System for sleep analysis and method of managing thereof

Abstract

Disclosed herein is a system designed to evaluate sleep quality through an interactive user interface The system initiates a series of questions, including bedtime, wake-up time, caffeine / alcohol consumption and other potential factors, to gather relevant sleep data; these evaluation parameters are diverse and comprehensive leading to a more personalized sleep evaluation result. The acquired data is further fed into regression based artificial neural network developed algorithms used to calculate sleep splits between deep sleep and light sleep durations, which generates a numerical measurement for the quality of sleep, known as the "sleep score", which is further classified into specific brackets. Based on the assigned sleep bracket, the system provides tailored feedback to enhance sleep quality and potentially elevate the sleep score. The suggested recommendations are curated from research articles and scientific journals exploring the correlation between sleep quality and overall efficiency. This system offers individualized inputs and recommendations, contributing to improved sleep health and overall well-being, thus enhancing productivity of the users.

Description

SYSTEM FOR SLEEP ANALYSIS AND METHOD OF MANAGING THEREOFTECHNICAL FIELD

[0001] The present invention relates to the field of sleep monitoring systems, and in particular it pertains to an evaluation method to improve sleep quality.BACKGROUND OF THE INVENTION

[0002] A large size of the population in today’s world suffer from various sleep disorders that affect their physical as well as mental conditions. Sleep disorders are affecting people of all ages. Poor sleep quality not only impacts physical health but also affects mental well-being and overall productivity. People who are deprived of restful sleep may experience difficulty in making decisions, lack of concentration, irritability, depression and anxiety.

[0003] Quality sleep is very important for the overall well-being of a person. Restful sleep can alleviate mental stress, ensure normal metabolism of the body, accelerating the healing of damaged tissues and also maintains mental health. Whereas, lack of sleep and sleep disorders can adversely affect human life by contributing to increased risk of obesity, diabetes, hypertension, heart attack, stroke and depression. Therefore, the quality of sleep directly affects human quality of life.

[0004] To tackle this situation, there are various sleep tracking / managing devices available in the market but the traditional sleep tracker methods primarily rely on limited objective measurements. The tracker does not identify the cause for variations in sleep quality. It follows a standardized mechanism to calculate users' sleep quality by focusing only on the sleep phase and body movements. These approaches often lack accuracy, comprehensive data collection, and it fails to address individual variations in sleep patterns.

[0005] A Japanese patent document: JP2009297474A provides a sleep stage determining means for determining the sleep stage using the 8-wave ratio of the brain wave which isobtained by the 5- wave ratio estimating means. However, the invention does not provide information on other critical aspects of sleep, such as REM sleep or deep sleep, which are important for a complete assessment of sleep quality.

[0006] A Korean patent document: KR102236420B1 provides Sleep analysis system using artificial Intelligence and method thereof that collects sleep analysis data according to the sleep state of a subject for sleep analysis by using an ultra wide band (UWB) radar, constructing a training dataset based on the sleep analysis data to train an artificial intelligence model, analyzing a sleep level (sleep quality) by applying the dataset for sleep analysis data for analyzing sleep quality to the trained artificial intelligence model, and providing a sleep care service guiding to take a deep sleep in response to the analyzed sleep level, and a method thereof. However, UWB radar signals might experience interference from other electronic devices or environmental factors, affecting the quality of data collection.

[0007] With advancements in technology, the sleep tracking industry has witnessed notable improvements. However, the current sleep monitoring systems may require end users to have complex and costly hardware. Further, they may have limited data interpretation capabilities and may not provide suggestive insights for users to improve their sleep quality effectively, by connecting data to behavioral modification needed. Additionally, these methods may not fully capture / measure the complexity of sleep, such as sleep continuity, disturbances, and activities before sleeping, which are crucial for comprehensive sleep evaluation. Given the limitations of the existing approaches, techniques, systems and methods towards their sustainable design, application, and economic feasibility, there exists a need to develop an improved sleep monitoring and analysis method that would address and resolve a variety of issues.

[0008] The present invention addresses the above challenges by incorporating advanced technology and novel algorithms to deliver a more accurate and user-friendly sleep model.SUMMARY OF THE INVENTION

[0009] Among the various negative impacts the pandemic has had on people, disruption of sleep schedules and irregularities in sleep patterns is often overlooked. Most people don't receive adequate sleep, which has ill effects on their health and efficiency. To help keep tabs on the deteriorating importance of sleep among the current generation, the current invention has been developed. Unlike wearable sleep trackers that need to be worn throughout the night, which can be a hassle at times, this invention eliminates the need for the same. With the help of a short questionnaire, it analyzes the sleep quality of the user and provides suggestions on how to improve the same. It also proves advantageous over wearable sleep trackers in the aspect of cost as it is solely software-based, saving on cost for hardware components.

[0010] It is an object of the present invention to provide a software program model for determining the quality of the user's sleep. The software program comprises an algorithm and / or a method that has been trained on sleep data collected by a wearable sleep tracker. The program further compares the input to the trends it developed based on its training data. The program code comprises instructions for performing the method required to determine the quality of sleep. One may note that the wearable sleep tracker data may not be of the end user, but is rather used to train the model.

[0011] A further object of the present invention is to develop a model integrated with a score based output system and provide suggestions to improve sleep quality.

[0012] The functionality, or rather the UI of the model is such that it asks a series of questions to the user and based on the responses, runs it through the Al algorithms and obtains a numerical value for sleep quality, called the sleep score. Then the sleep score is placed into brackets and accordingly assessed for shortcomings. These are then addressed in the form of suggestions and tips which are given to the user. These tips over a span of time can help improve sleep quality as proven by scientific research and prior datasets.

[0013] Usually, a wearable sleep tracker may not identify the cause for variations in sleep quality, nor does it discuss the effects of these variations. It follows a standardized mechanism to calculate your sleep quality by focusing only on the sleep phase. The present invention, however, gives users a more comprehensive idea on how their sleep score affects them.

[0014] In one embodiment, a system for sleep analysis and method of managing thereof, the method comprising steps of: gathering sleep data using user interface, presenting users with series of sleep related questions, collecting user data to input into a regression based artificial neural network, comparing user input with trends derived from wearable sleep tracker using Al algorithm, generating a sleep quality score, classifying sleep score and displaying the appropriate suggestion to improve sleep quality.

[0015] In one embodiment, the sleep data includes total hours of sleep per night, bedtime and wakeup time, time taken to fall asleep after lying down, activities before bedtime.

[0016] In one embodiment, series of sleep related questions includes how many times you wake up at night, how long will it take you to fall back asleep if you wake up at night, do you consume caffeine or alcohol close to bedtime, do you have any medical conditions that affect your sleep or other questions useful for identifying sleep issues.

[0017] In one embodiment, regression based neural network model is trained on 2000+ data points, collected first hand by using a sleep tracker and produces highly accurate output with a Mean Squared Error: 3.48, Root Mean Squared Error: 1.86 and a Mean Accuracy Error: 1.38. The model is integrated with a score based output system and corresponding suggestions to improve sleep quality as well.

[0018] In one embodiment, the method of obtaining output and corresponding suggestions to improve sleep quality, further comprising steps of collecting data from sleep tracking devices, that includes sleep duration, sleep cycles, and other relevant sleep-related metrics, calculating sleep duration and processing of other factors: generating deep sleep light sleep split, monitoring and analyzing of data collected through various wearable devices using digital tool called as sleep buddy, classification of obtained sleep score and finally displaying results with recommendations.

[0019] In one embodiment, the factors taken into consideration for sleep analysis include time of sleeping and waking up, screen time before bed, stress level, sleep continuity, caffeine alcohol consumption, heavy meal close to sleeping and sleep continuity.

[0020] In one embodiment, generating deep sleep light sleep split helps to calculate the duration of deep sleep and light sleep, enabling the determination of a comprehensive sleep score. The ratio of deep sleep to light sleep is approximately 37 to 63.

[0021] In an embodiment, the numerical measurement known as the sleep score, is categorized into different brackets i.e poor < 75 < moderate < 85 < good < 95 < excellent.

[0022] In one embodiment, the recommendations given to the user includes consistent sleep schedule, avoiding caffeine or heavy meals close to bedtime, and creating a relaxing pre-sleep routine.

[0023] In an embodiment, the data from sleep tracking devices includes sleep duration, sleep cycles, and other relevant sleep-related metrics.

[0024] In one embodiment, the method of collecting data as per claim further comprising steps of collecting data using a wearable sleep tracker that includes time of sleeping, waking up and sleep score, visualizing data to finalize data features, initiation using sleep split formula, predicting continuous output variables using regression based neural network which includes sleep duration, sleep stages, sleep percentages, sleep latency and other sleep metrics, training of ANN models over a period exceeding 6 months, and exporting Al model for usage and deployment.

[0025] In one embodiment, the sleep split formula divides sleep data into different stages or segments, such as light sleep, deep sleep.

[0026] In an embodiment, ANN model ran over 1800 epochs with data randomization and shuffling to avoid overfitting.

[0027] In one embodiment, the method of operating the sleep analysis system comprises steps of taking input data from the user, the input data is fed to Keras MLh5 model, the efficiency score is then shown as an output on the app.

[0028] In an embodiment, the Keras ML h5 model refers to a model saved in the HDF5 format. Keras, a popular high-level neural network API written in Python.

[0029] In an embodiment, the module analyzes sleep data in real-time and generates reports or visualizations of sleep patterns and quality as an output.

[0030] In one embodiment, the module is further supplemented by structured query language based user data collection and Java Script based user interface with a score dependent suggestion mechanism, and these components have been compiled using Python programming language.BRIEF DESCRIPTION OF THE ACCOMPANYING DRAWINGS

[0031] In order to understand the invention and to see how it may be carried out in practice, embodiments will now be described, by way of non-limiting examples only with reference to the accompanying figures.

[0032] Figure la, lb and lc are flowcharts illustrating the various steps employed in the system, in accordance with an embodiment of the present invention.

[0033] Figure 2 illustrates operation of the system, in accordance with an embodiment of the present invention.DETAILED DESCRIPTION OF THE INVENTION

[0034] Various embodiments described herein are intended only for illustrative purposes and subject to variations. It is understood that various omissions and substitutions of equivalents are contemplated as circumstances may suggest or render expedient, but are intended to cover the application or implementation without departing from the scope of the present invention. Also, it is to be understood that the phraseology and terminology used herein is for the purpose of description and should not be regarded as limiting.

[0035] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used herein, the term “and / or” includes any and all combinations of one or more of the associated listed items. As used herein, the singular forms “a”, “an” and “the” are intended to include the plural forms as well as the singular forms, unless the context clearly indicates otherwise. It will be further understood that the terms “comprises” and / or “comprising”, when used in this specification, specify the presence of stated features, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, steps, operations, elements, components, and / or groups thereof.

[0036] Unless, otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one having ordinary skill in the art to which this invention belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and the present disclosure and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.

[0037] In describing the invention, it will be understood that a number of techniques and steps are disclosed. Each of these has individual benefits and each can also be used in conjunction with one or more, or in some cases all, of the other disclosed techniques. Accordingly, for the sake of clarity, this description will refrain from repeating every possible combination of the individual steps in an unnecessary fashion.

[0038] With reference to Figure la, lb, le a method for sleep evaluation and treating poor sleep behavior is disclosed. The system consists of an Artificial Intelligence based sleep analyst that provides users with reports of their sleep quality and also suggests methods to improve them based on a questionnaire.

[0039] The model operates by gathering sleep data using a user interface that includes total hours of sleep per night, bedtime and wakeup time, time taken to fall asleep after lying down, activities before bedtime. Presenting users with a simple series of questions regarding their bedtime, wake-up time, and other potential factors that might impact their sleep. The gathered user data is then input into an artificial neural network (ANNs) thatutilizes regression techniques. This algorithm has undergone training using sleep data acquired from a sleep tracking device over a period exceeding 6 months. The algorithm then compares user input derived from wearable devices which provides information, including sleep duration, sleep cycles, and other relevant sleep-related metrics resulting in a numerical measurement known as the sleep score, determining the quality of sleep. Subsequently, the sleep score is categorized into different brackets i.e. poor < 75 < moderate < 85 < good < 95 < excellent. Furthermore, the model furnishes feedback based on the assigned sleep bracket and offers suggestions to improve users' sleep quality.

[0040] In user sleep data the factors taken into consideration for sleep analysis includes time of sleeping and waking up, screen time before bed, stress level, sleep continuity, caffeine alcohol consumption, heavy meal close to sleeping and sleep continuity. Further the neural network developed algorithms and formulas are used to calculate sleep splits between deep sleep and light sleep durations. The "deep sleep" and "light sleep" split refers to how sleep is categorized based on brain activity and physiological functions. Deep Sleep, also known as slow-wave sleep, is characterized by slower brain waves (delta waves). It's crucial for physical restoration and memory consolidation. During deep sleep, heart rate and breathing slow down, and it's harder to wake someone. The algorithm can identify deep sleep by analyzing brain wave patterns, heart rate variability, and movement data, whereas light sleep is characterized by a mix of alpha and theta brain waves. It's a transition period where the body and mind are preparing to enter deeper sleep stages. During light sleep, individuals are more easily aroused compared to deep sleep. The algorithm can detect light sleep by monitoring these brain wave patterns, along with slight movements and changes in heart rate. Sleep analysis systems use algorithms and machine learning to analyze data from wearable devices or sleep studies, often integrating multiple data sources such as EEG (electroencephalogram), heart rate, and motion sensors to accurately distinguish between these sleep stages. Based on analysis conducted on the data visualization graphs and scientific research, the deep sleep-light sleep ratio has been determined to be roughly 37-63 and is seen to dip by around 5% for each detrimental factor observed. This ratio is applied to the Al algorithm which then provides the sleep score accordingly.

[0041] A "sleep buddy" provided in a sleep analysis system is a digital assistant or tool designed to monitor, analyze, and improve sleep patterns. It uses artificial intelligence tocollect data through various sensors or input methods, such as wearable sleep trackers , smartphone apps, or smart home equipment. The sleep score is then classified evaluating several metrics to provide a comprehensive assessment of sleep quality. Following factors were taken into consideration for evaluating sleep scores:• Sleep Efficiency: This measures the percentage of time spent asleep while in bed. High sleep efficiency indicates good sleep quality, while low efficiency suggests difficulties falling or staying asleep.• Total Sleep Time (TST): This is the total amount of time spent sleeping. It helps in assessing whether the individual is getting enough sleep.• Sleep Latency: This measures how long it takes to fall asleep after going to bed. Shorter sleep latency generally indicates good sleep health.• Wake After Sleep Onset (WASO): This represents the amount of time spent awake after initially falling asleep. Lower WASO values are typically better, as they indicate fewer interruptions.• Sleep Stages Distribution: Sleep analysis system analyzes the proportions of light, deep, and REM sleep. Adequate amounts of each stage are crucial for restorative sleep.• Heart Rate Variability (HRV): Variations in heart rate during sleep can provide insights into autonomic nervous system activity and overall sleep quality.• Movement Data: This involves tracking body movements during sleep, which can indicate disturbances or restlessness.• Sleep Score: Based on these metrics, an overall sleep score is calculated. This score often ranges from 0 to 100 and summarizes sleep quality. A higher score typically reflects better sleep quality, while a lower score indicates potential issues.

[0042] The algorithms in the system use machine learning models to interpret various sleep metrics, integrating data from wearable sleep trackers, smart mattresses, or other sensors to generate a detailed and personalized sleep assessment. Further the users receive recommendations to enhance their sleep quality. These recommendations includes• Sleep Hygiene: Maintain a consistent sleep schedule, avoid caffeine or heavy meals close to bedtime, and create a relaxing pre-sleep routine.• Environment: Ensure your sleep environment is conducive to rest — cool, dark, and quiet.• Activity Level: Increase physical activity during the day to promote better sleep, but avoid vigorous exercise close to bedtime.• Sleep Patterns: Consider tracking and adjusting your sleep patterns, such as limiting naps during the day or managing time in bed to align with your body's natural rhythm.• Stress Management: Engage in stress-reducing activities like meditation, deep breathing, or light reading before bed.• Health Check: Consult with a healthcare professional if sleep issues persist, as underlying health conditions may affect sleep quality.

[0043] These suggestions are personalized based on the insights gathered by the Al system, aiming to address specific issues identified in your sleep patterns.

[0044] The process of analyzing sleep quality as described in Figure lc involves sleep data that is in the form of sleeping , waking up, sleep score. Sleep data is collected from various sources, such as wearable devices or sleep studies. This data might include variables like total sleep time, sleep efficiency, duration of REM sleep, and occurrences of wakefulness. This data is applied to identify latent factors that explain patterns in thesleep data. For example, factors might represent different aspects of sleep quality or disturbances.

[0045] The data analyzed is through primary factor variance analysis that will determine how much of the total variance in the sleep data can be attributed to these underlying factors. The "primary factor" is the one that explains the most variance in the dataset. For instance, a primary factor might be "overall sleep quality", which encompasses various measurements like sleep duration and interruptions. By examining the primary factor, researchers or clinicians can gain insights into the main contributors to sleep patterns or issues. This can help in designing targeted interventions or improving sleep-related recommendations. Overall, primary factor variance analysis in sleep data helps in simplifying complex data and focusing on the most significant factors affecting sleep quality or disturbances. This algorithm has undergone training using sleep data acquired from a sleep tracking device over a period exceeding 6 months.

[0046] Further data visualization in the context of finalizing data features for an sleep analysis system involves creating graphical representations of the data to understand its structure, patterns, and relationships better. This process can help in several ways like: Visualizing features can reveal how they relate to each other and to the target variable (e.g., sleep quality). For instance, scatter plots can show the relationship between sleep duration and sleep efficiency. Visualizations like box plots can highlight anomalies or outliers in the data, which might need to be addressed or removed. Histograms or density plots can illustrate the distribution of individual features, helping to understand their spread and skewness. Time series plots can help identify trends and patterns over time, such as changes in sleep patterns across different days or months. Visualizations can assist in identifying missing values or erroneous data points that need correction. Trends may hence refer to intelligent data or correlations drawn based on first hand data available of wearable sleep tracking devices, with the sleep parameters such as eating habit, electronic device use etc. In one exemplary embodiment, a regression based neural network model may be trained on 2000+ data points, collected first hand by using a sleep tracker to generate said trends.

[0047] Overall, effective data visualization helps in making informed decisions about which features to include, modify, or exclude, ultimately leading to a more robust and accurate sleep analysis system.

[0048] Once the data is visualized it is initiated using the sleep split formula, the sleep split formula helps in segmenting sleep data into distinct phases (like REM, light, and deep sleep). This initialization process is crucial for accurate analysis. By dividing sleep into these phases, the system can better analyze patterns, predict sleep quality, and provide insights. This segmented data serves as the foundation for training Al models to improve sleep recommendations and interventions. The "sleep split formula" typically refers to a method for dividing sleep data into different stages or segments, such as wake, light sleep, deep sleep, and REM. This is often done using algorithms based on polysomnography data or wearable sensors. For example, a common formula involves analyzing changes in heart rate, movement, and brain wave activity to determine sleep stages. These segments are then used to train Al models or for further analysis to understand sleep patterns and quality.

[0049] The Artificial Intelligence model used in the sleep analysis system is a regression based supervised learning Artificial Neural Network, trained on 2000+ data points and is collected directly by using a sleep tracker. The present invention is a software based application with a future scope of global coverage through websites or app integrations. The development of this invention is mainly centered around the Al model in its back end. The Al model is a regression based supervised learning Artificial Neural Network, trained on 2000+ data points, collected first hand by using a sleep tracker. The Al model ran over 1800 epochs with data randomization and shuffling to avoid overfitting. It produced highly accurate output with a Mean Squared Error: 3.48, Root Mean Squared Error: 1.86 and a Mean Accuracy Error: 1.38. The results were accurate and gave satisfactory insight on the sleep quality when compared to other popular and expensive sleep trackers. The model is integrated with a score based output system and corresponding suggestions to improve sleep quality as well.

[0050] A regression-based neural network for a sleep monitoring system uses Al to predict and analyze various aspects of sleep based on input data. The system collects data from sleep monitoring devices, which may include variables such as heart rate, bodymovement, sleep stage duration (light, deep, REM), and sleep onset time. It then extracts features from raw data, such as average heart rate during sleep, frequency of awakenings, total sleep duration, and time spent in different sleep stages. Multiple layers with neurons in a regression based neural network apply nonlinear transformations to the input data, and so provide continuous outputs, such as predicted total sleep time, sleep efficiency, or the duration of specific sleep stages. This network uses functions like ReLU (Rectified Linear Unit) in hidden layers and linear activation in the output layer to handle regression tasks and use historical sleep data with known outcomes to train the model. The data should be split into training and validation sets, and apply a loss function suitable for regression tasks, such as Mean Squared Error (MSE), to measure prediction accuracy. The network utilizes optimization algorithms like Adam or SGD (Stochastic Gradient Descent) to minimize the loss function and improve model performance and further uses the trained model to predict continuous variables related to sleep, such as:• Sleep Duration: Predict total sleep time.• Sleep Efficiency: Estimate the efficiency of sleep.• Sleep Stages: Predict the duration of different sleep stages.• Evaluation: Assess model performance using metrics like R-squared, Mean Absolute Error (MAE), or Root Mean Squared Error (RMSE).

[0051] Replacing a regression- based neural network in an Al-enabled sleep analysis system depends on your specific goals and data characteristics. Here are some alternatives to consider:A. Convolutional Neural Networks (CNNs): Useful if you have time-series data or sleep data that can be treated as spatially structured. CNNs can automatically learn hierarchical features from raw data, which might be advantageous for detecting patterns in sleep stages or anomalies.B. Recurrent Neural Networks (RNNs) and Long Short-Term Memory Networks (LSTMs): Effective for sequential data, such as time-series sleep data. LSTMs, in particular, are good at capturing long-term dependencies in sequences, which is useful for modeling sleep cycles over time.C. Transformers: Known for their strong performance in sequence modeling tasks. Transformers can handle long-range dependencies in data better than RNNs and might be suitable for analyzing complex sleep patterns.D. Autoencoders: If your goal is to extract features or reduce dimensionality, autoencoders can be used for unsupervised learning. They might help in feature extraction from sleep data, which can then be used with other models.E. Ensemble Methods: Combining multiple models, such as decision trees or gradient boosting machines, with neural networks might improve performance. Ensemble methods can aggregate predictions from different models to achieve better accuracy.F. Attention Mechanisms: Incorporating attention mechanisms into your model can help focus on specific parts of the data that are more relevant for sleep analysis, improving the model's interpretability and performance.

[0052] The choice of replacement depends on the nature of your sleep data, the problem you're addressing, and the performance metrics you aim to optimize.

[0053] The ANN model in the system can effectively analyze and predict sleep patterns, offering valuable insights into sleep quality and potential disorders. The model ran over 1800 epochs with data randomization and shuffling to avoid overfitting. It produced highly accurate output with a Mean Squared Error: 3.48, Root Mean Squared Error: 1.86 and a Mean Accuracy Error: 1.38.

[0054] Exporting an Al model for deployment in a sleep analysis system ensures that the model has been thoroughly tested and evaluated to confirm its performance meets the desired criteria. It converts the model to a format that can be used with differentplatforms or frameworks if needed, ensuring that all necessary dependencies (libraries, frameworks) are included or documented for the environment where the model will be deployed , also it provides clear documentation on how to use the model, including any required preprocessing steps or input / output formats.

[0055] Figure 2 illustrates operation of the system, in accordance with an embodiment of the present invention.

[0056] The gathered user data that is collected from various sources, such as wearable devices or sleep studies. This data might include variables like total sleep time, sleep efficiency, duration of REM sleep, and occurrences of wakefulness is fed as an input in Keras MLh5 module which typically refers to a module saved in the HDF5 format. Keras, a popular high-level neural network API written in Python, allows for easy creation, training, and deployment of deep learning modules. The module analyzes sleep data in real-time and generates reports or visualizations of sleep patterns and quality as an output. The module is further supplemented by structured query language based user data collection and Java Script based user interface with a score dependent suggestion mechanism, and these components have been compiled using Python programming language.

Claims

I / We claim:

1. A method for sleep analysis and management thereof (100), the method (100) comprising steps of: a. presenting users with series of sleep related questions; b. gathering sleep data using user interface; c. feeding user data into a regression based artificial neural network; d. comparing user input with trends derived from wearable sleep tracker using Artificial Intelligence algorithm; e. generating a sleep quality score; f. classifying sleep score; and g. displaying the appropriate suggestion to improve sleep quality.

2. The method for sleep analysis and method of managing thereof (100) as claimed in claim 1 wherein, the sleep data includes total hours of sleep per night, bedtime and wakeup time, time taken to fall asleep after lying down, activities before bedtime.

3. The method for sleep analysis and management thereof (100) as claimed in claim 1, wherein series of sleep related questions includes number and frequency of wake up cycle at night, time taken to fall back asleep if user wakes up at night, consumption of caffeine or alcohol close to bedtime, presence of any medical conditions that affect your sleep or other questions useful for identifying sleep issues.

4. A method for sleep analysis and management thereof (100) as claimed in claim 1 wherein, a regression based neural network model is trained on data points, collected first hand by using a sleep tracker to generate said trends.

5. The method for sleep analysis and management thereof (100) as claimed in claim 1 wherein, regression based neural network model is integrated with a score based output system and corresponding suggestions to improve sleep quality as well.

6. A method for sleep analysis and management thereof to improve sleep quality, comprising steps of: a. collecting data from sleep tracking devices that includes sleep duration, sleep cycles, and other relevant sleep-related metrics; b. calculating sleep duration and processing of other factors; c. generating deep sleep and light sleep split; d. monitoring and analyzing of data collected through various wearable devices using digital tool; e. classification of obtained sleep score; and f. displaying results with recommendations.

7. The method for sleep analysis and management thereof as per claim 6 wherein, the factors taken into consideration for sleep analysis includes time of sleeping and waking up, screen time before bed, stress level, sleep continuity, caffeine / alcohol consumption, heavy meal close to sleeping and sleep continuity.

8. The method for sleep analysis and management thereof as per claim 7 wherein, generating deep sleep, light sleep split is used to calculate the duration of deep sleep and light sleep, for the determination of a comprehensive sleep score.

9. The method for sleep analysis and management thereof as per claim 7 wherein, the data from sleep tracking devices includes sleep duration, sleep cycles, and other relevant sleep-related metrics.

10. The method of creating trends derived from wearable sleep tracker using Artificial Intelligence comprising steps of: a. collecting data using a wearable sleep tracker; b. visualizing data to finalize data features; c. initiation using sleep split formula; d. predicting continuous output variables using regression based neural network; e. training of ANN model; f. exporting Al model for usage and deployment.

11. The method of creating trends derived from wearable sleep tracker using Artificial Intelligence as in claim 10 wherein, the collected data is in the form of time of sleeping, waking up and sleep score.

12. The method of creating trends derived from wearable sleep tracker using Artificial Intelligence as in claim 10 wherein, the sleep split formula divides sleep data into different stages or segments, such as light sleep, deep sleep.

13. The method of creating trends derived from wearable sleep tracker using Artificial Intelligence as in claim 10 wherein, the continuous variable output includes sleep duration, sleep stages, sleep percentages, sleep latency and other sleep metrics.

14. The method of creating trends derived from wearable sleep tracker using Artificial Intelligence as in claim 10 wherein, the ANN model has undergone training using sleep data acquired from a sleep tracking device over a period of time.

15. The method of operating the sleep analysis system comprising steps ofa. taking input data from user; b. the input data is fed to Keras ML h5 model; and c. obtaining output as an efficiency score on the app.

16. The method of operating the sleep analysis system as per claim 20 wherein, the module analyzes sleep data in real-time and generates reports or visualizations of sleep patterns and quality as an output.

Images