Sleep management system based on CBT-i therapy

Abstract

The invention discloses a sleep management system based on CBT-i therapy, and belongs to the technical field of sleep quality management and improvement, and the system comprises a sleep restriction module which comprises a data acquisition unit, a data analysis unit and an intelligent suggestion unit; the stimulation control module comprises an intelligent event monitoring unit and a personalized reminding and suggestion unit; the relaxation training module is used for selecting a proper relaxation training item for the user and determining the most suitable relaxation training type through a recommendation algorithm according to the preference of the user and historical feedback data; the cognitive therapy module comprises a sleep quality analyzer and a sleep education content recommendation system; by integrating various intelligent algorithms and models, an all-around sleep management solution is provided for the user, the sleep quality is improved, the manual intervention requirement is reduced, and the method has the advantages of being easy and convenient to operate, low in cost and high in adaptability.

Description

Technical Field

[0001] The present invention belongs to the technical field of sleep quality management and improvement, and in particular relates to a sleep management system based on CBT-i therapy. Background Art

[0002] Sleep problems are becoming increasingly common, requiring effective management and improvement methods. Traditional sleep therapy usually requires professional equipment and personnel supervision, lacking flexibility and personalization. Therefore, it is particularly important to develop an intelligent and personalized sleep management system. Summary of the invention

[0003] The purpose of the present invention is to propose a sleep management system based on CBT-i therapy in order to solve the above-mentioned problems.

[0004] In order to achieve the above object, the present invention adopts the following technical solution: a sleep management system based on CBT-i therapy, comprising:

[0005] A sleep restriction module, which includes a data collection unit, a data analysis unit and an intelligent suggestion unit. The data collection unit uses a non-contact sleep belt device to record the user's sleep data. The data analysis unit system analyzes the collected data, calculates the actual sleep duration and the time after going to bed when not sleeping, and asks the user about the activities when not sleeping. The intelligent suggestion unit gives personalized sleep restriction suggestions based on the large language model CBT-i TherapyChat according to the results of the data analysis unit.

[0006] A stimulation control module, which includes an intelligent event monitoring unit and a personalized reminder and suggestion unit, wherein the intelligent event monitoring unit sets a state machine model called RestCycleTracker for monitoring and managing the user's sleep state, and the personalized reminder and suggestion unit provides personalized reminders and suggestions for the user;

[0007] The relaxation training module selects appropriate relaxation training programs for users and determines the most suitable relaxation training type through a recommendation algorithm based on user preferences and historical feedback data;

[0008] The cognitive therapy module includes a sleep quality analyzer and a sleep education content recommendation system. The sleep quality analyzer collects data on sleep quality, and the sleep education content recommendation system combines collaborative filtering and rule-based recommendation technology to push personalized education content and coping strategies to users.

[0009] As a further description of the above technical solution:

[0010] CBT-iTherapyChat includes a real-time interaction unit, a customized intervention unit, a personalized treatment session unit, a role-playing and scenario simulation unit, and a continuous learning and adaptation unit. The real-time interaction unit uses natural language processing to allow CBT-iTherapyChat to have meaningful conversations with users and extract relevant details that may affect sleep quality, such as sleep habits, environmental factors, and psychological stress. Based on the collected information, CBT-iTherapyChat provides targeted intervention measures. The personalized treatment session unit CBT-iTherapyChat provides tailored treatment sessions specifically for users' specific sleep problems. In order to strengthen CBT-i skills, CBT-iTherapyChat can simulate different scenarios and perform role-playing with users in the role-playing and scenario simulation unit. With the help of the constantly updated RAG knowledge base, CBT-iTherapyChat is always at the forefront of sleep management technology and incorporates the latest sleep research results in a timely manner.

[0011] As a further description of the above technical solution:

[0012] The relaxation training module includes a motion capture and analysis system and a real-time feedback generator. The OpenPose skeleton detection model in the motion capture and analysis system is used to capture the posture and key points of the user when performing relaxation training in real time. The ST-GCN spatiotemporal convolutional network in the motion capture and analysis system receives key point data from OpenPose and analyzes the dynamic changes of key points over time. The real-time feedback generator combines the analysis results of OpenPose and ST-GCN, and this component generates real-time guiding suggestions.

[0013] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0014] In the present invention, by integrating multiple intelligent algorithms and models, a comprehensive sleep management solution is provided to users, sleep quality is improved, and the need for manual intervention is reduced. The solution is easy to operate, low cost, and highly adaptable. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A control block diagram of a sleep management system based on CBT-i therapy. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0017] See also Figure 1 The present invention provides a technical solution: a sleep management system based on CBT-i therapy, comprising:

[0018] The sleep restriction module includes a data collection unit, a data analysis unit and an intelligent suggestion unit. The data collection unit uses a non-contact sleep belt device to record the user's sleep data, including the time of going to bed, the time of getting up and the duration of sleep, etc. The data analysis unit systematically analyzes the collected data, calculates the actual sleep duration and the time of not falling asleep after going to bed, and asks the user about the activities when not falling asleep. According to the results of the data analysis unit, the intelligent suggestion unit system gives personalized sleep restriction suggestions based on the large language model CBT-iTherapyChat. CBT-iTherapyChat is an advanced conversational artificial intelligence model designed for providing personalized cognitive behavioral therapy (CBT-i) for the treatment of insomnia. The model is based on the Qwen2-72B-I nstruct architecture and is fine-tuned by the LoRA (low-rank adaptation) method to enable it to specifically adapt to the requirements of CBT-i therapy, thereby performing well in understanding and responding to various sleep-related problems;

[0019] The stimulation control module includes an intelligent event monitoring unit and a personalized reminder and suggestion unit. The intelligent event monitoring unit sets a state machine model called RestCycleTracker to monitor and manage the user's sleep state, such as trying to fall asleep, falling asleep, waking up, etc. The state machine model can accurately track the state transition of the user from going to bed to trying to fall asleep, ensuring that the system can promptly identify the situation where the user fails to fall asleep on time. The personalized reminder and suggestion unit provides personalized reminders and suggestions for the user. A stimulation control system called StimulusControl Mode l has been developed. When the user fails to fall asleep within 20 minutes, the StimulusControl Mode l system will start. The system provides personalized suggestions based on the following two algorithms:

[0020] Collaborative filtering algorithm: By analyzing the behavioral data of similar user groups, it finds and recommends sleep-aiding activities that have helped other users fall asleep successfully. This method can use the experience of user groups to recommend effective sleep-aiding strategies.

[0021] Content filtering algorithm: Analyzes the user's historical preferences and feedback data to recommend highly personalized sleep-aiding activities, such as reading or listening to music. The content filtering algorithm makes recommendations that are most suitable for the user based on the user's past responses to specific types (such as preferred music types or reading materials).

[0022] By combining collaborative filtering and content filtering, the StimulusControl Mode system can provide personalized stimulus control suggestions, such as getting up and reading, listening to light music, etc.

[0023] The relaxation training module selects appropriate relaxation training items for users. Based on the user's preferences and historical feedback data, the most suitable relaxation training type is determined through a recommendation algorithm. After selection, the system automatically plays the corresponding guidance video and helps users perform correct relaxation techniques through the Motion Capture and Analysis System.

[0024] A cognitive therapy module, which includes a sleep quality analyzer and a sleep education content recommendation system. The sleep quality analyzer collects data on sleep quality. The user first completes a PSQI (Pittsburgh Sleep Quality Index) questionnaire. The sleep quality analyzer uses a logistic regression model to analyze the user's PSQI score to determine the severity of the sleep problem. The logistic regression model is trained based on a historical data set and can predict the specific type of sleep disorder that the user may have and classify the user as mild, moderate or severe sleep disorder. The sleep education content recommendation system combines collaborative filtering and rule-based recommendation technology to push personalized education content and coping strategies to the user;

[0025] Collaborative filtering algorithm: By analyzing the feedback of similar users and the interactive history of educational content, it finds and recommends educational content that is effective for similar problems. This allows the recommendation system to learn from the successful experiences of other users and provide strategies that help improve sleep.

[0026] Rule-based recommendation system: Based on the user's sleep quality index and the classification results of the logistic regression model, the system pushes relevant educational materials according to predefined rules. For example, for users with severe insomnia, in-depth cognitive behavioral therapy tutorials and more frequent interactions are recommended; for mild problems, general sleep hygiene education and light relaxation exercises are recommended.

[0027] CBT-iTherapyChat includes a real-time interaction unit, a customized intervention unit, a personalized treatment session unit, a role-playing and scenario simulation unit, and a continuous learning and adaptation unit. The real-time interaction unit uses natural language processing to allow CBT-iTherapyChat to have meaningful conversations with users and extract relevant details that may affect sleep quality, such as sleep habits, environmental factors, and psychological stress. Based on the collected information, CBT-iTherapyChat provides targeted intervention measures, such as adjusting sleep restriction parameters or suggesting changes in sleep hygiene habits. The model's recommendations are dynamically adjusted according to the user's specific needs to ensure a high degree of personalization of sleep management methods. The personalized treatment session unit CBT-iTherapyChat provides tailored treatment sessions specifically targeting the user's specific sleep problems. By leveraging the rich RAG knowledge base, the model is able to create customized treatment plans that address the root causes of insomnia, such as anxiety or poor sleep hygiene habits. These programs include a combination of cognitive restructuring exercises, relaxation techniques, and behavioral changes. The role-playing and scenario simulation unit can simulate different scenarios and role-play with users to strengthen CBT-i skills. For example, the model may simulate a stressful situation and guide users to challenge and modify unhealthy thinking patterns through cognitive restructuring. This interactive training helps users practice and internalize CBT-i strategies. The continuous learning and adaptation unit uses the constantly updated RAG knowledge base to keep CBT-iTherapyChat at the forefront of sleep management technology and incorporate the latest sleep research results in a timely manner, which ensures that interventions and treatment sessions remain relevant and effective as new insights and technologies emerge.

[0028] The relaxation training module includes a motion capture and analysis system and a real-time feedback generator. The OpenPose skeleton detection model in the motion capture and analysis system is used to capture the posture and key points of the user when performing relaxation training in real time. OpenPose can identify multiple key points of the human body, including arms, legs, torso, etc., which is crucial for analyzing the correctness of the user's movements. The ST-GCN spatiotemporal convolutional network in the motion capture and analysis system receives key point data from OpenPose and analyzes the dynamic changes of key points over time. ST-GCN can identify inaccuracies in user movements by learning a large amount of standard movement data. The real-time feedback generator combines the analysis results of OpenPose and ST-GCN. This component generates real-time guiding suggestions. If it is detected that a certain movement of the user does not meet the standard, the system will immediately provide specific correction suggestions, such as "adjust the breathing rhythm" or "relax the shoulder muscles". This instant feedback helps users correct their movements and ensure the effectiveness of relaxation training.

[0029] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A sleep management system based on CBT-i therapy, characterized by: include: A sleep restriction module, which includes a data collection unit, a data analysis unit and an intelligent suggestion unit. The data collection unit uses a non-contact sleep belt device to record the user's sleep data. The data analysis unit system analyzes the collected data, calculates the actual sleep duration and the time after going to bed when not sleeping, and asks the user about the activities when not sleeping. The intelligent suggestion unit gives personalized sleep restriction suggestions based on the large language model CBT-iTherapyChat according to the results of the data analysis unit. The stimulation control module includes an intelligent event monitoring unit and a personalized reminder and suggestion unit. The intelligent event monitoring unit sets a state machine model called RestCycleTracker to monitor and manage the user's sleep state. The personalized reminder and suggestion unit provides personalized reminders and suggestions for the user. The relaxation training module selects appropriate relaxation training programs for users and determines the most suitable relaxation training type through a recommendation algorithm based on user preferences and historical feedback data; The cognitive therapy module includes a sleep quality analyzer and a sleep education content recommendation system. The sleep quality analyzer collects data on sleep quality, and the sleep education content recommendation system combines collaborative filtering and rule-based recommendation technology to push personalized education content and coping strategies to users.

2. A sleep management system based on CBT-i therapy according to claim 1, characterized in that: CBT-iTherapyChat includes a real-time interaction unit, a customized intervention unit, a personalized treatment session unit, a role-playing and scenario simulation unit, and a continuous learning and adaptation unit. The real-time interaction unit uses natural language processing to allow CBT-iTherapyChat to have meaningful conversations with users and extract relevant details that may affect sleep quality, such as sleep habits, environmental factors, and psychological stress. Based on the collected information, CBT-iTherapyChat provides targeted intervention measures. The personalized treatment session unit CBT-iTherapyChat provides tailored treatment sessions specifically for users' specific sleep problems. In order to strengthen CBT-i skills, CBT-iTherapyChat can simulate different scenarios and perform role-playing with users in the role-playing and scenario simulation unit. With the help of the constantly updated RAG knowledge base, CBT-iTherapyChat is always at the forefront of sleep management technology and incorporates the latest sleep research results in a timely manner.

3. A sleep management system based on CBT-i therapy according to claim 1, characterized in that: The relaxation training module includes a motion capture and analysis system and a real-time feedback generator. The OpenPose skeleton detection model in the motion capture and analysis system is used to capture the posture and key points of the user when performing relaxation training in real time. The ST-GCN spatiotemporal convolutional network in the motion capture and analysis system receives key point data from OpenPose and analyzes the dynamic changes of key points over time. The real-time feedback generator combines the analysis results of OpenPose and ST-GCN, and this component generates real-time guiding suggestions.

Images