Personalized sound wave stress relief system based on stress prediction

Abstract

The present invention relates to a personalized sound wave stress relief system based on stress prediction, capable of calculating a stress index and a stress resistance index by analyzing a heart rate signal measured in a non-contact manner, predicting stress according to cumulative analysis thereof, and relieving stress through a sound source corresponding to a heart rate signal in a stable state.

Description

Stress Prediction-Based Personalized Soundwave Stress Relief System

[0001] The present invention relates to a personalized sound wave stress relief system, and more specifically, to a stress prediction-based personalized sound wave stress relief system that alleviates stress derived by analyzing heart rate signals measured by a user according to lifestyle patterns and time flow through a sound source corresponding to a heart rate signal in a resting state, and creates an environment capable of reducing stress by operating preemptively according to predictive information generated based on accumulated data.

[0002] The present invention is the result of the following research and development project.

[0003] - Project Name: Regional Specialized Industry Development+(R&D) - Regional Key Industry Development

[0004] - R&D Project Title: Customized Soundwave Solution Based on AI Stress Prediction System

[0005] - R&D Project Number: S3403431

[0006] - R&D Project Support Agency: Ministry of SMEs and Startups

[0007] Anxiety and stress are increasing due to the occurrence of various social problems in modern society, such as the rapidly changing social environment, excessive competition, dehumanization, socio-psychological crimes, the emergence of infectious diseases, and aging, posing a significant threat to mental and physical health.

[0008] If continuous stress is not properly relieved and accumulates chronically, it can lead to cardiovascular and digestive system diseases, as well as psychological symptoms such as depression, anger, agitation, emotional instability, and anxiety. In severe cases, it can result in sleep disorders, panic disorders, depression, addiction, burnout syndrome, and autonomic nervous system imbalance; therefore, interventions to alleviate stress are crucial.

[0009] To this end, various programs and intervention methods to alleviate stress are currently being studied, and research results are also being presented indicating that they are practically helpful in reducing stress and promoting calmness.

[0010] However, while the need for stress management is increasing, the mental care healthcare market remains severely underdeveloped, and the number of people receiving mental health care is extremely low due to reasons such as "issues of time and cost and social stigma."

[0011] In particular, existing methods are focused on addressing problems resulting from stress, such as sleep disorders, addiction, and violence. However, they suffer from a lack of fundamental stress data management and direct alternatives for stress solutions. Furthermore, they are limited to serving as platforms for connecting with experts, resulting in limitations such as a shortage of professionals and time.

[0012] Consequently, there is an increasing demand for non-face-to-face digital healthcare services for stress management that enable self-management anytime and anywhere.

[0013] The present invention was created in response to the above-mentioned needs, and the objective of the present invention is to provide a stress prediction-based personalized sound wave stress relief system that can reduce stress by relieving stress through a sound source corresponding to the user's resting state heart rate signal, while preemptively operating based on stress prediction information generated from analyzing and accumulating heart rate signals measured in real time through a wearable device according to lifestyle patterns and time flow.

[0014] For the above purposes, the present invention is characterized by comprising: a measurement module for measuring a subject's heart rate signal and location-time information; an analysis module having a heart rate analysis unit that collects and analyzes waveform information including the amplitude and frequency of the heart rate through the calculated heart rate signal, and a waveform information acquisition unit that acquires waveform information of a psychological stable period; a calculation module that calculates a stress index through a heart rate variability analysis index; a sound source management module having an extraction unit that extracts a feature vector of waveform information of a stable period, and a processing unit that processes a sound source having the same BPM as the heart rate of a psychological stable period so that the feature vector is reflected therein; a prediction module having a storage unit that continuously collects and stores the stress index according to a set cycle, a pattern analysis unit that classifies and analyzes the stored stress index according to location-time information to generate pattern information, and a prediction unit that generates prediction information using the stress index calculated in real time and the pattern information; and an output module that outputs a sound source processed in response to the stress index and the prediction information so that a user can hear it.

[0015] At this time, the above-mentioned computation module may include a first computation unit that calculates a physical stress index based on the analysis of the heart rate signal, a second computation unit that calculates a mental stress index based on the analysis of the heart rate signal, and a third computation unit that calculates a stress resistance index based on the analysis of the heart rate signal.

[0016] Additionally, the first operation unit uses the number of heart rate intervals (NNI), the mode, and the number of modes and the standard deviation (SDNN), and the second operation unit may use the ratio of low-frequency band and high-frequency band power (LF / HF ratio) of the heart rate variability power spectrum (HRV power spectrum) as a heart rate variability analysis indicator.

[0017] In addition, the measurement module may further include a group management module having a first averaging unit configured to process sound sources by averaging feature vectors for each user in conjunction with the sound source management module in a situation where multiple users are located within a specific space, as a wearable device.

[0018] Additionally, the group management module further includes a grouping unit that groups users within a space by reflecting their calculated stress index and location, and a second averaging unit configured to process group-specific sound sources by averaging the feature vectors of each user within the group, and the output module may be configured to be provided for each group location within the space to output adjacent group-specific sound sources.

[0019] The present invention provides various wearable devices that acquire a user's heart rate signal to calculate stress and resistance indices and enable stress monitoring, and provides a sound source that reflects the user's resting heart rate signal, thereby providing effects of improving mental and physical stress, including stress relief, improved concentration, improved exercise efficiency, and improved sleep quality.

[0020] Furthermore, active and effective stress control can be achieved through the prediction of stress based on rhythmic cycles and preemptive responses reflecting these results, utilizing the results of collecting and analyzing stress resistance indices according to cycles. Moreover, since this is based on sound sources rather than the use of medication or methods that affect brainwaves, anyone, regardless of age or gender, can safely control stress without any side effects or discomfort.

[0021] FIG. 1 is a block diagram showing the configuration and connection relationships according to an embodiment of the present invention,

[0022] Figure 2 is an example diagram showing an electrocardiogram waveform,

[0023] Figure 3 is an explanatory diagram of the QRS complex,

[0024] Figure 4 is an example of RR or NN,

[0025] Figure 5 is an example diagram of an RR Tachogram and an HR Tachogram,

[0026] Figure 6 is an example of an RR Tachogram,

[0027] Figure 7 is an example diagram of the HRV Power Spectrum.

[0028] Figure 8 is an example of an RR histogram.

[0029] Figure 9 is a graph comparing the results of the physical stress algorithm according to the present invention and "CANOPY9 RSA".

[0030] FIG. 10 is a graph comparing the physical stress algorithm according to the present invention and Baevsky's algorithm,

[0031] FIG. 11 is a graph comparing the results of the mental stress algorithm according to the present invention and "CANOPY9 RSA",

[0032] FIG. 12 is a stress index calculation screen according to an embodiment of the present invention,

[0033] FIG. 13 is a stress prediction information output screen according to an embodiment of the present invention.

[0034] The configuration of the stress prediction-based personalized sound wave stress relief system of the present invention will be described in detail below with reference to the attached drawings.

[0035] FIG. 1 is a block diagram showing the configuration and connection relationships according to an embodiment of the present invention. The present invention includes, as main components, a measurement module (110), an analysis module (120), a calculation module (130), a sound source management module (140), a prediction module (150), an output module (160), and a group management module (170).

[0036] The above measurement module (110) is for acquiring heart rate signals and location-time information from the subject's body and is equipped with a heart rate sensor unit (111) for measuring the subject's heart rate signal and a location tracking unit (112).

[0037] The heart rate sensor unit (111) above acquires a biosignal based on the heartbeat of the subject, and in the present invention, it basically measures an ECG (Electrocardiogram). Although the accuracy of the ECG is high for the analysis of heart rate waveforms used in the present invention, some papers have compared various indicators comparing the ECG and PPG and found that the largest error among the indicators is only 2.46%, so the use of PPG is also possible.

[0038] In addition, the heart rate sensor unit (111) may be configured in a form that the user can always carry and use by holding it in the hand, but may also be configured in a form that can be worn or attached to various parts of the body, such as the wrist or other parts capable of measuring an electrocardiogram, to obtain the user's electrocardiogram signal.

[0039] The above location tracking unit (112) is applied to a wearable device to obtain the user's location information over time. Depending on the usage environment, the following can be appropriately used: GNSS (Global Navigation Satellite System) based technology represented by GPS, A-GPS (Assisted GPS) which uses information from mobile communication base stations alone or in combination, Cell ID-based location tracking technology, Wi-Fi and Bluetooth-based location tracking technology specialized for indoor areas, and location tracking technology utilizing a combination of accelerometer + gyroscope + geomagnetic sensor as a complementary technology in indoor environments where GNSS signals are weak.

[0040] The above analysis module (120) is configured to analyze a heart rate signal measured through the heart rate sensor unit and includes a heart rate analysis unit (121) and a waveform information acquisition unit (122).

[0041] The heart rate analysis unit (121) is configured to collect and analyze waveform information including the amplitude and frequency of the heart rate through the calculated heart rate signal.

[0042] The waveform information acquisition unit (122) is configured to acquire waveform information of a psychological stable period. Preferably, in order to obtain an electrocardiogram signal of a stable period or a state as close as possible thereto, an electrocardiogram signal that is judged to be in the most stable state is obtained through multiple measurements over a period of time during which the stressor is removed, rather than a single measurement.

[0043] In addition, electrocardiogram signals can be continuously measured and cardiac information received in a state where stressors are absent or eliminated as much as possible, or the Balance of Autonomic Nervous System can be quantified using the standard lead method based on Heart Rate Variability (HRV) after continuous measurement for a set period using autonomic nervous system measurement equipment and utilized as psychological information. Furthermore, electrocardiogram waveform information is extracted from the electrocardiogram signal during a period of psychological stability, and the period of psychological stability can be determined by reflecting the stress index described later and identifying the point in time when stress is judged to be lowest.

[0044] The above calculation module (130) is configured to calculate a stress index based on a heart rate signal, and in the present invention, the stress index is a concept that includes a physical stress index, a mental stress index, and a stress resistance index derived by the applicant through research. To this end, the above calculation module (130) is equipped with a first calculation unit (131) that calculates a physical stress index (PSI) based on the analysis of the heart rate signal of the heart rate analysis unit (121), a second calculation unit (132) that calculates a mental stress index (MSI) based on the analysis of the heart rate signal, and a third calculation unit (133) that calculates a stress resistance index (SRI) based on the analysis of the heart rate signal.

[0045] In the present invention, the physical stress index calculated by the first calculation unit (131) uses the number of heart rate intervals (NNI), the mode, and the number of modes and the standard deviation (SDNN), and the specific algorithm is described later.

[0046] In the present invention, the second operation unit (132) and the third operation unit (133) use the ratio of the low-frequency band and high-frequency band power of the heart rate variability power spectrum (HRV power spectrum) (LF / HF ratio) as a heart rate variability analysis indicator for calculating the mental stress index, and the specific algorithm will be described later.

[0047] The above sound source management module (140) is configured to generate and manage sound sources for stress relief and includes an extraction unit (141) that extracts a feature vector of waveform information of a stabilizer and a processing unit (142) that processes the sound source having a BPM equal to the heart rate of a psychological stabilizer so as to reflect the feature vector.

[0048] That is, the extraction unit (141) extracts feature vectors for amplitude and frequency through the analysis of changes in the normal rhythm P-QRS-T waveform of the electrocardiogram in the user's stable state.

[0049] In addition, the processing unit (142) receives sound sources such as various music or nature sounds that can provide mental and physical stability in advance and processes them so that the feature vector is reflected in a sound source having a BPM equal to the heart rate of a psychologically stable period. In particular, since a sound source with a BPM equal to the heart rate of a stable period is used, it is possible to artificially adjust the speed of the sound source to have a specific BPM, but it is preferable to use a separate sound source having that BPM.

[0050] That is, the extracted sound source is processed by reflecting the sound waves generated through the extraction unit (141), and the sound wave characteristics of the sound source are processed to be identical to the sound waves extracted from the feature vector of the user's resting heart rate waveform.

[0051] The above prediction module (150) is configured to enable active and effective stress control through stress prediction based on rhythm cycles and preemptive response reflecting the results obtained from collecting and analyzing stress indices according to cycles, and is equipped with a storage unit (151), a pattern analysis unit (152), and a cycle prediction unit (153).

[0052] The above storage unit (151) is configured to continuously collect and store the stress index according to a set period. That is, storage is performed continuously over time, but the stress index collected at a set period can also be linearized to continuously connect.

[0053] The pattern analysis unit (152) classifies and analyzes the stored stress resistance index according to a set period to generate pattern information. The set period can reflect a typical lifestyle pattern and can generate pattern information based on cycles of days, weeks, months, quarters, seasons, and years, including daily work hours, work, rest, meals, and sleep times.

[0054] The above-mentioned period prediction unit (153) generates prediction information corresponding to the pattern information of the change in the stress resistance index calculated in real time.

[0055] That is, the stress change calculated in real time is compared with the pattern generated by the pattern analysis unit (152) mentioned above. When the change reflects location and time and exhibits the same or similar characteristics within a set range, the subsequent change in the stress index can be predicted through the corresponding pattern information, and this is generated as prediction information.

[0056] The output module (160) is configured to output a processed sound source so that a subject can hear it. It may be a speaker, headphones, earphones, bone conduction headphones, or various other known sound output means. However, as described in the embodiment below, when applied to a group facility, it is preferable to use a speaker that allows multiple users to hear it simultaneously.

[0057] In particular, the present invention operates the output module according to the stress index and prediction information. That is, based on the operation of outputting a processed sound source in a high-stress situation using the stress index, if the stress resistance index is predicted to decrease over time using the prediction information, a sound source with waveform characteristics identical to the electrocardiogram waveform during rest is output, thereby enabling effective stress management by proactively responding to the stress situation.

[0058] This invention is based on the Healing Beat Program developed by the applicant and applies a Beat Induction intervention program utilizing the described beats and sound waves to various patients to induce sedation in users exposed to various stressors by outputting a sound source with waveform characteristics identical to a resting electrocardiogram waveform. The effects of autonomic nervous system stabilization and stress relief through such sound sources have been proven through various research papers in which the applicant and inventor participated.

[0059] In particular, the present invention can be applied under the premise of a situation in which multiple users are located within a specific space in various workplace environments where modern people spend a lot of time and which can be a major cause of stress.

[0060] Of course, in work environments where one can work quietly alone, such as for freelancers, it is desirable to use an output module that can be used individually to process sound sources optimized for the individual. However, in work conditions where various customer services and interactions take place, such as in department stores and supermarkets, individual workers cannot use individual output modules. Therefore, in the present invention, stress can be managed by group through a group management module (170).

[0061] In other words, as mentioned earlier, in environments where a large number of workers are located within a specific space, background music (BGM) is typically output at all times to create a specific atmosphere. By processing and outputting sound sources obtained from the market as is, in accordance with the present invention, an additional effect of reducing worker stress can be achieved along with creating an atmosphere.

[0062] The above group management module (170) has a first averaging unit (171), a grouping unit (172), and a second averaging unit (173) as detailed components.

[0063] The first averaging unit (171) is configured to process sound sources by averaging the feature vectors for each user in conjunction with the sound source management module (140). The first averaging unit (171) is useful when workers performing the same or similar duties are located within a relatively small unit space, and applies the average of the waveform information of the psychological stabilizer obtained by collecting and analyzing the waveform information of heart rate signals collected from each user located in the space. That is, the feature vector of the waveform information of the stabilizer for each user is extracted, and the sound source is processed so that the average of the feature vectors is reflected in the average BPM of the heart rate of the stabilizer for all users.

[0064] In this process, it is desirable to calculate the average while excluding outliers, which are heart rate signals outside the normal range due to reasons such as specific constitution or cardiac abnormalities. Such outlier removal can be achieved by applying the Interquartile Range (IQR) method, which calculates the first quartile (Q1) and third quartile (Q3) of the data, calculates IQR = Q3 - Q1, and sets the outlier criteria to values ​​outside Q-1.5×IQR and Q3+1.5×IQR; the Mean ± nσ method, which uses the mean (μ) and standard deviation (σ) to consider values ​​outside the ranges of μ-2σ and μ+2σ as outliers; unsupervised learning algorithms that identify outliers by clustering normal and abnormal groups or detect outliers through random partitioning; or machine learning-based methods that detect outliers by comparing density with surrounding data.

[0065] If the applicable space is large or there are many included users, or if the space is separated or the tasks of the included users can also be specifically separated, they can be grouped, and the grouping unit (172) groups the users within the space by reflecting the calculated stress index and location.

[0066] For example, even within the same store, such as the aforementioned supermarket or department store, groups can be classified into those performing sales promotion and sales, those performing cashier and settlement tasks, those performing inventory and logistics management tasks, and those performing customer service and response tasks. Additionally, the level of stress may vary depending on the task. It is also possible to group based on this information. For convenience, a more effective stress reduction effect can be obtained by grouping users within a set range based on the location identified through the measurement module (110), including the stress index calculated through the calculation module (130) for each user.

[0067] The second averaging unit (173) is configured to process group-specific sound sources by averaging the feature vectors for each user within the group. That is, for users grouped into groups through the grouping unit (172), the feature vector of the waveform information of each user's stabilizer is extracted, and the average of the feature vector is processed so that it is reflected in the sound source of the average BPM of the heart rate of all users' stabilizers. In this case, the average can be calculated by excluding outliers by applying methods such as the interquartile range method, the standard deviation method, and the machine learning-based method.

[0068] In response to this group management module (170), the output module (160) is configured to be provided for each group location within the space and to output sound sources for adjacent groups, thereby effectively managing the stress of workers and visitors who stay in a specific space with BGM and perform services.

[0069] The algorithm for calculating physical and mental stress indices and stress resistance indices according to the present invention is described as follows.

[0070] The heart acts as a sophisticated pump that supplies blood to all cells, tissues, and organs of the body, performing repetitive cycles of contraction and expansion to achieve this. The phase of the heart cycle during which the ventricles contract and expel blood from the heart is called "systole," while the phase during which the ventricles relax and blood flows from the atria into the ventricles is called "diastolicity." Blood pressure measuring devices measure the systolic and diastolic blood pressure of the heart cycle; for example, if the reading is 120 / 80, 120 represents the systolic blood pressure, which indicates the pressure generated by the contraction of the left ventricle, and 80 represents the diastolic blood pressure, which signifies the pressure when blood flows in as the left ventricle relaxes.

[0071] The autonomic nervous system regulates these heartbeats; the sympathetic nervous system increases the heart rate, while the parasympathetic nervous system decreases it. In other words, at a given point in time, the heart rate is determined by the interaction between the parasympathetic nervous system (PSNS), which slows it down, and the sympathetic nervous system (SNS), which speeds it up.

[0072] The sympathetic nervous system is primarily activated during daytime activity or exercise, increasing heart rate, constricting blood vessels to raise blood pressure, and reducing the movement of the digestive organs. The parasympathetic nervous system is activated during rest or meals, smoothing the heartbeat, dilating blood vessels to promote blood flow, adjusting the mind and body to a relaxed state, and promoting the secretion of digestive juices or bowel movements.

[0073] Figure 2 is an example diagram showing an electrocardiogram waveform, and Figure 3 is an explanatory diagram of the QRS complex. An electrocardiogram is a graph in which the electrical activity of the heart is measured as voltage at a set time and recorded.

[0074] The heart acts as a small power plant; electricity generated in the sinoatrial node is transmitted to the atria, the upper part of the heart, where it flows. At this time, a small P wave is recorded on the electrocardiogram (ECG), and the atria contract to pump blood into the ventricles. When the electrical waves passing through the atria reach the ventricles, the lower part of the heart, and excite them, a large, sharp QRS wave appears on the ECG; at this moment, the ventricles contract to send blood throughout the body. Subsequently, the ventricles rest to recover their excitability in preparation for the next contraction. During this time, a broad, relatively small peak-shaped T wave appears on the ECG, and the ventricles mechanically relax. In other words, the P wave represents atrial contraction, the QRS wave represents ventricular contraction, and the T wave represents ventricular relaxation; the QRS wave is called the QRS complex.

[0075] Figure 4 is an example of RR or NN, where the time interval between adjacent PQR complexes in the ECG is denoted as 'RR'. Here, 'R' refers to the R point of the PQR complex. It is common to denote this 'RR' as 'NN'. Here, 'N' indicates 'normal', meaning a normally processed signal free from noise, bias, or abnormal fluctuations. In other words, 'NN' signifies 'normal RR'. The unit used for RR or NN is milliseconds (ms).

[0076] Figure 5 is an example of an RR Tachogram and an HR Tachogram. The change in the minute time interval between one heart cycle and the next is called heart rate variability (HRV). When the standard for heart rate variability is expressed as RR (ms) or NN (ms), it becomes an RR Tachogram (green line), and when expressed as heart rate (bpm), it becomes an HR Tachogram (red line).

[0077] Since HR(bpm) refers to the frequency during one minute (60s, 60,000ms), NN and HR are expressed by the following [Equation 1].

[0078] [Mathematical Formula 1]

[0079]

[0080] Figure 6 is an example of an RR Tachogram, where the graph in Figure 5 represents HRV over 10 seconds, and when HRV over 1 hour is shown as an example, it is displayed as in Figure 6.

[0081] A healthy heart exhibits complex and continuous non-linear trends of change to respond rapidly to sudden physical or psychological problems, whereas, conversely, the complexity of heart rate variability significantly decreases under conditions of disease or stress. However, in some cases, pathological conditions can generate variations, requiring differentiation through close interpretation of the electrocardiogram.

[0082] To increase the reliability of HRV analysis, it is recommended that ECG data be measured at 250–500 Hz. Although the number of measurements can be reduced using interpolation techniques, it is required to measure at least 100 Hz. Additionally, the measurement time can be divided into short-term (5 minutes) and long-term (24 hours), and 2.5 minutes can be applied as an attribute.

[0083] Since heart rate variability occurs through the interaction between the sympathetic and parasympathetic nervous systems, heart rate variability is believed to be associated with autonomic nervous system activity; furthermore, HRV is easy to apply clinically because it is a simple, non-invasive test that provides various information regarding the activity of autonomic nerves secreting to the heart.

[0084] Figure 7 is an example of an HRV Power Spectrum, and by applying the Fast Fourier Transform (FFT) to the RR Tachogram, the HRV Power Spectrum of the frequency domain data can be obtained from the time domain data of the RR Tachogram.

[0085] HRV Power Spectrum can be classified by frequency range as shown in [Table 1] below, and in the case of LF, long-term measurement can reflect sympathetic nervous system activity more.

[0086] [Table 1]

[0087]

[0088] Figure 8 is an example of an RR histogram. An RR histogram is a graph of a frequency table obtained by classifying the NN (=RR) variables shown in an RR Tachogram into specific intervals (classes, bins) and calculating the frequency (number) belonging to each class.

[0089] Although a bin size of 1 / 128(s) = 7.8125(ms) is recommended for HRV analysis, most experiments have shown that a bin size of about 8ms is appropriate for drawing a smooth histogram that is neither too fine nor too coarse.

[0090] When heart rate variability is high, the NM interval is large and the Y value is low, forming a low and wide triangle shape, whereas when heart rate variability is low, it forms a narrow and high triangle shape.

[0091] The autonomic nervous system functions to maintain homeostasis in response to internal and external changes; stress stimulates the autonomic nervous system, and healthier individuals exhibit a wider range of heart rate variability, preparing them to respond to various stimuli. Consequently, individuals more vulnerable to stress exhibit reduced heart rate variability. Furthermore, stress increases the activity of the sympathetic nervous system or causes a decrease in the activity of the parasympathetic nervous system.

[0092] As previously discussed, heart rate variability arises from the interaction between the sympathetic and parasympathetic nervous systems on the sinoatrial node; therefore, the intervals between heartbeats continuously change depending on the influence of the autonomic nervous system on the sinoatrial node, which varies according to internal and external environments. In other words, even when the body maintains homeostasis, the heart rate changes every moment in a complex and non-linear manner. By continuously observing heart rate fluctuations over time, periodic changes can be identified, and analyzing this allows for the prediction of the state of the autonomic nervous system.

[0093] First, through time domain analysis, the time interval between normal QRS complexes or the instantaneous heart rate at a specific point in time is measured to determine the time interval between consecutive normal QRS complexes (NN interval, normal to normal interval, NN interval), and the indicators of the following [Table 2] and [Equation 2] can be calculated statistically.

[0094] [Table 2]

[0095]

[0096] [Mathematical Formula 2]

[0097]

[0098] SDNN is generally calculated from 24-hour measurements, and the shorter the electrocardiogram measurement time, the smaller this value becomes. In 5-minute measurements, it is referred to as the SDNN index. In this invention, we will primarily explain the interpretation of 5-minute analysis results, which are widely used in clinical practice, and the SDNN presented in this invention may be understood as identical to the SDNN index. Furthermore, since the SDNN value increases as heart rate variability becomes greater and more irregular, the SDNN value increases with physiological health. Therefore, SDNN can be considered an indicator reflecting physiological resilience to stress. In other words, a decrease in the SDNN value reflects a decline in the ability to cope with stress.

[0099] RMSSD reflects short-term variation in continuous heart rates and is primarily used to predict heart rate variability in the high-frequency range, mainly reflecting parasympathetic nervous system activity. Generally, a higher RMSSD value is interpreted as indicating a physiologically healthy and relaxed state. While time-domain analysis indicators show overall heart rate variability and are easy to calculate arithmetically, they cannot distinguish between sympathetic and parasympathetic nervous system activity or quantify the balance of the autonomic nervous system.

[0100] Next is frequency domain analysis, which allows a complex heart rate variability signal to be separated into frequency signals of different bands (Fourier Transformation) and evaluates the relative strength (power, variance) of each frequency signal band.

[0101] Using frequency domain analysis in HRV analysis allows for the separation of the activity levels of the sympathetic and parasympathetic nervous systems.

[0102] While time-domain analysis metrics such as SDNN and RMSSD can be effectively utilized in short-term measurements, results obtained from frequency-domain analysis offer better explanatory power for describing the physiological significance of heart rate variability.

[0103] The heart rate variability power spectrum provides information on the intensity of high frequency (HF; 0.15–0.4 Hz), low frequency (LF; 0.04–0.15 Hz), and very low frequency (VLF; 0.003–0.04 Hz) bands, which reflect the state of the autonomic nervous system. These three major spectral components are primarily measured through short-term measurements between 2 and 5 minutes. The indices derived from the frequency domain analysis are as shown in the following [Table 3].

[0104] [Table 3]

[0105]

[0106] VLF is a component of heart rate variability with a very long period. Therefore, since it is difficult to accept VLF indicator values ​​obtained from short-term measurements of 5 minutes or less as they are, it is advisable to exclude the VLF component when interpreting power spectrum analysis indicator values ​​of short-term electrocardiograms.

[0107] The physiological significance of the VLF band is generally known to reflect sympathetic nervous system activity.

[0108] The prevailing view is that the LF band primarily reflects the activity of the sympathetic nervous system. Although LF is known to reflect the activity of both the sympathetic and parasympathetic nervous systems, analysis of heart rate measurements over a long period indicates that it reflects sympathetic nervous system activity more strongly. Therefore, it should be understood that LF reflects both sympathetic and parasympathetic nervous system activity simultaneously in short-term measurements, while reflecting the sympathetic nervous system more strongly in long-term measurements. Additionally, LF is associated with mental stress.

[0109] The HF band shows rapid changes in inter-beat variation, which is due to parasympathetic nervous system stimulation, and HF is a representative measure of parasympathetic nervous system activity.

[0110] TP corresponds to total heart rate variability and can be considered an indicator reflecting the overall regulatory capacity of the autonomic nervous system. In time-domain analysis, it has a similar significance to SDNN. In cases of chronic stress or physical illness, TP decreases due to a decline in the regulatory capacity of the autonomic nervous system.

[0111] LF / HF reflects the overall balance of the autonomic nervous system. A high value indicates that the sympathetic nervous system is relatively overactive or that the activity of the parasympathetic nervous system is suppressed. This value increases when sympathetic nervous system activity is relatively dominant, such as in states of anxiety, fear, tension, distractibility, and hyperarousal. Conversely, this value decreases when parasympathetic nervous system activity is dominant, such as in states of lethargy, neurasthenia, depression, and low arousal.

[0112] LF_norm and HF_norm are used, along with LF / HF, to reflect the balance between the sympathetic and parasympathetic nervous systems. Their meaning can be accurately understood only when expressed with their absolute values.

[0113] In the present invention, the physical stress index (PSI) is an indicator that evaluates the activity of the autonomic nervous system controlling heart rhythm, and is expressed by defining physical stress as changes in the body's central circuit that narrow the range of RR variation. It is an index that measures physiological changes occurring in response to daily or prolonged physical activity, particularly in response to stress, and may include heart rate, blood pressure, respiratory rate, hormone levels, etc. In the present invention, the PSI is calculated using the following [Equation 3] using indicators of time domain analysis.

[0114] [Mathematical Formula 3]

[0115]

[0116] Figure 9 is a graph comparing the results of the physical stress algorithm according to the present invention and "CANOPY9 RSA", and it can be seen that the results of comparing the algorithm of the present invention with the results of "CANOPY9 RSA" for PSI are completely consistent.

[0117] In the present invention, physical stress is evaluated according to the PSI value as shown in [Table 4].

[0118] [Table 4]

[0119]

[0120] Baevsky's algorithm, as shown in the following [Equation 4], is widely used to calculate the stress index and was compared with the algorithm according to the present invention.

[0121] [Mathematical Formula 4]

[0122]

[0123] FIG. 10 is a graph comparing the physical stress algorithm according to the present invention with Baevsky's algorithm, wherein SI B and the PSI of the algorithm according to the present invention RThe graph showing the relationship indicates a very high correlation with a correlation coefficient of r=0.98.

[0124] Next, the mental stress index (MSI) is an indicator that evaluates the degree of imbalance between the sympathetic and parasympathetic nervous systems; if sympathetic activity is relatively dominant compared to the parasympathetic nervous system, mental stress appears high, and it is also referred to as the "autonomic nervous system balance index."

[0125] Factors related to mental burden, emotional anxiety, depression, etc., can be measured. Mental stress primarily occurs in cognitive, emotional, and social aspects. MSI is calculated using the following [Equation 5] with the indicators of frequency domain analysis.

[0126] [Mathematical Formula 5]

[0127]

[0128] Figure 11 is a graph comparing the results of the mental stress algorithm according to the present invention and "CANOPY9 RSA", and it can be seen that the results of comparing the algorithm of the present invention and the results of "CANOPY9 RSA" for MSI are completely consistent.

[0129] In the present invention, mental stress is evaluated according to the MSI value as shown in the following [Table 5].

[0130] [Table 5]

[0131]

[0132] Finally, the stress resistance index (SRI) is an index that indicates how well an individual copes with and overcomes stressful situations, and it can evaluate the ability to respond to life's difficulties, resilience, and problem-solving skills. The SRI is calculated using the following [Equation 6] with indicators from frequency domain analysis.

[0133] [Mathematical Formula 6]

[0134]

[0135] In the present invention, mental stress is evaluated according to the MSI value as shown in the following [Table 6].

[0136] [Table 6]

[0137]

[0138] FIG. 12 is a stress index calculation screen according to an embodiment of the present invention, in which the stress index by date is displayed in a color range from low to dangerous along with numeric points, and an option is provided to provide a stress index by period such as 1 day, 1 week, 1 month, and 1 year.

[0139] In addition, information such as Avg BPM (Average Heart Rate), Fatigue, and SRI (Stress Resistance Index) is displayed. An Auto Measure button is provided to check the stress level. Subsequently, the Stress Rec. Music function outputs audio processed to correspond to the stress level, allowing the user to listen to it.

[0140] FIG. 13 is a stress prediction information output screen according to an embodiment of the present invention, which monitors the current heart rate and provides a function to predict changes in the user's stress through Today's Stress Prediction. For example, if the user's stress level is predicted to peak at 4:09 PM, a message recommending that the user control stress using the "Healing Beat" function is displayed 15 minutes before (3:54 PM). Based on the user's heart rate and other biometric data, it predicts when stress will reach its peak and can perform preventive measures if stress is expected to increase at a specific time.

Claims

1. A measurement module that measures the subject's heart rate signal and location-time information; An analysis module comprising a heart rate analysis unit that collects and analyzes waveform information including the amplitude and frequency of the heart rate through a calculated heart rate signal, and a waveform information acquisition unit that acquires waveform information of a psychological stable period; A computation module that calculates a stress index using heart rate variability analysis indicators; A sound source management module comprising an extraction unit for extracting a feature vector of waveform information of a stabilizer, and a processing unit for processing the sound source having a BPM equal to the heart rate of a psychological stabilizer so as to reflect the feature vector; A prediction module comprising: a storage unit that continuously collects and stores the stress index according to a set period; a pattern analysis unit that classifies and analyzes the stored stress index according to location-time information to generate pattern information; and a prediction unit that generates prediction information using the stress index calculated in real time and the pattern information. A smart stress prediction and alleviation system characterized by having an output module that outputs a processed sound source corresponding to a stress index and the prediction information so that a user can hear it.

2. In Paragraph 1, The above computational module is, A smart stress prediction and alleviation system characterized by comprising a first calculation unit that calculates a physical stress index based on the analysis of the heart rate signal, a second calculation unit that calculates a mental stress index based on the analysis of the heart rate signal, and a third calculation unit that calculates a stress resistance index based on the analysis of the heart rate signal.

3. In Paragraph 2, The above first operation unit uses the number of heart rate intervals, the mode, the number of modes, and the standard deviation, and The smart stress prediction and alleviation system is characterized by the fact that the second calculation unit above uses the ratio of the low-frequency band and high-frequency band power of the heart rate variability power spectrum as a heart rate variability analysis indicator.

4. In Paragraph 1, The above measurement module is a wearable device, A smart stress prediction and alleviation system characterized by further including a group management module having a first averaging unit configured to process sound sources by averaging feature vectors for each user in conjunction with the sound source management module in a situation where multiple users are located within a specific space.

5. In Paragraph 4, The above group management module is, It further includes a grouping unit that groups users within a space by reflecting the calculated stress index and location, and a second averaging unit configured to process group-specific sound sources by averaging the feature vector for each user within the group. A smart stress prediction and alleviation system characterized by the above-mentioned output module being configured to be provided for each group location within the space and to output sound sources for adjacent groups.

Images