain level, predicted sleep disturbance events and sleep assistance device settings. The display unit 995 may also display a cursor and dialogue boxes and screens enabling interaction between a user and the programs and data stored on the computing device. The input mechanisms 996 may enable a user to input data and instructions to the computing device. The components are connectable to one another via a bus 992.

[0157] The memory 994 may include a computer readable medium, which term may refer to a single medium or multiple media (e.g., a centralized or distributed database and/or associated caches and servers) configured to carry computer-executable instructions or have data structures stored thereon. Computer-executable instructions may include, for example, instructions and data accessible by and causing a general purpose computer, special purpose computer, or special purpose processing device (e.g., one or more processors) to perform one or more functions or operations. Thus, the term computer-readable storage medium may also include any medium that is capable of storing, encoding or carrying a set of instructions for execution by the machine and that cause the machine to perform any one or more of the methods of the present disclosure. The term computer-readable storage medium may accordingly be taken to include, but not be limited to, solid-state memories, optical media and magnetic media. By way of example, and not limitation, such computer-readable media may include non-transitory computer-readable storage media, including Random Access Memory (RAM), Read-Only Memory (ROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Compact Disc Read-Only Memory (CD-ROM) or other optical disk storage, magnetic disk storage or other magnetic storage devices, flash memory devices (e.g., solid state memory devices).

[0158] The processor 993 is configured to control the computing device and execute processing operations, for example executing code stored in the memory to implement the various different functions described here and in the claims. The memory 994 stores data being read and written by the processor 993, such as the inputs (such as, for example, the received sensor data), interim results (such as, for example, analysis of the sensor data and determined pain level) and results of the processes referred to above (such as, for example, the control signal to be transmitted). As referred to herein, a processor may include one or more general-purpose processing devices such as a microprocessor, central processing unit, or the like. The processor may include a complex instruction set computing (CISC) microprocessor, reduced instruction set computing (RISC) microprocessor, very long instruction word (VLIW) microprocessor, or a processor implementing other instruction sets or processors implementing a combination of instruction sets. The processor may also include one or more special-purpose processing devices such as an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a digital signal processor (DSP), network processor, or the like. In one or more embodiments, a processor is configured to execute instructions for performing the operations and steps discussed herein.

[0159] The display unit 995 may display a representation of data stored by the computing device and may also display a cursor and dialog boxes and screens enabling interaction between a user and the programs and data stored on the computing device. The input mechanisms 996 may enable a user to input data and instructions to the computing device. The display unit 995 and input mechanisms 996 may form the output 26.

[0160] The network interface (network I/F) 997 may be connected to a network, such as the Internet, and is connectable to other such computing devices via the network. The network I/F 997 may control data input/output from/to other apparatus via the network. Other peripheral devices such as microphone, speakers, printer, power supply unit, fan, case, scanner, trackerball etc. may be included in the computing device.

[0161] Methods embodying the present invention may be carried out on a computing device such as that illustrated in FIG. 7. Such a computing device need not have every component illustrated in FIG. 7 and may be composed of a subset of those components. A method embodying the present invention may be carried out by a single computing device in communication with one or more data storage servers via a network. The computing device may be a data storage itself storing the input content before and after processing and thus for example, the dialogue and/or trained model.

[0162] A method embodying the present invention may be carried out by a plurality of computing devices operating in cooperation with one another. One or more of the plurality of computing devices may be a data storage server storing at least a portion of the data.

[0163] Other hardware arrangements, such as laptops, iPads and tablet PCs in general could alternatively be provided. The software for carrying out the method of invention embodiments as well as input content, and any other file required may be downloaded, for example over a network such as the internet, or using removable media. Any dialogue or trained model may be stored, written onto removable media or downloaded over a network.

[0164] The invention embodiments may be applied to any field in which a user is monitored during sleep and measures are taken to reduce the user's pain and avoid sleep disturbance. The invention embodiments may preferably applied to the medical and healthcare field.

[0165] Artificial neural networks are widely employed to perform pattern matching and diagnostic procedures, using so-called machine learning. A typical structure of an artificial neural network is a three-layer structure, having an input layer at which observations are input to the network, a hidden or processing layer, at which further processing operations are carried out on information received from the input layer, and an output layer, at which an output signal is generated based on information received from the processing layer. The precise structure of the artificial neural network is not limited, neither are the specific functions of the layers.

[0166] A suitable neural network system may include a training processor which utilizes test data and annotated data to generate a trained model for an AI system, which trained model is accessible by an AI system. Detection is performed with reference to a similarity value computation processor.

[0167] Such a system comprises a hardware architecture such as that illustrated in FIG. 7 (described above), which may be used for training a model using the sensor data. In particular, processor 993 may perform the processing instructions for training the model, determining the pain level of the user and/or predicting a sleep disturbance event and providing the response and the associated processing. One or more storage units (memories) 994 store the processing instructions, the trained model query response and the dialogue.

[0168] Variations to the disclosed embodiments may be understood and effected by those skilled in the art in practicing the principles and techniques described herein, 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 fulfil 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 or 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. Any reference signs in the claims should not be construed as limiting the scope.

[0169] The above-described embodiments of the present invention may advantageously be used independently of any other of the embodiments or in any feasible combination with one or more others of the embodiments.