KR20220124457A - Smart system for stress measure and management - Google Patents

Abstract

The present invention provides a smart system for stress measurement and management, which may collect stress data, extract patterns which analyzed the stress data, generate a feedback model for a control sequence generation model of a smart device learned using the extracted data through user and sensor data collection, obtain the control sequence generation model by additional learning for individual users based on the feedback model, and detect and manage the adolescent stress early.

Description

Stress measurement and management smart system {Smart system for stress measure and management}

The present invention relates to a stress measurement and management smart system, and to a youth customized stress measurement and management smart system.

Korean adolescents (13-24 years old) are exposed to daily stress due to excessive competition for entrance exams and worries about an uncertain future. It is known that adolescent stress not only increases depression, anxiety, delinquency, addiction, and suicidal thoughts, but also lowers school adaptation, and has a longitudinal effect on physical and mental health in later adulthood. essential for later development.

Accordingly, attempts have been made in academia and industry to measure and manage stress, but stress measurement and management programs that reflect neurophysiological mechanisms by focusing on adolescent characteristics and brain developmental characteristics are insufficient. In addition, self-report questionnaires or measurements in experimental situations do not reflect the actual stress change pattern in reality, and reliability or validity is somewhat lacking.

On the other hand, stress measurement technology based on biosignals can monitor the stress level as an objective indicator that goes beyond subjective perception. In particular, stress measurement through machine learning can 1) efficiently quantitatively model human responses to various stresses, and 2) empirically predict changes in individual stress based on these data. 3) Through this, it is possible to diagnose stress at an early stage before it worsens into a chronic disease and to make scientific predictions based on the evidence-based change pattern. Therefore, it is necessary to develop a stress smart device and system tailored to adolescents that can detect and manage adolescent stress at an early stage by applying an artificial intelligence machine learning algorithm.

Korean Patent Publication No. 10-2019-0092691 (published on August 8, 2019)

It is an object of the present invention to provide a smart system for measuring and managing youth customized stress that can detect and manage youth stress early.

To achieve the above object, the stress measurement and management smart system according to an embodiment of the present invention extracts the stress data collection and analysis pattern, and uses the extracted data to learn about the control sequence generation model of the smart device. It is obtained by generating a feedback model through user and sensor data collection, and further learning the control sequence generation model for each user based on the feedback model.

The stress measurement and management smart system may generate stress coping, management, and stress scoring results according to time series data accumulation, and may warn the user of chronic stress and related risks according to the generated stress scoring results.

Therefore, the stress measurement and management smart system according to an embodiment of the present invention helps adolescents who are accustomed to using smart devices to measure and manage stress on a daily basis. I can prevent suicide and improve the quality of life.

1 shows a method of constructing a stress measurement and management smart system.
2 shows an example of a smart device capable of measuring stress.
3 shows an example of time series data prediction using a sequence generation model.
4 and 5 show an example of an artificial neural network for processing an input signal and an artificial neural network for generating a result, respectively.
6 shows an example of sequence calibration through a feedback signal.

In order to fully understand the present invention, the operational advantages of the present invention, and the objects achieved by the practice of the present invention, reference should be made to the accompanying drawings illustrating preferred embodiments of the present invention and the contents described in the accompanying drawings.

Hereinafter, the present invention will be described in detail by describing preferred embodiments of the present invention with reference to the accompanying drawings. However, the present invention may be embodied in several different forms, and is not limited to the described embodiments. In addition, in order to clearly explain the present invention, parts irrelevant to the description are omitted, and the same reference numerals in the drawings indicate the same members.

Throughout the specification, when a part "includes" a certain component, it does not exclude other components unless otherwise stated, meaning that other components may be further included. In addition, terms such as "... unit", "... group", "module", and "block" described in the specification mean a unit that processes at least one function or operation, which is hardware, software, or hardware. and a combination of software.

In general, stress management is based on a professional model in which a client visits a center prepared in a psychotherapist, counselor, or medical clinic setting and manages it in the form of a social service through one-on-one meeting with an expert. However, this research team intends to utilize smart systems and smart devices to detect and manage stress in daily life. To this end, it is necessary to include time and spatial freedom in stress detection and management, and above all, to provide personalized judgment of stress situations specialized for the user's condition and feedback according to the situation sequentially. It consists of four parts: 1) a smart device equipped with a sensor that can be worn in daily life, 2) developing and mounting a stress measurement model, 3) generating a recommendation for stress management and coping behavior, and 4) creating an individually customized sequence based on feedback. Including, the system is built with the following process. This system aims to provide a customized stress management solution through the provision of stress measurement and stress level alarm using smart devices, including the above functions, and feedback on stress coping using learning data.

1 shows the process of developing a stress measurement and management smart system.

Referring to FIG. 1 , the stress measurement and management smart system according to the present embodiment may be developed through the following process.

1. A smart device equipped with a sensor that can be worn in daily life

2 shows an example of a smart device capable of measuring stress.

The smart device shown in FIG. 2 can be viewed as including the configurations 1) and 2).

1) Smart device with stress sensor: Includes sensors that measure heart rate variability, surface temperature, and pulse, which can be measured through the wearer's skin surface contact with the smart device.

2) Signal processing system: includes a part that collects and processes real-time sensor information obtained from the system, a part that generates a feedback signal according to a specified model, a part that records previous stress information, and a communication part with an external system.

2. Development and implementation of stress measurement model

Sample data on stress is collected through the smart system selected above. Standard data is established by analyzing the stress response pattern according to the individual's stressful/non-stressed situation based on previously known information on changes in heart rate or changes in biological responses using real sensors. Through the data, it learns and models the biomarker change pattern according to the stress response range and intensity for each stress situation.

3. Create Stress Management and Coping Behavior Recommendations

1) Stress Coping Reminder: The stress measurement model proceeds with initial settings according to the user's condition, records past stress information, and induces and guides stress coping behavior through notification of appropriate coping behavior when stress exceeds a certain level. . The purpose of this system is to provide a stress detection alarm. It uses the sensed information to prevent psychological pain or chronic stress accumulation outside of the normal range, and at the same time extracts a pattern that deviates from the normal value of the biosignal to alert the user to the problem situation. can be directly connected with

2) Individual optimization: Acquire the individual's initial input information, initial stress measurement data, and sensing information to determine the individual's condition. Based on this, it includes a function to manage the stress cycle and intensity through learning and to recommend appropriate stress coping behavior. It has the advantage of being able to recommend the type and composition of coping behavior with high satisfaction according to the user's pattern based on the actual collected information.

3 shows an example of time series data prediction using a sequence generation model, and shows an example of time series data prediction using ARIMA.

Time series data prediction has been studied in various fields such as stock price prediction and weather forecasting, and various traditional studies such as traditional feature extraction, dynamic regression, stratification and randomization exist based on attempts such as regression models or signal decomposition. In particular, research on model-based guessing models such as Auto-regressive Integrated Moving Average (ARIMA), ARIMA, and Seasonal ARIMA is widely known. Also, along with ARIMA, ETS (Error, Trend, Seasonal, Exponential Smoothing) is famous.

This was later studied as data analysis and state transition shape prediction, and developed into Finite State Machine (FSM) and Hidden Markov Model (HMM) that advanced it.

That is, as shown in FIG. 3 , a study was conducted to construct an individually customized stress sequence generation model for each user so that the user's future stress state could be predicted based on the data obtained for each individual user.

Such a sequence generation model is well known because it is mainly applied to weather prediction, and recently it can be implemented as an artificial neural network by combining with deep learning technology.

4 and 5 show an example of an artificial neural network for processing an input signal and an artificial neural network for generating a result, respectively.

As shown in FIG. 4 , the artificial neural network for input signal processing may be configured as a combination of a CNN stage for input signal processing and a recurrent neural network (RNN) for continuous time series data processing. And, as shown in Fig. 5, the artificial neural network for generating a result can be implemented as an LSTM-attention model with a technique such as generating a result through attention like a recent transformer network.

4. Create individual custom sequences based on feedback

1) Optimal sequence generation: In the above process, each process utilizes the user's feedback and sensing information to identify each individual's status, and through customized fine-tuning of the model, a customizable sequence generation specialized model can be configured separately. have.

6 shows an example of sequence calibration through a feedback signal.

Simultaneous Localization And Mapping (SLAM) is a representative study for correcting sequences by using actual engineering sensor information in time series. This is a technique frequently used for navigation by probabilistically updating its location based on previously recorded landmark information based on the Kalman filter. FIG. 6 shows an example of landmark observation and location correction of EFK SLAM.

2) Collection and reflection of various information: The individual operation sequence obtained in the above process can be corrected by sensing information of the smart system and user feedback. This system intends to additionally utilize the following information for personal optimization of sequence generation.

- Subjective assessment = self-assessment of stress level and stress management feedback

- Smart device sensor value = stress response biomarker measured through the contact surface

As a result, the stress measurement and management smart system according to the present invention is configured through the following steps 1 to 5, and can notify the user as shown in 6 and 7.

1. Stress measurement index selection and measurement system design

2. Stress data collection and pattern extraction by analyzing it

3. Learning a control sequence generation model of a smart device using the data

4. Create feedback model through user and sensor data collection

5. Further learning progress for individual users

6. Create stress coping, management and stress scoring results according to time series data accumulation

7. Stress Chronic and Associated Risk Warnings

The method according to the present invention may be implemented as a computer program stored in a medium for execution by a computer. Here, the computer-readable medium may be any available medium that can be accessed by a computer, and may include all computer storage media. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, and read dedicated memory), RAM (Random Access Memory), CD (Compact Disk)-ROM, DVD (Digital Video Disk)-ROM, magnetic tape, floppy disk, optical data storage, and the like.

Although the present invention has been described with reference to the embodiment shown in the drawings, which is only exemplary, those skilled in the art will understand that various modifications and equivalent other embodiments are possible therefrom.

Accordingly, the true technical protection scope of the present invention should be defined by the technical spirit of the appended claims.

Claims (2)

Stress data collection and pattern analysis are extracted, and a feedback model for a control sequence generation model of a smart device learned using the extracted data is generated through user and sensor data collection, and the control is performed based on the feedback model. Stress measurement and management smart system obtained by further training the sequence generation model for individual users. The method of claim 1, wherein the stress measurement and management smart system is
A stress measurement and management smart system that generates stress coping, management, and stress scoring results according to the accumulation of time series data, and warns users of chronic stress and related risks according to the generated stress scoring results.

Images