DESC:SYSTEM AND METHOD FOR MONITORING CHARACTERISTICS OF SLEEP BY GENERATING SLEEP CYCLE INTERVAL USING ML
Technical Field
[0001] The embodiments herein generally relate to monitoring characteristics of sleep of a subject, more particularly, a system and method of generating a sleep time cycle interval based on sensor data using a machine learning model that enables to evaluate one or more health conditions of the subject.
Description of the Related Art
[0002] With the economic, health, development of medical standards, the average life expectancy is extended comparatively. At the same time, due to the increased competitive pressure and many other factors, on a global scale, the incidence of mental illness is growing year by year and if becomes one of the main causes of death. Some psychiatric conditions may cause sleep problems, and sleep disturbances may also exacerbate the symptoms of many mental conditions including depression, anxiety, and bipolar disorder. The relationship between sleep and mental health is complex. The lack of sleep triggers the onset of certain psychological conditions. The circular relationship between sleep patterns and mental state helps medical practitioners to identify the severity of the illness.
[0003] Many existing systems are developed to identify the severity of the illness. Once such an existing system may include inbuilt sensors or modules in a variety of products like wearable devices such as watches, armbands, head-worn devices, and contactless products to collect data related to the sleep of a person from the sensors data. Another existing system may include sensors inbuilt in mattresses to monitor the sleep pattern of the person by collecting the sensors' data.
[0004] The existing system analyses the sleep pattern of the user by collecting sound signals near the bed of the user with the help of a recorder. The snoring sound signals during the sleep of the user are recorded and analyzed the sleep quality of the user. This existing system may fail if the user does not snore or if it records sound signals from multi-sources apart from the user.
[0005] Accordingly, there is a need for a system and method of analyzing sleep patterns during sleep cycle intervals in a more accurate and efficient way.
SUMMARY
[0006] In view of foregoing an embodiment herein provides a system for monitoring characteristics of sleep of a subject by generating a sleep time cycle interval based on a sensor data using a machine learning model that enables to evaluate one or more health conditions of the subject. The system includes a digital device that includes at least one of a screen button, a gyroscope, or an accelerometer, the digital device obtains the sensor data during a pre-defined time interval, the sensor data includes at least one of a timestamp of screen on, a timestamp of screen off, a timestamp of subject interaction, a three-dimensional accelerometer data, or a three-dimensional gyroscope data. The system includes a sleep pattern analyzing server that acquires the sensor data from the digital device, and processes, using the machine learning model, the sensor data. The sleep pattern analyzing server includes a memory that stores a database and a set of instructions; and a processor that is configured to execute the machine learning model and is configured to (i) extract, using a data extraction method, data points after cleaning at least one of noise or out-of-sync data from the sensor data; (ii) group the data points based on the pre-defined time interval to obtain one or more data points groups; (iii) generate, using a domain feature technique, one or more domain features from the one or more data points groups, the one or more domain features includes time-domain features, and frequency domain features; (iv) estimate, using a Bayesian model, time points of user’s sleeping time and user’s wake-up time by stacking the one or more domain features that are correlated, the one or more domain features that are correlated are selected using filtering, wrapping, and an embedding technique; (v) assign, using a Bayesian model, a weight to each time point of the user’s sleeping time and the user’s wake-up time that is estimated; and (vi) generate, using a trained machine learning model, the sleep time cycle interval using a switch point to monitor the characteristics of the sleep of the subject that enables to evaluate one or more health conditions of the subject, the switch point is calculated using the weight of each time point of the user’s sleeping time and the user’s wake-up time, the characteristics of sleep includes a sleep interruption, type of interruption, and a quality of sleep.
[0007] In some embodiments, the processor is configured to train the machine learning model by providing one or more historical weights and one or more historical sleep time cycle intervals associated with one or more subjects as training data to obtain the trained machine learning model.
[0008] In some embodiments, the processor is configured to reduce dimensions of the one or more domain features using at least one of principal component analysis or factor analysis to reduce a count of the one or more domain features, the count is used to provide the one or more domain features for training the machine learning model.
[0009] In some embodiments, the time-domain features and the frequency domain features comprise a frequency, a skewness, a mean, a kurtosis, and higher-order intensities.
[0010] In some embodiments, the processor is configured to generate a motion intensity score using the data points of the three-dimensional accelerometer data by, (i) determining acceleration vector components for each axis of the data points of the three-dimensional accelerometer data; and (ii) generating the motion intensity score by determining magnitudes of the acceleration vector components over a period of time.
[0011] In some embodiments, the processor is configured to measure a pacing intensity score based on a walking pattern during the sleep cycle interval by, (i) determining a pacing moment by combining the motion intensity score with a time-duration, and a walking axis; and (ii) determining the pacing intensity score using a duration of the pacing moment and a speed of motion.
[0012] In some embodiments, the motion intensity score is used to determine the one or more domain features.
[0013] In some embodiments, the pacing intensity score is used to determine the quality of sleep.
[0014] In some embodiments, the processor is configured to detect a pattern across the data points for grouping the data points using k-means clustering and t-distributed stochastic neighbor embedding.
[0015] In one aspect, a processor-implemented method for monitoring characteristics of sleep of a subject by generating a sleep time cycle interval based on a sensor data using a machine learning model that enables to evaluate one or more health conditions of the subject is provided. The method includes obtaining the sensor data from a digital device that includes at least one of a screen button, a gyroscope, or an accelerometer during a pre-defined time interval, the sensor data includes at least one of a timestamp of screen on, a timestamp of screen off, a timestamp of user interaction, a three-dimensional accelerometer data, or a three-dimensional gyroscope data. The method includes extracting, using a data extraction method, data points after cleaning at least one of noise or out-of-sync data from the sensor data. The method includes grouping the data points based on the pre-defined time interval to obtain one or more data points groups. The method includes generating, using a domain feature technique, one or more domain features from the one or more data points groups, the one or more domain features includes time-domain features, and frequency domain features. The method includes estimating, using a Bayesian model, time points of user’s sleeping time and user’s wake-up time by stacking one or more domain features that are correlated, the one or more domain features that are correlated are selected using filtering, wrapping, and an embedding technique. The method includes assigning, using a Bayesian model, a weight to each time point of the user’s sleeping time and the user’s wake-up time that is estimated. The method includes generating, using a trained machine learning model, the sleep time cycle interval using a switch point to monitor the characteristics of the sleep of the subject that enables to evaluate one or more health conditions of the subject, the switch point is calculated using the weight of each time point of the user’s sleeping time and the user’s wake-up time, the characteristics of sleep includes a sleep interruption, type of interruption, and a quality of sleep.
BRIEF DESCRIPTION OF THE DRAWINGS
[0016] The embodiments herein will be better understood from the following detailed description with reference to the drawings, in which:
[0017] FIG. 1 illustrates a block diagram of a system for monitoring characteristics of sleep of a subject by generating a sleep time cycle interval based on a sensor data using a machine learning model that enables to evaluate one or more health conditions of the subject according to some embodiments herein;
[0018] FIG. 2 illustrates a block diagram of a sleep pattern analyzing server according to some embodiments herein;
[0019] FIG. 3A is an exemplary data table that includes data points that are collected when a subject interacts with a digital device for sleep pattern analysis according to some embodiments herein;
[0020] FIG. 3B is an exemplary data table that includes three-dimensional accelerometer data points collected from the accelerometer sensor for sleep pattern analysis according to some embodiments herein;
[0021] FIG. 4 is an exemplary user interface view of the digital device that depicts the output of sleep pattern analysis of the subject according to some embodiments herein;
[0022] FIG. 5 is a flow diagram that illustrates a method for monitoring characteristics of sleep of a subject by generating a sleep time cycle interval based on sensor data using a machine learning model that enables to evaluate one or more health conditions of the subject according to some embodiments herein; and
[0023] FIG. 6 is a schematic diagram of a computer architecture in accordance with the embodiments herein.
DETAILED DESCRIPTION OF THE DRAWINGS
[0024] The embodiments herein and the various features and advantageous details thereof are explained more fully with reference to the non-limiting embodiments that are illustrated in the accompanying drawings and detailed in the following description. Descriptions of well-known components and processing techniques are omitted so as to not unnecessarily obscure the embodiments herein. The examples used herein are intended merely to facilitate an understanding of ways in which the embodiments herein may be practiced and to further enable those of skill in the art to practice the embodiments herein. Accordingly, the examples should not be construed as limiting the scope of the embodiments herein.
[0025] As mentioned, there is a need for a system and method for monitoring characteristics of sleep of a subject by generating a sleep time cycle interval based on sensor data using a machine learning model that enables an evaluation of one or more health conditions of the subject. Referring now to the drawings, and more particularly to FIGS. 1 through 7, where similar reference characters denote corresponding features consistently throughout the figures, preferred embodiments are shown.
[0026] FIG. 1 illustrates a block diagram of a system 100 for monitoring characteristics of sleep of a subject 102 by generating a sleep time cycle interval based on sensor data using a machine learning model 110 that enables to evaluate one or more health conditions of the subject 102 according to some embodiments herein. The system 100 view includes a digital device 104, and a sleep pattern analyzing server 108. In some embodiments, the system 100 includes an android application package (APK), iOS App Store Package (IPA), or any such application packages that is installed in the digital device 104 of the subject 102. In some embodiments, the digital device 104 may include a mobile phone, a kindle, a PDA (Personal Digital Assistant), a tablet, or a smartphone. In some embodiments, the system 100 may include an application that may be installed in android based devices, windows-based devices, or any such mobile operating systems devices. The sleep pattern analyzing server 108 is configured to connect with at least one of a charging port, an accelerometer, a camera, a gyroscope, and an inbuilt speaker of the digital device 104.
[0027] The digital device 104 collects sensor data during a pre-defined time interval of the subject 102 and communicates to the sleep pattern analyzing server 108 through the network 106. In some embodiments, the pre-defined time interval maybe a day. In some embodiments, the digital device 104 stores the sensor data of the subject 102 in its local database before sending it to the sleep pattern analyzing server 108. In some embodiments, the network 106 is a wired network or a wireless network such as Bluetooth, Wi-Fi, ZigBee, cloud, or any other communication networks. In some embodiments, the sensor data of the subject 102 includes a user id, a timestamp of screen on, a timestamp of screen off, a timestamp of user interaction, a number of distinct applications associated with digital device 104 that is used by the subject 102, a total number of applications launched in the digital device 104, three-dimensional accelerometer data, three-dimensional gyroscope data, and a sleep cycle interval.
[0028] In some embodiments, data of the sensor data of the subject 102 are collected actively in real-time when the subject 102 is interacting with the digital device 104. In some embodiments, verification labels are assigned to such interactive subject 102. In some embodiments, the sensor data of the subject 102 are collected from the digital device 104 passively in real-time, for example, users with high risk, like individuals prone to erratic behavior, and a few collaborative users. In some embodiments, the sensor data of the subject 102 are collected from the data points obtained from real-time user events. In some embodiments, the data points obtained from the real-time user events may be data collected from a finger print scanner of the digital device 104, a face id of the digital device 104, a lock screen of the digital device 104, no lock data of the digital device 104. In some embodiments, the user id is a distinct identity given to each subject 102 who installs the system 100 in the digital device 104. In some embodiments, the timestamp of screen on includes a recorded data point when the screen of the digital device 104 is turned on. In some embodiments, the timestamp of screen off includes a recorded data point when the screen of the digital device 104 is turned off. In some embodiments, the number of distinct applications used by the subject 102 includes the applications used during user interaction in any one of the events that are either screen on event or screen off event. In some embodiments, the total number of applications launched in the digital device 104 includes the total number of applications that are launched during interaction of the subject 102 with the digital device 104.
[0029] In some embodiments, the three-dimensional accelerometer data points are collected during the pre-specified time interval. In some embodiments, the three-dimensional gyroscope data points are collected during a pre-defined time interval. In some embodiments, the three-dimensional gyroscopic data is collected when the subject 102 uses the digital device 104 before and after the sleep cycle interval. In some embodiments, the subject 102 who is active is asked to confirm an estimated sleep time or a wakeup time as a confirmation label using the machine learning model 110, which improves accuracy of the system 100. The machine learning model 110 may be a semi-supervised learning model. In some embodiments, the collected three-dimensional gyroscopic data points from the subject 102 are used to train the semi-supervised learning model.
[0030] In some embodiments, the sleep pattern analyzing server 108 dimensions of the domain features using at least one of principal component analysis or factor analysis to reduce a count of the domain features. The count is used to provide the domain features for training the machine learning model 110.
[0031] In some embodiments, for example, data points of three-dimensional gyroscope, three-dimensional accelerometer and their corresponding features that are angle of the digital device 104, altitude, each of them are taken as X-Y-Z dimensions of acceleration and gyroscope for reduction of dimensionality.
[0032] In some embodiments, if the subject 102 is prone to wandering behavior, the three-dimensional accelerometer data is collected, and if the subject 102 is not prone to the wandering behavior, then the data is not collected. In some embodiments, to assess performance of the sleep pattern of the subject 102, sensitivity, the sleep pattern analyzing server 108 may receive specificity, accuracy scores provided by collaborative users based on few user labels. In some embodiments, the sleep cycle interval is an average time at which the subject 102 goes to sleep.
[0033] The sleep pattern analyzing server 108 extracts data points from the sensor data using a data extraction method. In some embodiments, the data points which include noise, out-of-sync data are cleaned before extraction of data points using libraries of python. The sleep pattern analyzing server 108 groups the data points based on the pre-defined time interval to obtain data points groups. In some embodiments, for example, the data points that are collected in 5 minutes-30 minutes interval boxes are grouped. The sleep pattern analyzing server 108 generates domain features from the data points groups using a domain feature technique. The domain features include time-domain features, and frequency domain features. In some embodiments, the generated domain features are frequency, skewness, mean, kurtosis, and other higher-order intensities. The domain features are used as priors. The sleep pattern analyzing server 108 estimates time points of the user’s sleeping time and the user’s wake-up time by stacking the domain features that are correlated using a Bayesian model. The domain features that are correlated are selected using filtering, wrapping, and an embedding technique. In some embodiments, the estimation of time points is done separately/independently. In some embodiments, the stacks are provided based on a rule-based model.
[0034] The sleep pattern analyzing server 108 converts the three-dimensional accelerometer data into a motion intensity score using an acceleration vector. The sleep pattern analyzing server 108 generates the motion intensity score using the data points of the three-dimensional accelerometer data by, (i) determining acceleration vector components for each axis of the data points of the three-dimensional accelerometer data; and (ii) generating the motion intensity score by determining magnitudes of the acceleration vector components over a period of time. For example, the motion intensity score may be 0,1,2,3, etc. In some embodiments, the motion intensity score is obtained based on a rule-based model. In some embodiments, the motion intensity score may help to eliminate notifications or false positives, and quantity motion intensity. The sleep pattern analyzing server 108 measures a pacing intensity score based on a walking pattern during the sleep cycle interval by, (i) determining a pacing moment by combining the motion intensity score with a time-duration, and a walking axis; and (ii) determining the pacing intensity score using a duration of the pacing moment and a speed of motion. In some embodiments, the values of pacing intensity scores may be 0,1,2,3, etc. In some embodiments, the pacing intensity scores are measured using a rule-based model. In some embodiments, the motion intensity score is used to determine the one or more domain features. In some embodiments, the pacing intensity score is used to determine the quality of sleep.
[0035] The sleep pattern analyzing server 108 assigns a weight to each time point of the user’s sleeping time and the user’s wake-up time that is estimated. In some embodiments, the assigned weights are, for example, the weight of timestamp of screen on as 0.5, the weight of timestamp of screen off as 0.25, the weight of timestamp of user interaction as 0.4. In some embodiments, the weights are assigned using a rule-based model.
[0036] The sleep pattern analyzing server 108 generates the sleep time cycle interval using a switch point to monitor the characteristics of the sleep of the subject 102 that enables it to evaluate one or more health conditions of the subject 102. The switch point is calculated using the weight of each time point of the user’s sleeping time and the user’s wake-up time. The characteristics of sleep include a sleep interruption, type of interruption, and a quality of sleep. The sleep pattern analyzing server 108 analyses the sleep duration of the subject 102 over multiple days to identify changes in the sleep activity pattern. In some embodiments, based on the usage of digital device 104 patterns, motion features, and corresponding outputs of sleep quality, the sleep pattern analyzing server 108 evaluates the behavioral insights of the subject 102 that are related to mental health.
[0037] In some embodiments, the sleep pattern of the subject 102 is optimized based on the assessment of sleep/wake timing of the subject 102.
[0038] The output of the system 100 includes data points of a sleep time, a wake time, a sleep duration, a sleep onset latency, a wake onset latency, sleep disturbances, a duration of sleep disturbance, sleep quality. In some embodiments, the sleep time is a time point at which the subject 102 goes to bed, for example, if the subject goes to bed at 11:30 p.m., then the sleep time is 11:30 p.m. In some embodiments, the wakeup time is a time point at which the subject 102 wakes up from sleep, for example, if the subject wakes up from sleep at 06:30 a.m., then the wake time is 06:30 a.m. In some embodiments, the sleep duration is a duration of sleep of the subject 102, for example, the sleep duration of the subject is 6 hours, 30 minutes that is the duration in between the sleep time that is 11:30 p.m. to the wake-up time that is 06:30 a.m. In some embodiments, the wake onset latency is a time difference when the subject 102 went to bed and fell asleep. In some embodiments, the sleep disturbances are considered when the subject 102 uses the digital device 104 during their sleep cycle interval, for example, if the subject 102 uses the digital device 104 three times during their sleep cycle interval. In some embodiments, the duration of sleep disturbance is a total time spent by the subject 102 on the digital device 104 during their sleep cycle interval. In some embodiments, the sleep quality is estimated based on time spent trying to fall asleep and sleep disturbances.
[0039] FIG. 2 illustrates a block diagram of the sleep pattern analyzing server 108 according to some embodiments herein. The sleep pattern analyzing server 108 includes a database 202, a sensor data acquiring module 204, a data points extraction module 206, a data points grouping module 208, a domain features generating module 210, a time points estimating module 212, weights assigning module 214, and a sleep time cycle interval generating module 216. The sensor data acquiring module 204 acquires the sensor data of the subject 102 from the digital device 104 through the network 106 and stores it in the database 202 at the end of each day. In some embodiments, the sensor data includes at least one of a subject id, a timestamp of screen on, a timestamp of screen off, a timestamp of subject interaction, a number of distinct applications used by the subject, a total number of applications launched in the digital device 104, three-dimensional accelerometer data, three-dimensional gyroscope data, and a sleep cycle interval. In some embodiments, data of the sensor data are collected actively in real-time where the subject 102 is interacting with the digital device 104. The data points extraction module 206 extracts data points from the sensor data using a data extraction method. In some embodiments, the data points which include noise, out-of-sync data are cleaned before extraction of data points using libraries of python. In some embodiments, the specified time interval maybe a day.
[0040] The data points grouping module 208 groups the data points based on the pre-defined time interval to obtain data points groups. In some embodiments, for example, the data points that are collected in 5 minutes-30 minutes interval boxes are grouped. The domain features generation module 210 generates domain features from the data points groups using a domain feature technique. The domain features include time-domain features, and frequency domain features. In some embodiments, the generated domain features are frequency, skewness, mean, kurtosis, and other higher-order intensities. The time points estimating module 212 estimates time points of the user’s sleeping time and the user’s wake-up time by stacking the domain features that are correlated using a Bayesian model. The domain features that are correlated are selected using filtering, wrapping, and an embedding technique. In some embodiments, the estimation of time points is done separately/independently. In some embodiments, the stacks are performed based on a rule-based model. The weights assigning module 214 assigns a weight to each time point of the user’s sleeping time and the user’s wake-up time that is estimated. In some embodiments, the assigned weights are, for example, the weight of timestamp of screen on as 0.5, the weight of timestamp of screen off as 0.25, the weight of timestamp of user interaction as 0.4. In some embodiments, the weights are assigned using a rule-based model. The sleep time cycle interval generating module 216 generates the sleep time cycle interval using a switch point to monitor the characteristics of the sleep of the subject 102 that enables to evaluate one or more health conditions of the subject 102. The switch point is calculated using the weight of each time point of the user’s sleeping time and the user’s wake-up time. The characteristics of sleep include a sleep interruption, type of interruption, and a quality of sleep.
[0041] The machine learning model 110 is trained by providing historical weights and historical sleep time cycle intervals associated with one or more subjects as training data to obtain the trained machine learning model.
[0042] FIG. 3A is an exemplary data table 300A that includes data points that are collected when a subject 102 interacts with a digital device 104 for sleep pattern analysis according to some embodiments herein. The exemplary data table 300A includes data points of a subject id, a timestamp of screen on, a timestamp of screen off, a timestamp of subject interaction, a number of distinct applications used by the subject 102, a total number of applications launched in the digital device 104. In some embodiments, the subject id is the distinct identity given to each subject. In some embodiments, the timestamp of screen on includes a recorded data point when the screen of the digital device 104 is turned on. In some embodiments, the timestamp of screen off includes a recorded data point when the screen of the digital device 104 is turned off. In some embodiments, the number of distinct applications used by the subject includes the applications used during subject interaction in one event either screen on or screen off. In some embodiments, the total number of applications launched in the digital device 104 includes the total number of applications that are launched during interaction of the subject 102 with the digital device 104.
[0043] FIG. 3B is an exemplary data table 300B that includes three-dimensional accelerometer data points collected from the accelerometer sensor sleep pattern analysis according to some embodiments herein. The exemplary data table 300B includes data points of three-dimensional accelerometer data points associated with the subject 102 during his usage of the digital device 104 and an average sleep interval. In some embodiments, the three-dimensional accelerometer data points are collected at pre-specified time intervals.
[0044] FIG. 4 is an exemplary user interface view 400 of the digital device 104 that depicts the output of sleep pattern analysis of the subject 102 according to some embodiments herein. The exemplary user interface view 400 depicts output of sleep pattern analysis comprising a sleep time, a wake time, a sleep duration, a sleep onset latency, a wake onset latency, sleep disturbances, a duration of sleep disturbance, sleep quality. In some embodiments, the sleep time is a time point at which the subject 102 goes to bed, for example, if the subject goes to bed at 11:30 p.m., then the sleep time is 11:30 p.m. In some embodiments, the wake time is a time point at which the subject 102 wakes up from sleep, for example, if the subject wakes up from sleep at 06:30 a.m., then the wake time is 06:30 a.m. In some embodiments, the sleep duration is a duration of sleep of the subject 102, for example, the sleep duration of the subject is 6 hours, 30 minutes that is the duration in between the sleep time that is 11:30 p.m. to wake up time that is 06:30 a.m. In some embodiments, the wake onset latency is a time difference when the subject 102 went to bed and fell asleep. In some embodiments, the sleep disturbances are considered when the subject 102 uses the digital device 104 during their sleep cycle interval, for example, if the subject uses the digital device 104 for 3 times during their sleep cycle interval. In some embodiments, the duration of sleep disturbance is a total time spent by the subject 102 on the digital device 104 during their sleep cycle interval. In some embodiments, the sleep quality is estimated based on time spent trying to fall asleep and sleep disturbances.
[0045] FIG. 5 is a flow diagram 500 that illustrates a method for monitoring characteristics of sleep of a subject 102 by generating a sleep time cycle interval based on a sensor data using a machine learning model 110 that enables to evaluate one or more health conditions of the subject 102 according to some embodiments herein. At step 502, the method includes the step of obtaining the sensor data from a digital device that includes at least one of a screen button, a gyroscope, or an accelerometer during a pre-defined time interval. The sensor data includes at least one of a timestamp of screen on, a timestamp of screen off, a timestamp of subject interaction, a three-dimensional accelerometer data, or a three-dimensional gyroscope data. At step 504, the method includes the step of extracting data points after cleaning at least one of the noise or out-of-sync data from the sensor data using a data extraction method. At step 506, the method includes the step of grouping the data points based on the pre-defined time interval to obtain one or more data points groups. At step 508, the method includes the step of generating one or more domain features from one or more data points groups using a domain feature technique. In some embodiments, one or more domain features include time-domain features, and frequency domain features. At step 510, the method includes the step of estimating time points of subject’s sleeping time and subject’s wake-up time by stacking one or more domain features that are correlated. In some embodiments, one or more domain features that are correlated are selected using filtering, wrapping, and an embedding technique. At step 512, the method includes the step of assigning a weight to each time point of the subject’s sleeping time and the subject’s wake-up time that is estimated using a Bayesian model. At step 514, the method includes the step of generating the sleep time cycle interval using a switch point to monitor the characteristics of the sleep of the subject that enables it to evaluate one or more health conditions of the subject. In some embodiments, the switch point is calculated using the weight of each time point of the subject’s sleeping time and the subject’s wake-up time. In some embodiments, the characteristics of sleep include a sleep interruption, type of interruption, and a quality of sleep.
[0046] In some embodiments, the processor is configured to train the machine learning model by providing one or more historical weights and one or more historical sleep time cycle intervals associated with one or more subjects as training data to obtain the trained machine learning model.
[0047] In some embodiments, the processor is configured to reduce dimensions of the one or more domain features using at least one of principal component analysis or factor analysis to reduce a count of the one or more domain features, the count is used to provide the one or more domain features for training the machine learning model.
[0048] In some embodiments, the time-domain features and the frequency domain features comprise a frequency, a skewness, a mean, a kurtosis, and higher-order intensities.
[0049] In some embodiments, the processor is configured to generate a motion intensity score using the data points of the three-dimensional accelerometer data by, (i) determining acceleration vector components for each axis of the data points of the three-dimensional accelerometer data; and (ii) generating the motion intensity score by determining magnitudes of the acceleration vector components over a period of time.
[0050] In some embodiments, the processor is configured to measure a pacing intensity score based on a walking pattern during the sleep cycle interval by, (i) determining a pacing moment by combining the motion intensity score with a time-duration, and a walking axis; and (ii) determining the pacing intensity score using a duration of the pacing moment and a speed of motion.
[0051] In some embodiments, the motion intensity score is used to determine the one or more domain features.
[0052] In some embodiments, the pacing intensity score is used to determine the quality of sleep.
[0053] In some embodiments, the processor is configured to detect a pattern across the data points for grouping the data points using k-means clustering and t-distributed stochastic neighbor embedding.
[0054] A representative hardware environment for practicing the embodiments herein is depicted in FIG. 6, with reference to FIGS. 1 through 5. This schematic drawing illustrates a hardware configuration of a sleep pattern analyzing server 108 /computer system/ computing device in accordance with the embodiments herein. The system includes at least one processing device CPU 10 that may be interconnected via system bus 15 to various devices such as a random access memory (RAM) 12, read-only memory (ROM) 16, and an input/output (I/O) adapter 18. The I/O adapter 18 can connect to peripheral devices, such as disk units 58 and program storage devices 50 that are readable by the system. The system can read the inventive instructions on the program storage devices 50 and follow these instructions to execute the methodology of the embodiments herein. The system further includes a subject interface adapter 22 that connects a keyboard 28, mouse 50, speaker 52, microphone 55, and/or other subject interface devices such as a touch screen device (not shown) to the bus 15 to gather subject input. Additionally, a communication adapter 20 connects the bus 15 to a data processing network 52, and a display adapter 25 connects the bus 15 to a display device 26, which provides a graphical subject interface (GUI) 56 of the output data in accordance with the embodiments herein, or which may be embodied as an output device such as a monitor, printer, or transmitter, for example.
[0055] The system and/or method is used for generating suggestions to healthcare experts related to psychiatric, neuropsychiatric, mental illness, neurological, neuro-psychotic disorders using sleep pattern analysis through the sleep cycle interval of a patient. The system or method may help the experts in smart-phone based sleep assessments and interventions that require an emphasis on promoting long-term adherence, exploring possibilities of adaptive and personalized systems to predict risk/relapse, and determining the impact of sleep monitoring on improving patients’ quality of life and clinically meaningful outcomes. In psychiatric disorders, sleep staging including quantification of slow-wave sleep and spindles may have potential diagnostic and prognostic applications. Changes in sleep are also early warning signs of relapse in schizophrenia or conversion in schizophrenia prodrome. Also, the system or method may be beneficial to assess sleep pattern analysis during the sleep cycle interval of a patient.
[0056] The foregoing description of the specific embodiments will so fully reveal the general nature of the embodiments herein that others can, by applying current knowledge, readily modify and/or adapt for various applications such specific embodiments without departing from the generic concept, and, therefore, such adaptations and modifications should and are intended to be comprehended within the meaning and range of equivalents of the disclosed embodiments. It is to be understood that the phraseology or terminology employed herein is for the purpose of description and not of limitation. Therefore, while the embodiments herein have been described in terms of preferred embodiments, those skilled in the art will recognize that the embodiments herein can be practiced with modification within the spirit and scope of the appended claims.
,CLAIMS:CLAIMS
I/We Claim:

1. A system (100) for monitoring characteristics of sleep of a subject (102) by generating a sleep time cycle interval based on a sensor data using a machine learning model (110) that enables to evaluate one or more health conditions of the subject (102), the system (100) comprising:
a digital device (104) that comprises at least one of a screen button, a gyroscope, or an accelerometer, wherein the digital device (104) obtains the sensor data during a pre-defined time interval, wherein the sensor data comprises at least one of a timestamp of screen on, a timestamp of screen off, a timestamp of subject interaction, a three-dimensional accelerometer data, or a three-dimensional gyroscope data;
a sleep pattern analyzing server (108) that acquires the sensor data of the subject (102) from the digital device (104), and processes, using the machine learning model (110), the sensor data, wherein the sleep pattern analyzing server (108) comprises:
a memory that stores a database and a set of instructions;
a processor that is configured to execute the machine learning model (110) and is configured to
extract, using a data extraction method, data points after cleaning at least one of noise or out-of-sync data from the sensor data;
group the data points based on the pre-defined time interval to obtain a plurality of data points groups;
characterized in that,
generate, using a domain feature technique, a plurality of domain features from the plurality of data points groups, wherein the plurality of domain features comprises time-domain features, and frequency domain features;
estimate, using a Bayesian model, time points of subject’s sleeping time and subject’s wake-up time by stacking the plurality of domain features that are correlated, wherein the plurality of domain features that are correlated are selected using filtering, wrapping, and an embedding technique;
assign, using a Bayesian model, a weight to each time point of the subject’s sleeping time and the subject’s wake-up time that is estimated; and
generate, using a trained machine learning model, the sleep time cycle interval using a switch point to monitor the characteristics of the sleep of the subject (102) that enables to evaluate one or more health conditions of the subject (102), wherein the switch point is calculated using the weight of each time point of the subject’s sleeping time and the subject’s wake-up time, wherein the characteristics of sleep comprises a sleep interruption, type of interruption, and a quality of sleep.

2. The system (100) as claimed in claim 1, wherein the processor is configured to train the machine learning model (110) by providing a plurality of historical weights and a plurality of historical sleep time cycle intervals associated with a plurality of subjects as training data to obtain the trained machine learning model.

3. The system (100) as claimed in claim 1, wherein the processor is configured to reduce dimensions of the plurality of domain features using at least one of principal component analysis or factor analysis to reduce a count of the plurality of domain features, wherein the count is used to provide the plurality of domain features for training the machine learning model (110).

4. The system (100) as claimed in claim 1, wherein the time-domain features and the frequency domain features comprise a frequency, a skewness, a mean, a kurtosis, and higher-order intensities.

5. The system (100) as claimed in claim 1, wherein the processor is configured to generate a motion intensity score using the data points of the three-dimensional accelerometer data by,
determining acceleration vector components for each axis of the data points of the three-dimensional accelerometer data; and
generating the motion intensity score by determining magnitudes of the acceleration vector components over a period of time.

6. The system (100) as claimed in claim 1, wherein the processor is configured to measure a pacing intensity score based on a walking pattern during the sleep cycle interval by,
determining a pacing moment by combining the motion intensity score with a time-duration, and a walking axis;
determining the pacing intensity score using a duration of the pacing moment and a speed of motion.

7. The system (100) as claimed in claim 5, wherein the motion intensity score is used to determine the plurality of domain features.

8. The system (100) as claimed in claim 6, wherein the pacing intensity score is used to determine the quality of sleep.

9. The system (100) as claimed in claim 1, wherein the processor is configured to detect a pattern across the data points for grouping the data points using k-means clustering and t-distributed stochastic neighbor embedding.

10. A processor-implemented method for monitoring characteristics of sleep of a subject (102) by generating a sleep time cycle interval based on a sensor data using a machine learning model (110) that enables to evaluate one or more health conditions of the subject (102), the method comprising:
obtaining the sensor data from a digital device (104) that comprises at least one of a screen button, a gyroscope, or an accelerometer during a pre-defined time interval, wherein the sensor data comprises at least one of a timestamp of screen on, a timestamp of screen off, a timestamp of subject interaction, a three-dimensional accelerometer data, or a three-dimensional gyroscope data;
extracting, using a data extraction method, data points after cleaning at least one of noise or out-of-sync data from the sensor data;
grouping the data points based on the pre-defined time interval to obtain a plurality of data points groups;
generating, using a domain feature technique, a plurality of domain features from the plurality of data points groups, wherein the plurality of domain features comprises time-domain features, and frequency domain features;
estimating, using a Bayesian model, time points of subject’s sleeping time and subject’s wake-up time by stacking the plurality of domain features that are correlated, wherein the plurality of domain features that are correlated are selected using filtering, wrapping, and an embedding technique;
assigning, using a Bayesian model, a weight to each time point of the subject’s sleeping time and the subject’s wake-up time that is estimated; and
generating, using a trained machine learning model, the sleep time cycle interval using a switch point to monitor the characteristics of the sleep of the subject t(102) hat enables to evaluate one or more health conditions of the subject (102), wherein the switch point is calculated using the weight of each time point of the subject’s sleeping time and the subject’s wake-up time, wherein the characteristics of sleep comprises a sleep interruption, type of interruption, and a quality of sleep.
Dated this day of 08th December, 2021.

Bala Arjun Karthik
IN/PA -1021