KR102552787B1 - Method for snoring analysis service providing snoring analysis and disease diagnosis prediction service based on snoring sound analysis - Google Patents

Abstract

A snoring analysis service system according to the present invention includes a snoring analysis module that analyzes a snoring sound and detects a snoring state and a snoring position; a posture-snoring relationship modeling unit that detects a sleeping posture, matches it with snoring state information, and displays the snoring position on a human body modeling picture; Using the snoring monitoring information and snoring location information of the snoring analysis module and the correlation information between snoring and posture of the posture and snoring relationship modeling unit, snoring evaluation and sleep evaluation are performed by a predetermined evaluation algorithm. A disease prediction module that diagnoses and predicts a disease that can cause snoring by analyzing the snoring monitoring and location information, the evaluation information, and the cumulatively managed information change trend, while performing the cumulative storage and management in the database. ; includes

Description

Method for snoring analysis service providing snoring analysis and disease diagnosis prediction service based on snoring sound analysis}

The present invention relates to a snoring analysis service method based on snoring sound analysis, and more particularly, by analyzing snoring sound, estimating and detecting a location where snoring-sleep apnea (hereinafter referred to as “snoring”) occurs, and detecting it The present invention relates to a snoring analysis service method for providing a snoring analysis based on bio-signal analysis and a disease diagnosis prediction service that analyzes and predicts the diagnosis of a disease that causes snoring and provides a snoring analysis service.

Snoring is a respiratory noise caused by vibration of body tissues such as the soft palate due to narrowing of the upper airway for various reasons during sleep. Snoring reduces the amount of coral intake during breathing, which can lead to symptoms such as daytime fatigue, headache, daytime sleepiness, and personality changes (aggressive personality, irritability, anxiety, depressive reaction, etc.). It can cause Obstructive Sleep Apnea Syndrome (OSAS).

As such, snoring can cause abnormalities in the body's endocrine system. However, it is impossible for sleepers to self-determine their own sleep disorders. A test for diagnosing sleep disorders in a specialized medical institution causes inconvenience such as attaching measurement equipment such as electrocardiography and electroencephalography to various places and having to sleep in an unfamiliar place, and is also expensive.

Therefore, it is necessary to find out whether snoring occurs during sleep in an unrestrained and insensitive way. According to this need, Korean Patent Laid-Open Publication No. 10-2000-0063265 (published on November 6, 200) as a conventional technique for detecting snoring signals from snoring sounds "Snoring Acoustic Identification Based on Acoustic Identification Using a Neural Network" Method", Korean Patent Publication No. 10-2013-0015071 (published on Feb. 13, 2013) "Anti-snoring system and method for preventing snoring using the same", Korean Patent Publication No. 10-1299679 (published on Aug. 27, 2013) , "Anti-snoring pillow", Korean Patent Publication No. 10-2016-0096905 (published on August 17, 2016) "Method for detecting snoring by reconstructing acoustic signals", (Patent Document 005) Korean Patent Registration No. 10-2068484 Issue (2020.01.15 registration) "Method of generating sleep apnea prediction model and sleep apnea prediction method using this model", etc.

The prior art as described above analyzes sound to determine whether snoring occurs, provides information on the snoring state, makes a sound or gives motion to prevent snoring, controls a snoring prevention pillow or mattress, and the like. is supposed to do

By the way, since snoring is greatly affected by the sleeping posture, snoring can be prevented to some extent by correcting the posture during sleep. In addition, in the case of an obstructive sleep apnea (OSA) patient, it is possible to refer to a request for a professional medical diagnosis or whether surgery is possible by knowing the location of the obstruction where snoring occurs. However, in the prior art, a method for finding the snoring location using only the snoring sound has not been proposed.

Republic of Korea Patent Publication No. 10-2000-0063265 (published on November 6, 200) "Acoustic identification method for snoring based on acoustic identification using a neural network", Republic of Korea Patent Publication No. 10-2013-0015071 (published on February 13, 2013) "Snoring prevention system and snoring prevention method using the same", Korean Patent Registration No. 10-1299679 (published on August 27, 2013), "anti-snoring pillow", Korean Patent Publication No. 10-2016-0096905 (published on August 17, 2016) "Method for detecting snoring by reconstructing acoustic signals" Korean Patent Registration No. 10-2068484 (registered on January 15, 2020) "Method of generating sleep apnea prediction model and prediction method of sleep apnea using this model"

A study on snoring detection method using LDA. Korean Society for Communications and Communications 2015 Winter Conference, 2015.01

An object of the present invention is to detect the snoring position according to the snoring sound, bio-signals such as oxygen saturation, posture during sleep and snoring sound, and change trend according to the snoring monitoring cumulative information in consideration of snoring and sleep A snoring analysis service method that provides snoring analysis and disease diagnosis prediction services based on bio-signal analysis to provide diagnosis and prediction of the disease structure that causes snoring by evaluation and recommendation information for improving the user's sleep quality, and to provide the system.

A snoring analysis service method for providing a snoring sound analysis-based snoring analysis and disease diagnosis prediction service according to an embodiment of the present invention for solving the above object is a snoring analysis service by detecting a snoring state by sound analysis. A snoring analysis service method providing information service, comprising: a snoring detection step of analyzing audio collected during a user's sleeping time to detect snoring monitoring information including sleep apnea and snoring location; a posture detection step of detecting the user's sleeping posture by analyzing the sleeping posture information measured during the user's sleeping time; matching the posture information with snoring monitoring information during sleep and matching the snoring location information by modeling a breathing path as an anatomical path; a step of diagnosing and predicting a disease that may cause snoring by accumulatively managing the snoring monitoring information, snoring evaluation information, and sleep evaluation information, and analyzing a snoring state and change trend of the evaluation information; a service step of providing the snoring state monitoring information, snoring evaluation information, and sleep evaluation information, as well as providing the diagnosis prediction result and recommendation information for preventing snoring; includes

In addition, the snoring analysis service system, which provides snoring analysis based on snoring sound analysis and disease diagnosis prediction service according to the present invention for implementing the above method, collects snoring sound and analyzes audio characteristics to monitor snoring information. and a snoring analysis module for detecting snoring location information; a posture-snoring relationship modeling unit that detects a sleeping posture, maps the snoring monitoring information, and maps snoring position information to human body modeling information; A disease prediction module that accumulates and manages the snoring monitoring information, snoring location information, and correlation information between posture and snoring in a database and analyzes the cumulatively managed information and their change trends to diagnose and predict diseases that will cause snoring. ; includes

The present invention can detect the snoring state by sound analysis, and by matching the snoring state information to the posture information during sleep, it is possible to extract and provide information on the position where snoring occurs and information on the degree of snoring. . By extracting the correlation between snoring and posture, the user can easily find and apply a posture that can prevent or reduce snoring and an auxiliary means such as an anti-snoring pillow.

In addition, according to the present invention, snoring and snoring location can be detected by sound analysis, so snoring occurrence location information can be easily grasped without using expensive professional equipment by snoring and snoring location analysis. .

In addition, according to the present invention, snoring detection by snoring sound analysis, snoring location detection, and posture detection sensors can be used to detect posture during sleep and provide monitoring information. In addition, it is possible to analyze the snoring condition by posture by matching snoring, snoring position, and posture information, accumulate monitoring information and manage it as a history, analyze the status change trend to predict possible diseases, or consult a specialist for snoring. By providing this monitoring information, it is possible to provide guidance for accurate diagnosis, and it is effective to provide recommendation information for improving sleep quality and preventing snoring.

1 is a schematic diagram of a snoring and posture analysis service system according to the present invention.
2 is a schematic diagram of a user smart terminal according to the present invention.
3 is a conceptual diagram of a snoring and posture analysis unit of a server according to the present invention.
Figure 4 (a) and (b) is an exemplary view of the sleeping posture according to the sleeping posture and the use of a pillow for explaining the present invention.
5 is an exemplary view of a snoring position for explaining the present invention.
6 is a schematic diagram of a service providing unit of a server according to the present invention.
7 is an exemplary view of an analysis result report generation unit according to the present invention.
8 is a conceptual diagram of a snoring analysis apparatus composed of a stand-alone device according to another embodiment of the present invention.
9 is an explanatory diagram for explaining a snoring and posture monitoring method according to an embodiment of the present invention.
10 is an explanatory diagram for explaining a method for detecting a snoring position by analyzing a snoring sound according to an embodiment of the present invention.
11 is an explanatory diagram explaining a method of providing a diagnosis prediction and recommendation information service based on snoring analysis according to an embodiment of the present invention.

The following merely illustrates the principles of the present invention. Therefore, those skilled in the art can invent various devices and methods that embody the principles of the present invention and fall within the concept and scope of the present invention, even though not explicitly described or illustrated herein. In addition, all conditional terms and embodiments listed in this specification are, in principle, expressly intended only for the purpose of understanding the concept of the present invention, and should be understood not to be limited to such specifically listed embodiments and conditions. do.

In addition, it should be understood that all detailed descriptions reciting specific embodiments, as well as principles, aspects and embodiments of the present invention, are intended to encompass structural and functional equivalents of these matters. In addition, it should be understood that such equivalents include not only currently known equivalents but also equivalents developed in the future, that is, all devices invented to perform the same function regardless of structure.

Thus, for example, the block diagrams herein should be understood to represent conceptual views of illustrative circuits embodying the principles of the present invention. Similarly, all explanatory diagrams, state transition diagrams, pseudo code, etc. are meant to be substantially readable on computer readable media and represent various processes performed by a computer or processor, whether or not the computer or processor is explicitly depicted. It should be understood.

In addition, the explicit use of terms presented as processor, control, or similar concepts should not be construed as exclusively citing hardware capable of executing software, and without limitation, digital signal processor (DSP) hardware, ROM for storing software (ROM), random access memory (RAM) and non-volatile memory. Other hardware for the governor's use may also be included.

The above objects, features and advantages will become more apparent through the following detailed description in conjunction with the accompanying drawings, and accordingly, those skilled in the art to which the present invention belongs can easily implement the technical idea of the present invention. There will be. In addition, in carrying out the present invention, if it is determined that the detailed description of the known technology related to the present invention may unnecessarily obscure the subject matter of the present invention, the detailed description will be omitted.

Terms used in this application are only used to describe specific embodiments, and are not intended to limit the present invention. Singular expressions include plural expressions unless the context clearly dictates otherwise. In this application, the terms "include" or "have" are intended to designate that there is a feature, number, step, operation, component, part, or combination thereof described in the specification, but one or more other features It should be understood that the presence or addition of numbers, steps, operations, components, parts, or combinations thereof is not precluded.

Hereinafter, with reference to the accompanying drawings, preferred embodiments of the present invention will be described in more detail. In order to facilitate overall understanding in the description of the present invention, the same reference numerals are used for the same components in the drawings, and redundant descriptions of the same components are omitted.

Hereinafter, an embodiment of the present invention will be described in detail with reference to the accompanying drawings.

1 is a schematic diagram of a snoring and posture analysis service system according to the present invention.

As shown in FIG. 1, it includes an audio input means 10, a sleep posture measurement means 20, a smart terminal 40, and a service server 100.

The audio input means 10 may be a microphone that receives sound including snoring sound generated during sleep of the user.

The sleep posture measurement unit 20 may be a position awareness device (eg, gyroscope) for detecting a sleeping posture of a user during sleep, another type of sensor, or a camera device for taking pictures. As the sleep posture measuring means 20, a gyro sensor that can detect posture by placing it on the chest and another sensor attached to the forehead is used, or a lidar sensor that measures position change for each part of the user's body is used to measure sleep The middle posture can also be detected.

Meanwhile, in the present invention, a biometric information measuring unit 30 may be further included. In order to further increase accuracy in predicting a disease based on snoring analysis information during sleep and snoring analysis information of a user only by sound analysis and posture analysis, the biometric information measuring means 30 can be additionally configured.

The biometric information measuring means 30 may include an oxygen saturation detection sensor. In addition to the oxygen saturation detection sensor, biometric information measurement means 30 for snoring analysis such as heart rate detection means, brain wave detection means, respiration detection means, etc., which are common biosignal detection means, may be further included.

The smart terminal 40 may be implemented as any one of a smart phone, a tablet PC, or a dedicated smart terminal, and a snoring analysis application may be installed. The audio signal and the sleep position measurement signal may be received and preprocessed, respectively, to synchronize the audio data and the sleep position information data, and transmit the same to the service server 100 .

The smart terminal 40 transmits audio data and sleep posture data during the sleep period from the start time to the end time of running the snoring analysis application to the service server 100 so that the service server can perform all analysis processing there is.

However, it is not limited to this, and the smart terminal detects the snoring section and the apneic section between the snoring sections using frequency analysis of the audio data, transmits the snoring data service server, and the service server transmits the audio of the snoring section. The snoring position can be detected by analyzing the data through the learning module.

In addition, in the case of using an imaging camera as the sleep posture measuring means 20, the operation of the imaging camera may be controlled using a measurement event of the biometric information measuring means 30. That is, only the sleep position image data of the time corresponding to the snoring detection section may be extracted and transmitted to the service server 100 .

The service server 100 may be a cloud service server capable of communicating with a plurality of smart terminals 40 through a network. The service server 100 includes an audio analyzer 110, a sleep posture analyzer 120, a biometric information analyzer 130, a snoring and posture analyzer 200, a service provider 400, and a database 400. ).

The snoring and posture analysis unit 200 analyzes the received audio data through the audio analysis unit 110 to detect snoring and snoring location information to analyze the snoring state. In addition, the user's sleeping posture is extracted by analyzing the received sleeping posture information through the sleeping posture analyzing unit 120 . In addition, the received biometric information may be analyzed through the biometric information analysis unit 130 and reflected in snoring and snoring location analysis, posture analysis, and disease prediction, or disease prediction or recommendation by analyzing and processing the biometric information itself. It can also be used as informational information.

In addition, the snoring analysis and the snoring location and posture information are matched to analyze the relationship between the posture and the snoring and the relationship between the posture and the snoring position. Such snoring analysis, snoring location analysis, posture analysis, mutual relationship analysis, etc. are analyzed by artificial intelligence learning by constructing a learning module based on learning data and using the input audio analysis information and posture analysis information. .

In addition, diagnosis and evaluation of obstructive apnea based on snoring information analyzed by the snoring and posture analysis unit 200, snoring location and posture information, etc., analysis of change trend using accumulated information, etc. Disease prediction and diagnosis, recommendation information generation, and the like may be further included.

The service providing unit 400 may provide various service information based on snoring analysis information, snoring location information, posture analysis information, snoring and posture, and snoring position and posture correlation analysis information. . The snoring condition monitoring information during one sleep period, the relationship between snoring and posture, and the snoring location and posture information can be expressed in graphic or 3D format and provided. , disease prediction and diagnosis, and recommendation information for snoring prevention can be provided.

2 is a schematic diagram of a user smart terminal according to the present invention. As shown in Figure 2, a snoring analysis application may be installed in the smart terminal 40, an audio processing unit 41 that receives an audio signal, removes noise, pre-processes it, and converts it into audio data, and image information. A sleep position information processor 42 that receives input and converts it into image data by removing noise and preprocessing; A synchronization and control unit 44 for synchronizing and controlling the transmission of posture information data, and transmitting the audio data and sleep posture information data to the service server 100 through a network under the control of the time synchronization and control unit 44, A data transmission/reception unit 45 for receiving snoring analysis result information provided from the service server 100, a user interface for providing a user interface and outputting the result information analyzed by the service server 100, and a service provision Includes study (46).

The audio processing unit 41 receives sound input as an analog signal, that is, an audio signal. The input audio signal is noise-filtered and converted into a data signal. Similarly, the sleep position information processing unit 42 converts the signal input from the sleep position measuring device into position signal data through noise filtering. In the case of input from a digital camera, preprocessing may be completed with image data only by filtering noise. In addition, the biometric information processor 43 performs noise processing on the signals of sensors that measure biosignals and converts them into data signals. When an oxygen saturation measuring sensor is provided as the biological signal measuring means 30, the oxygen saturation sensing signal is filtered and converted into data.

The synchronization and controller 44 can control the snoring analysis app execution mode, such as audio input and sleep posture measurement, and can control the level or time condition desired by the user. The audio data and the video data are synchronized control to transmit. The reason for synchronization is to meet the time standard that matches the audio and posture.

In the present invention, sound is detected to analyze the snoring state. At this time, the posture is detected to detect posture information during sleep, that is, in the snoring occurrence section. Therefore, the posture detection may control the shooting start and shooting standby based on the motion detection, and in this case, the audio and synchronization signal may be matched.

The user interface and service providing unit 46 may provide a user interface such as running or stopping the user's snoring analysis application, viewing the user's monitoring information history, checking status analysis information, or inputting a related questionnaire.

In addition, the snoring analysis monitoring information provided from the service server 100 may be displayed as a graph or numerical value and modeling of the mouth and throat for snoring occurrence as a location and stricture state indicator, or statistics or changes in the snoring analysis information. You can print status etc.

In addition, it is possible to provide diagnostic prediction, alternative guide information output, and recommendation information. Such information output may be simply output by receiving full information from the service server, but processing such as graphs, human body modeling, and analysis report generation may be performed directly in a smart terminal using data received from the service server.

3 is a conceptual diagram of a snoring and posture analysis unit of the service server according to the present invention. As shown in Figure 3,

An audio analysis unit 110, a snoring analysis module 200, a snoring information generator 230, a sleep posture analysis unit 120, a posture and snoring relationship modeling unit 250, It may include a snoring relation generator 260, a biometric information analyzer 130, a disease prediction module 270, and a disease prediction information generator 280.

The audio analysis unit 110 performs audio analysis for snoring detection. For example, in order to determine snoring, audio characteristics such as a frequency-related index, a frequency change-related index, an intensity change-related index, a noise-related index of the snoring sound, a tremor index, and a disconnection index are analyzed.

The snoring analysis module 200 uses the audio characteristics analyzed by the audio analysis unit 110 to analyze whether snoring, apnea, the snoring state, and the snoring position are analyzed, and the snoring pattern and snoring Conduct a comprehensive analysis. It is desirable to perform artificial intelligence analysis using a learning model. In the case of detecting the snoring location, it is necessary to construct learning data by analyzing the correlation between the snoring characteristic information and the snoring location.

The learning data is pitch and formant according to the part of the upper airway obstruction identified by drug-induced sleep endoscopy or sleep videofluoroscopy, which is a method for measuring the position of snoring in a hospital. The information of the change in the acoustic characteristics of the snoring sound is analyzed, and the result is constructed as learning data. Snoring acoustic parameters such as pitch (minimum, average, maximum) and formant are analyzed, and learning data is built by reflecting them. The site of obstruction is determined by SVF and can be classified for anatomical structure and level of obstruction.

After constructing the learning data as described above, a learning module is constructed using the learning data. When the learning module is built, the user's audio analysis characteristic information is applied to the learning module to extract snoring location information according to the user's snoring sound analysis. For example, the mean value of the peak frequency can analyze a significant difference between soft palate and separate tongue-basal or epiglottis obstruction and soft palate and tongue-basal or epiglottis-related complex obstruction.

On the other hand, in the snoring analysis, in order to accurately detect the sleep apnea state or determine the degree of snoring, oxygen saturation is detected using a sensor, and if oxygen saturation information is reflected in snoring detection, more accurate snoring state information is analyzed. You will be able to.

In addition, in analyzing the state of snoring, sound detection information at different locations using a plurality of sensors at different locations, that is, time difference information or feature information such as the degree of sound attenuation at each location is extracted, , this can be reflected in snoring detection by sound analysis. This can be used more effectively to analyze the location of snoring or the degree of snoring.

The snoring information generating unit 230 generates information to be serviced using information analyzed through the learning model. Snoring monitoring information such as snoring status overnight, that is, snoring occurrence time, number of times, and snoring intensity, snoring location information at each snoring occurrence, and snoring information including information on the degree of blockage of snoring location, etc. generate relevant information.

* The sleep posture analysis unit 120 analyzes the sleep posture information to detect the user's posture information during sleep. This can acquire posture information by detecting the user's body part as an object and analyzing the relationship between each body part. In particular, it is necessary to extract posture information corresponding to the correlation between the user's body, neck and head. Because the cause of snoring is greatly influenced by posture, it is necessary to analyze the correlation between posture and snoring.

As the means for measuring the sleeping posture, a video camera may be used, and a method of detecting the user's sleeping posture by photographing the sleeping posture from a plurality of angles and combining the images from each direction into a 3D modeling image may be used. . In addition, in the case of using a depth sensor camera for imaging, an object can be analyzed in three dimensions, and the entire head and neck, including the bones and soft structures of the upper respiratory tract, can be visualized and expressed.

In the embodiment of the present invention, the description is based on image capture, but is not limited thereto, and as a sleep posture measuring means 20, a gyro sensor that can detect the posture by placing it on the chest and another sensor attached to the forehead It is also possible to detect the user's posture during sleep by using a lidar sensor or the like that measures a change in position of each body part of the user.

The posture and snoring relationship modeling unit 250 uses the posture information analyzed by the sleep posture analysis unit 120 to generate posture modeling showing postures of the user's body, neck, and head.

Figure 4 (a) and (b) is an exemplary view of the sleeping posture according to the sleeping posture and the use of a pillow for explaining the present invention.

Postures during sleep can be classified according to the postures of the entire body when viewed from above. For example, as shown in (a) of FIG. 4, the posture crouched sideways (fetal type), the posture stretched sideways (log type), the posture bent sideways (craving type, the correct lying posture (soldier type, up It can be classified into various postures such as a raised posture, a lying posture with arms and legs bent, and also, as shown in FIG. Each pose is different depending on how you use it.

It can be seen that the above sleeping posture has a sufficient effect on snoring. Therefore, if the correlation between the posture where snoring occurs and the degree of snoring by each posture is analyzed, snoring can be prevented to some extent just by correcting the posture during sleep.

As for the method of measuring the posture during sleep, it is preferable to measure both the body posture measured from the top of the sleeping posture and the head posture according to the use of a pillow from both sides. When sleeping, etc., it will be difficult to measure an accurate user posture. Therefore, based on the body posture measured from the top and the head posture measured from both sides of the head and neck without using the table, each learning data is built, an artificial intelligence learning model using this is built, and sleep when using the table is If the artificial intelligence learning method is applied using the learning model by measuring the posture, the user's posture information can be sufficiently measured even when using the table, and the sleeping posture information can be analyzed and modeled.

Therefore, in the present invention, the posture during sleep is measured by measuring the posture of the user's body during sleep and the head posture according to the relationship between the neck, head, and pillow, and synchronizing the body posture and head posture to visualize each. It is modeled in a human body shape, and the snoring monitoring information is mapped to the modeling information so that the correlation between snoring and body posture and snoring and head posture can be visually expressed and modeled.

* Figure 5 is an exemplary position of snoring occurring in the nasal cavity and airway for explaining the present invention. As shown in this, snoring can be divided into nasal snoring and airway snoring. In the case of nasal snoring, nasal congestion due to deterioration of the nasal cavity and paranasal sinuses, symptoms such as runny nose may appear, and breathing holes when sleeping Snoring is the sound produced by the narrowing of the nose.

In addition, as the narrowing of the nasal cavity and paranasal sinuses worsens, it becomes difficult to snore and breathe, resulting in natural mouth breathing. As a result, when the tongue is relaxed and rolled backwards to close the air passage, sleep apnea may occur, and as the airway is narrowed, vibration of the palate or neck occurs, resulting in oral snoring. In the current treatment method, treatment is performed on the uvula, soft palate, and tongue to resolve airway obstruction during sleep.

Therefore, in analyzing snoring as described above, it is very important to specify the location of snoring, and it is very important to specify the location of snoring so that the cause of snoring can be identified to find a treatment method or a method to prevent snoring. It will be.

In the present invention, the snoring position is extracted by sound analysis based on the correlation between the characteristics of the sound analysis and the snoring position, and the snoring position and the degree of snoring are visualized through human body modeling. Based on the same modeling, it is possible to display the location of snoring.

Snoring detection information and snoring location information are received from the snoring analysis module 220, and snoring detection and snoring location information are matched with the posture modeling and anatomical mouth and larynx air passage modeling information, respectively, and snoring relationship modeling information is generated. In addition, when information about the degree of narrowing of the airway is extracted from snoring location detection information by audio analysis, the degree of narrowing of the airway may be modeled and generated using this. Then, the relationship between posture and snoring and snoring position is analyzed.

The posture and snoring relationship generating unit 260 creates visualization information by visually matching the relationship between the posture modeled by the modeling unit 250 and snoring.

The disease prediction learning module 270 enables disease prediction and recommendation information to be provided using monitoring information on the user's snoring and cumulatively managed information on snoring and postural relationship information. This is to construct current and future disease prediction and recommendation information that can be considered for snoring status information, snoring status change information, posture and snoring information, and snoring location information as learning data, and to build disease using learning data Build a prediction and recommendation information learning model. This allows the learning module to be applied by constructing a large amount of learning data in parallel with supervised learning or unsupervised learning.

The disease prediction information generator 280 diagnoses snoring and snoring location, diagnoses status change, diagnoses snoring and posture and posture change, and generates recommendation information through a learning module.

Meanwhile, in the present invention, the biometric information measurement unit 30 may be further included. As the biometric information measuring means 30, various biometric information measuring means such as a breathing detection sensor, an oxygen saturation detection sensor, a heart rate measuring sensor, an EEG measuring sensor, and the like can be further included and utilized. When such biometric information is used, the severity of sleep apnea can be analyzed in more detail by reflecting the detected oxygen saturation in the case of sleep apnea detection, compared to analyzing snoring only by simple sound analysis.

In addition, it would be more effective to reflect biometric information in sleep quality evaluation or disease prediction. In addition, it is possible to extract event occurrence (e.g., event occurrence where oxygen saturation falls below the threshold value, arrhythmia judgment event, change event of EEG information during sleep, etc.) from biometric information and reflect it in posture detection or sound analysis in response to event occurrence. It can be used as cumulative change information for disease prediction.

6 is a schematic diagram of a service providing unit of a server according to the present invention. As shown in FIG. 6 , the service provider 300 of the server includes an analysis result report generator 310 and a service information output unit 320 .

7 is an exemplary view of an analysis result report generation unit according to the present invention. As shown in FIG. 7 , the analysis result report generating unit 310 includes a snoring state information generating unit 311, a posture modeling unit 312, a snoring position modeling unit 313, and evaluation information generation. It may include a unit 314, a diagnosis prediction information generator 315, and a recommendation information generator 316.

The snoring state information generation unit 311 determines whether snoring occurs during the sleep period, the intensity of snoring, and sleep based on the snoring state information and snoring location information from the snoring and posture analysis unit 200 The apneic state is created as state information including a graph.

The posture modeling unit 312 models the sleeping posture of the user during sleep based on the posture information and the relationship information between the posture and snoring, and associates it with snoring so that it is possible to know which snoring occurs in which posture. to model In addition, the posture modeling unit may be implemented as a video including snoring audio so that the posture of the snoring section and the apnea section can be easily checked.

The snoring position modeling unit 313 anatomically visualizes and models the oral cavity and larynx of the human body based on air passages according to respiration, and expresses snoring position and stricture information according to the degree of snoring. to model In addition, by modeling air flow due to breathing and providing it as a dynamic expression, a modeled image including the cause of snoring, the location of snoring, the degree of stricture, and the sound of snoring may be generated and provided.

The evaluation information generation unit 314 uses the snoring section for one day of sleep, the sleep apnea section, the snoring intensity, the snoring location, the posture at the time of snoring, the number of occurrences or statistical information thereof Evaluation information may be generated by evaluating sleep and quantifying analysis statistical information related to snoring and posture.

The diagnostic prediction information generation unit 315 considers the currently detected snoring state information, location information, posture information, etc., and tracks the state change of the user's snoring analysis information accumulated and managed in the database 400, Based on the user's snoring state and state change, disease diagnosis prediction information that may cause snoring may be generated and provided.

The recommendation information generation unit 316 may generate recommendation information for preventing snoring based on the snoring and posture analysis and recommendation information based on the diagnosis prediction information. The recommendation information generation may use a learning module that generates optimal recommendation information by analyzing snoring state information, location information, posture information, and respective change trend information using an artificial intelligence learning module.

8 is a conceptual diagram of a snoring analysis apparatus composed of a stand-alone device according to another embodiment of the present invention. As shown in this,

As shown in the outline of FIG. 1, the present invention may be configured as an online system that transmits measurement information from a smart terminal to a service server and provides information analyzed by the service server, but is not limited thereto, as shown in FIG. It can also be configured as a stand-alone device with measuring means. If the snoring analyzer is configured as a stand-alone device type, it can be implemented offline without going through various communication limitations for health management.

In addition, if the analysis and service providing functions of the service server of the present invention are integrated into a smart terminal, the smart terminal itself can be configured as a stand-alone device and implemented by connecting only the measuring means to the smart terminal. A simple function can be implemented by using a built learning model or learning data.

In addition, the apparatus of the present invention described above is a snoring posture monitoring method for selectively matching and monitoring snoring and posture, a snoring position detection method for detecting and analyzing snoring and snoring position by sound analysis, and In addition, it is possible to provide a method for providing a snoring detection and analysis service that matches and analyzes snoring, posture, and snoring location, and provides disease prediction and recommendation services using these history information.

9 is an explanatory diagram for explaining a snoring and posture monitoring method according to an embodiment of the present invention. As shown in Figure 9,

The snoring monitoring method according to an embodiment of the present invention for solving the problem of the first object is,

A snoring detection step (S10) of collecting sound generated during the user's sleep time and detecting snoring and apnea by audio analysis; A posture detection step (S20) of acquiring sleep posture information during the user's sleeping time and detecting a posture during sleep; and outputting snoring posture information by matching the snoring detection information to the sleeping posture information to analyze a correlation between snoring and posture (S30).

In the snoring detection step (S10), the service server 100 receives audio data including the user's snoring sound during sleep through the smart terminal 40 and analyzes the audio. Through the audio analysis unit 110, sound characteristic parameters including frequency and frequency change amplitude and pattern are extracted.

By inputting the analyzed sound characteristic parameters to the snoring analysis module 220, the snoring state can be detected through the snoring detection learning model. Of course, it is possible to have a learning module that detects only snoring and apnea status, but simple snoring monitoring and apnea status information detection can be performed by using only parameters such as frequency characteristics and sound intensity without using the learning module. It can be detected only with an algorithm that detects this.

In the posture detection step (S20), the service server 100 receives image information of a user during sleep from the smart terminal 40, extracts a user object from the image information analyzer 120, and detects motion information. Thus, posture change information according to movement may be detected, and posture change information may be analyzed by modeling user object information.

In the snoring and posture relationship matching step (S30), when the posture information is analyzed as described above, the snoring information is matched with the posture information, and the interaction between snoring and posture is performed through the posture and snoring relationship modeling unit 250. Relationship information can be modeled and provided. That is, it is possible to analyze and provide information such as the posture and posture change in the snoring section, the degree of snoring in each posture, and whether apnea occurs.

Therefore, when snoring and posture information are matched and provided, it is possible to know in what position the user snoring occurs during sleep, the occurrence of snoring and the frequency of snoring according to the sleep progress time information, and the snoring by posture Analysis information such as intensity, duration, and apnea interval can be obtained. In this way, if snoring and posture relationship analysis information is known, it is possible to take measures to correct posture during sleep or use related products, and to recommend sleeping posture or related tools or products to the user according to the analyzed results. There is. A snoring and postural relationship monitoring service and product recommendation information can also be used as a profit model.

10 is an explanatory diagram for explaining a method for detecting a snoring position by analyzing a snoring sound according to an embodiment of the present invention. As shown in Figure 10,

A snoring sound analysis step (S110) of collecting sounds generated during the user's sleeping time and extracting snoring sound characteristics by analyzing the collected sound; and constructing learning data based on clinical data on the correlation between snoring sound characteristics and snoring position, and analyzing the snoring sound characteristics extracted in the sound analysis step through a learning model built using the learning data. Detecting the corresponding snoring location (S120); Anatomically visualizing and modeling the breathing path of the human body and matching the snoring location information (S130); includes.

The snoring sound analysis step (S110) collects and receives the user's sound during sleep from the smart terminal 40, analyzes it through the audio analysis unit 110 to detect sound characteristic information, and snoring analysis module ( 100), the snoring state is detected and the snoring position is detected. At this time, the snoring analysis module 100 based on the snoring sound characteristics and the snoring location relationship data based on the audio analysis is established as the learning data, and the snoring status information and snoring location are detected using the learning module, Analyze. The breathing path of the human body is anatomically visualized and modeled, and the snoring location information is matched using the analyzed information to provide a service as snoring location analysis information.

As described above, snoring state monitoring information and snoring location information may be detected by sound analysis. Since the snoring location information detects the location of the snoring in the section where the obstructive apnea symptom occurs, it is possible to analyze information such as whether the snoring is complex or one part, and the degree of snoring. If the snoring condition and snoring location information analysis information are known, it can be used as information for the purpose of preventing and treating snoring, such as whether a professional medical diagnosis is required, whether surgery is necessary, and information on the location to be operated.

11 is an explanatory diagram explaining a method of providing a diagnosis prediction and recommendation information service based on snoring analysis according to an embodiment of the present invention. As shown in Figure 11,

A snoring detection step (S210) of collecting sound generated during the user's sleep time and detecting snoring by audio analysis; a snoring location detection step (S230) of detecting snoring sound characteristics extracted by the audio analysis and detecting snoring location information correlated with sound characteristics through a learning model; Posture detection step (S230) of detecting a sleeping posture using the posture information collected during the user's sleeping time; modeling the posture information to match snoring state information during sleep, modeling a breathing path according to the posture information as an anatomical path, and matching the snoring position information (250); a snoring evaluation step 260 of mutually matching the snoring detection information, the posture detection information, and the snoring position information to provide snoring monitoring information, as well as performing snoring state evaluation analysis and sleep evaluation analysis; And cumulatively managing the snoring state monitoring information, snoring evaluation information, and sleep evaluation information, and analyzing the change trend of the snoring state and evaluation information to predict diagnosis of a disease that may cause snoring, diagnosis prediction, and A service step (270) of providing recommendation information for preventing snoring; includes

The snoring state detection step (S210) collects the sound of the user during sleep and detects the snoring state by audio analysis, and the snoring position detection step (S220) is based on the snoring characteristics extracted by audio analysis. The snoring state and the snoring position can be extracted using the parameters for the snoring and using the learning module 201 built in advance. Of course, the snoring state detection (S210) and the snoring position detection (S220) may be performed together using the sound characteristic parameters analyzed by the audio analysis using the learning module.

In the posture detection step (S230), an image of the user during sleep is captured, the image is analyzed, and the user's posture is detected as an object by image analysis (S230) to perform posture modeling and anatomical breathing path modeling (S231). there is.

In step S250 of matching the positional relationship between the posture and snoring, the snoring detection information and the snoring position information are matched with the posture information. Posture, snoring, and snoring position information are matched with modeling information so that the mutual relationship can be known by matching the posture modeling information and the breathing path modeling information.

In the snoring evaluation step (S260), snoring detection information, posture information, and matched snoring location relationship information are provided as monitoring information, and snoring and sleep analysis and evaluation are performed. At this time, the learning module (S202) based on evaluation analysis, disease prediction corresponding to change trend, and recommendation information learning data can be used.

In the service step (S270), snoring monitoring information and evaluation information are accumulated and managed, and disease prediction and recommendation information are provided by analyzing the trend of change. At this time, the learning module (S202) based on evaluation analysis, disease prediction corresponding to change trend, and recommendation information learning data can be used.

In this way, snoring analysis by sound analysis, snoring location analysis, and posture analysis during sleep are performed. The modeled posture information, snoring information, and snoring location information are matched and provided as snoring monitoring service information. It is cumulatively managed and analyzed to provide snoring condition analysis information and sleep evaluation information. It is possible to provide disease prediction information that can cause snoring and corresponding recommended information as service information by analyzing the trend of snoring state information accumulated through the learning module and the posture and sleep evaluation information.

The method according to the present invention described above may be produced as a program to be executed on a computer and stored in a computer-readable recording medium. Examples of the computer-readable recording medium include ROM, RAM, CD-ROM, and magnetic tape. , floppy disks, optical data storage devices, and the like, and also includes those implemented in the form of carrier waves (for example, transmission through the Internet).

The computer-readable recording medium is distributed to computer systems connected through a network, so that computer-readable codes can be stored and executed in a distributed manner. In addition, functional programs, codes, and code segments for implementing the method can be easily inferred by programmers in the technical field to which the present invention belongs.

In addition, although the preferred embodiments of the present invention have been shown and described above, the present invention is not limited to the specific embodiments described above, and the technical field to which the present invention belongs without departing from the gist of the present invention claimed in the claims. Of course, various modifications can be made by those skilled in the art, and these modifications should not be individually understood from the technical spirit or perspective of the present invention.

*10: Audio input means
20: means for measuring sleep posture information
30: biometric information measuring means
40: smart terminal
100: service server
110: audio analysis unit
120: sleep posture analysis unit
130: biometric information analysis unit
200: snoring and posture analysis unit
220: snoring analysis module
250: Posture and snoring relationship modeling unit
270: disease prediction module
300: service provider
400: database

Claims (3)

A snoring analysis service method for detecting a snoring state by sound analysis and providing snoring analysis information,
a snoring detection step of analyzing audio collected during a user's sleep time to detect snoring monitoring information including sleep apnea and a snoring location;
From the image information obtained by photographing the user's sleeping posture for a plurality of angles during the user's sleeping time, the user's body posture during sleep measured from the upper part of the sleeping posture, and the neck and neck according to the use of a pillow measured from both sides of the sleeping posture A sleeping posture modeling step of measuring the user's head posture according to the relationship between the head and the pillow, and modeling the user's body posture and the head posture by using pre-established learning data to create a visualized human body shape;
By matching snoring monitoring information during sleep to the modeled sleep posture information, modeling is performed to visually express the correlation between snoring and body posture and snoring and head posture, and mapping snoring location information to human body modeling information. modeling a posture and snoring relationship so that a snoring location is displayed on anatomical mouth and larynx air passage modeling information;
The posture and snoring correlation information modeled in the posture and snoring relationship modeling step, the snoring monitoring information, snoring evaluation information, and sleep evaluation information are accumulated and managed, and the posture and snoring correlation information, a disease diagnosis and prediction step for a disease that may cause snoring by analyzing a change trend of the snoring monitoring information, snoring evaluation information, and sleep evaluation information;
In addition to providing the posture and snoring correlation information, the snoring monitoring information, snoring evaluation information, and sleep evaluation information, diagnosis prediction results according to the execution of the disease diagnosis prediction step and recommendation information for preventing snoring are provided. service step; containing
Snoring analysis service method. According to claim 1,
The snoring detection step,
a snoring sound analysis step of collecting sounds generated during the user's sleeping time and extracting snoring sound characteristic information by analyzing the collected sounds;
The snoring sound characteristic information is analyzed to detect snoring and the sleep apnea state, and snoring monitoring including at least one of a snoring section during sleep, snoring intensity of each snoring section, and snoring time information. information detection step;
Learning data is constructed based on clinical data on the correlation between the snoring sound characteristics and the snoring position, and the snoring sound extracted in the snoring sound analysis step is performed through a learning model built using the learning data. a snoring position detection step of detecting a snoring position corresponding to a characteristic; containing
Snoring analysis service method. According to claim 1,
The snoring detection information, the modeled sleep posture information, and the snoring position information are mutually matched and provided as snoring monitoring information, as well as the snoring condition evaluation analysis and sleep evaluation analysis to perform disease diagnosis prediction step disease Further comprising a snoring evaluation step of providing evaluation information to be reflected as analysis information for prediction and reflected as service information in the service step.
Snoring analysis service method.

Images