KR102900126B1 - Method, apparatus, and program for estimating health status based on refined ppg data - Google Patents

Abstract

A method for estimating health status information based on refined PPG data according to various embodiments of the present invention is disclosed. The method may include: acquiring a plurality of PPG data measured from a plurality of areas of a user's skin through a plurality of PPG sensors; acquiring at least one refined PPG data from the plurality of PPG data; and estimating the user's health status information based on the at least one refined PPG data.

Description

Method, apparatus, and program for estimating health status information based on refined PPG data {Method, apparatus, and program for estimating health status based on refined PPG data}

The present invention relates to a method, device, and program for estimating health status information based on refined PPG data, and more particularly, to a method, device, and program for estimating health status information including a user's blood pressure using refined PPG data obtained by refining a plurality of PPG data acquired using a plurality of PPG sensor data.

Portable health monitoring devices, such as wearables, have recently attracted attention in the medical and healthcare fields. Consequently, biosignal measurement based on photoplethysmography (PPG) technology is becoming more widespread. PPG is a technology that allows for noninvasive measurement of heart rate, blood pressure, and vascular status through simple optical measurements, and is being applied to various wearable devices.

PPG-based healthcare technology has the advantage of detecting signal changes related to blood flow, allowing for real-time estimation of various biometric data. In particular, devices such as smartwatches and fitness trackers utilize these PPG signals to monitor heart rate, which is useful for personalized healthcare.

However, existing PPG technology has several limitations in providing accurate health status information. Because PPG signals are sensitive to external factors and the user's movements, the quality of the measured data can deteriorate. Furthermore, data acquired from a single PPG sensor is highly dependent on ambient conditions, which leads to signal noise and makes it difficult to accurately estimate biosignals based on these signals.

For example, in the case of blood pressure measurement, PPG-based measurements provided by wearable devices require periodic calibration and often suffer from low accuracy. Specifically, some devices require users to regularly calibrate using an external device, such as a medical blood pressure monitor, and automated continuous measurement is difficult, reducing the reliability of blood pressure estimation.

To overcome these limitations, a measurement method utilizing multiple PPG sensors is needed to increase signal precision. Furthermore, research and development are needed on methods for estimating health status information, such as blood pressure, based on high-precision signals. In this regard, Republic of Korea Patent No. 10-2661814 discloses a method and device for estimating biosignals.

The present invention has been conceived in response to the aforementioned background technology, and is intended to provide a method, device, and program for estimating health status information based on purified PPG data.

The technical problems of the present invention are not limited to the technical problems mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art from the description below.

According to one embodiment of the present invention for solving the aforementioned problem, a method for estimating health status information based on refined PPG data is disclosed. The method may include: acquiring a plurality of PPG data measured in a plurality of areas of a user's skin through a plurality of PPG sensors; acquiring at least one refined PPG data from the plurality of PPG data; and estimating the user's health status information based on the at least one refined PPG data.

In an alternative embodiment, the method may further include, prior to acquiring the refined PPG data, synchronizing the scale and time of the plurality of PPG data; passing the synchronized plurality of PPG data through a specific bandpass filter to remove a signal component corresponding to a specific frequency band; and removing a linear trend of the plurality of PPG data.

In an alternative embodiment, the step of obtaining at least one refined PPG data from the plurality of PPG data may include the step of determining at least one of first PPG data having the highest quality among the plurality of PPG data, second PPG data calculated based on a weighted average of the plurality of PPG data, and third PPG data satisfying a specific measurement condition among the plurality of PPG data as main PPG data; and the step of removing noise from the main PPG data to obtain the at least one refined PPG data.

In an alternative embodiment, the step of obtaining at least one refined PPG data from the plurality of PPG data may include: synchronizing scales and times of the plurality of PPG data; comparing signal components of each of the synchronized plurality of PPG data to compare correlations at specific points in time or frequency bands; and recognizing at least one signal component having the correlation as a biosignal component and combining the at least one signal component to obtain the at least one refined PPG data.

In an alternative embodiment, the step of estimating the health status information of the user based on the at least one refined PPG data may include the step of obtaining arterial blood pressure data based on the at least one refined PPG data; and the step of estimating the health status information including blood pressure based on the arterial blood pressure data.

In an alternative embodiment, the step of obtaining arterial blood pressure data based on the at least one refined PPG data includes the step of inputting two or more refined PPG data into an arterial blood pressure conversion algorithm to obtain the arterial blood pressure data, wherein the arterial blood pressure conversion algorithm estimates a pulse wave transit time based on a time difference between the two or more refined PPG data, calculates a pulse wave velocity based on the pulse wave transit time, and calculates the arterial blood pressure data based on the pulse wave velocity and a distance between PPG sensors corresponding to each of the two or more refined PPG data.

In an alternative embodiment, the step of estimating the health status information including blood pressure based on the arterial blood pressure data includes the step of inputting the arterial blood pressure data into a pre-learned blood pressure estimation model to obtain the blood pressure; wherein the blood pressure estimation model extracts features for the pulse wave velocity and the pulse wave transit time from the arterial blood pressure data, predicts systolic blood pressure based on the features for the pulse wave velocity and the pulse wave transit time, and predicts diastolic blood pressure based on the features for the pulse wave transit time and the user's bio-information.

In an alternative embodiment, the step of estimating the health status information of the user based on the at least one refined PPG data may further include the steps of: obtaining at least one numerical value corresponding to each item included in the health status information based on the at least one refined PPG data; integrating the at least one numerical value to determine a final numerical value for each item included in the health status information; and generating the health status information including the final numerical value for each item.

According to one embodiment of the present invention for solving the above-described problem, a health status information estimation device is disclosed. The device includes: a plurality of PPG (Photoplethysmography) sensors; a memory storing one or more instructions; and a processor executing the one or more instructions related to processing of a plurality of PPG data measured by the plurality of PPG sensors; wherein the processor can acquire a plurality of PPG data measured in a plurality of areas of the user's skin through the plurality of PPG sensors, acquire at least one refined PPG data from the plurality of PPG data, and estimate the user's health status information based on the at least one refined PPG data.

According to one embodiment of the present invention for solving the above-described problem, a computer program is disclosed, which is coupled to a computer as hardware and is stored in a computer-readable recording medium to perform the steps of: acquiring a plurality of PPG (Photoplethysmography) data measured in a plurality of areas of a user's skin through a plurality of PPG sensors; acquiring at least one refined PPG data from the plurality of PPG data; and estimating health status information of the user based on the at least one refined PPG data.

Other specific details of the present invention are included in the detailed description and drawings.

The present invention refines multiple PPG data sets to estimate accurate and reliable health status information in real time. This allows users to accurately monitor various health indicators, such as blood pressure, stress levels, and blood sugar levels, even in everyday environments.

The effects of the present invention are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the description below.

FIG. 1 is a diagram illustrating a system according to one embodiment of the present invention.
Figure 2 is a hardware configuration diagram of a health status information estimation device according to one embodiment of the present invention.
FIG. 3 and FIG. 4 are drawings for explaining an example of a wearable health status information estimation device according to one embodiment of the present invention.
FIG. 5 and FIG. 6 are drawings for explaining an example of a stationary health status information estimation device according to one embodiment of the present invention.
FIGS. 7 to 10 are diagrams for explaining an example of a method for estimating health status information based on purified PPG data according to one embodiment of the present invention.

Various embodiments are now described with reference to the drawings. In this specification, various descriptions are provided to facilitate an understanding of the present invention. However, it will be apparent that these embodiments may be practiced without these specific details.

As used herein, the terms "component," "module," "system," and the like refer to computer-related entities, hardware, firmware, software, a combination of software and hardware, or an execution of software. For example, a component may be, but is not limited to, a procedure running on a processor, a processor, an object, a thread of execution, a program, and/or a computer. For example, both an application running on a computing device and the computing device may be a component. One or more components may reside within a processor and/or a thread of execution. A component may be localized within a single computer. A component may be distributed between two or more computers. Furthermore, these components may execute from various computer-readable media having various data structures stored therein. Components may communicate via local and/or remote processes, for example, by signals comprising one or more data packets (e.g., data from one component interacting with another component in a local system, a distributed system, and/or data transmitted to another system via a network such as the Internet via signals).

Furthermore, the term "or" is intended to mean an inclusive "or" rather than an exclusive "or." That is, unless otherwise specified or clear from context, "X employs A or B" is intended to mean either of the natural inclusive permutations. That is, if X employs A; X employs B; or X employs both A and B, "X employs A or B" can apply to any of these cases. Furthermore, the term "and/or" as used herein should be understood to refer to and include all possible combinations of one or more of the associated items listed.

Additionally, the terms "comprises" and/or "comprising" should be understood to imply the presence of the features and/or components in question. However, it should be understood that the terms "comprises" and/or "comprising" do not exclude the presence or addition of one or more other features, components, and/or groups thereof. Furthermore, unless otherwise specified or clear from the context to refer to the singular form, the singular in the specification and claims should generally be construed to mean "one or more."

Those skilled in the art should further recognize that the various illustrative logical blocks, components, modules, circuits, means, logics, and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both. To clearly illustrate the interchangeability of hardware and software, various illustrative components, blocks, components, means, logics, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the overall system. Skilled artisans may implement the described functionality in varying ways for each particular application. However, such implementation decisions should not be interpreted as causing a departure from the scope of the present disclosure.

The description of the disclosed embodiments is provided to enable those skilled in the art to make or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art. The general principles defined herein may be applied to other embodiments without departing from the scope of the present invention. Therefore, the present invention is not limited to the disclosed embodiments. The present invention is to be construed in the widest scope consistent with the principles and novel features disclosed herein.

In this specification, the term "computer" refers to any type of hardware device including at least one processor, and may also be understood to encompass software components operating on the hardware device, depending on the embodiment. For example, the term "computer" may be understood to encompass, but is not limited to, smartphones, tablet PCs, desktops, laptops, and all user clients and applications running on each device.

Hereinafter, embodiments of the present invention will be described in detail with reference to the attached drawings.

Although each step described in this specification is described as being performed by a computer, the subject of each step is not limited thereto, and at least some of each step may be performed by different devices depending on the embodiment.

FIG. 1 is a diagram illustrating a system according to one embodiment of the present invention.

Referring to FIG. 1, a system according to one embodiment of the present invention may include a health status information estimation device (100), a user terminal (200), and an external server (300). The system illustrated in FIG. 1 is according to one embodiment, and its components are not limited to the embodiment illustrated in FIG. 1, and may be added, changed, or deleted as needed.

The present invention relates to a technology for simultaneously measuring PPG data at various locations using two or more PPG sensors and performing noise removal and data fusion through various AI-based algorithms.

Through this technology, the present invention can improve measurement accuracy and reliability using a multi-sensor approach, and provide real-time biosignal monitoring and data analysis capabilities. Furthermore, the present invention can ensure user convenience through a non-invasive method. Furthermore, the technology of the present invention can be used for worker health monitoring, chronic disease patient health management, real-time biosignal analysis, and industrial safety management.

In one embodiment, the health status information estimation device (100) may include a wearable health status information estimation device (101). That is, the health status information estimation device (100) may be implemented in the form of a wearable device. For example, the wearable health status information estimation device (101) may be attached to a user's body in the form of a smartwatch or wristband, collect PPG data in real time, and estimate health status information based on the data.

In various embodiments, the health status information estimation device (100) may include a stationary health status information estimation device (102). That is, the health status information estimation device (100) may be implemented in a form that is fixed to a specific location and usable by various users. For example, the stationary health status information estimation device (102) may be a stationary device installed in a medical clinic or health care center, and may be designed so that multiple users can place their measurement areas on it and measure PPG data.

Below, descriptions of various forms in which the health status information estimation device (100) is implemented will be provided with reference to FIGS. 3 to 6.

In one embodiment, the health status information estimation device (100) can estimate (or measure) health status information based on refined PPG data.

Specifically, the health status information estimation device (100) can acquire multiple PPG data measured from multiple areas of the user's skin via multiple PPG sensors. In addition, the health status information estimation device (100) can acquire at least one refined PPG data from the multiple PPG data. In addition, the health status information estimation device (100) can estimate the user's health status information based on the at least one refined PPG data.

For example, the health status information estimation device (100) can estimate health status information related to blood pressure, heart rate, stress, and blood sugar based on refined PPG data.

Accordingly, the health status information estimation device (100) of the present invention can accurately estimate the user's real-time health status information based on refined data while using the non-invasive PPG measurement method. In particular, the health status information estimation device (100) of the present invention can maximize the user's convenience by estimating the user's blood pressure without a cuff (i.e., eliminating the pressure on the body), and the absence of the cuff's pressure can increase durability.

Hereinafter, an example of a health status information estimation device (100) based on refined PPG data will be described with reference to FIGS. 7 and 10.

In various embodiments, the health status information estimation device (100) may provide a web- or application-based service. However, the present invention is not limited thereto.

The health status information estimation device (100) may include any type of computer system or computer device, such as, for example, a microprocessor, a mainframe computer, a digital processor, a portable device, and a device controller, but is not limited thereto.

Below, a description of the hardware configuration of the health status information estimation device (100) will be provided with reference to FIG. 2.

Meanwhile, the user terminal (200) may be connected to the health status information estimation device (100) via a network (400) and may be a user terminal related to the health status information estimated by the health status information estimation device (100). That is, the user terminal (200) may include a terminal of a user wearing a wearable health status information estimation device (100) or a user who provides PPG data through a stationary health status information estimation device (100).

Here, the user terminal (200) may include, for example, various types of computer devices. For example, the user terminal (200) may refer to various terminal devices such as a smartphone, tablet PC, desktop, or laptop.

The user terminal (200) includes a display on at least a portion of the terminal, and may include an operating system for driving an application or extension program-based service provided from the health status information estimation device (100). For example, the user terminal (200) may be a smart phone, but is not limited thereto, and the user terminal (200) may include all types of handheld-based wireless communication devices such as navigation, PCS (Personal Communication System), GSM (Global System for Mobile communications), PDC (Personal Digital Cellular), PHS (Personal Handyphone System), PDA (Personal Digital Assistant), IMT (International Mobile Telecommunication)-2000, CDMA (Code Division Multiple Access)-2000, W-CDMA (W-Code Division Multiple Access), Wibro (Wireless Broadband Internet) terminals, smart pads, tablet PCs, etc., as a wireless communication device that ensures portability and mobility.

An external server (300) can be connected to a health status information estimation device (100) via a network (400), and the health status information estimation device (100) can transmit and receive various information/data necessary for measuring cardiovascular metabolic indices using a plurality of PPG sensors, and the health status information estimation device (100) can store and manage various information/data generated as the health status information estimation device (100) measures cardiovascular metabolic indices using a plurality of PPG sensors.

For example, the external server (300) may be a database server that stores information used in a health status information estimation method based on refined PPG data. As another example, the external server (300) may be a server that provides information used in a health status information estimation method based on refined PPG data.

A network (400) may refer to a connection structure that enables information exchange between each node, such as a computing device, multiple terminals, and servers. For example, the network (400) includes a local area network (LAN), a wide area network (WAN), the Internet (WWW), a wired and wireless data communication network, a telephone network, a wired and wireless television communication network, etc.

Wireless data communication networks include, but are not limited to, 3G, 4G, 5G, 3GPP (3rd Generation Partnership Project), 5GPP (5th Generation Partnership Project), LTE (Long Term Evolution), WIMAX (World Interoperability for Microwave Access), Wi-Fi, the Internet, LAN (Local Area Network), Wireless LAN (Wireless Local Area Network), WAN (Wide Area Network), PAN (Personal Area Network), RF (Radio Frequency), Bluetooth network, NFC (Near-Field Communication) network, satellite broadcasting network, analog broadcasting network, and DMB (Digital Multimedia Broadcasting) network.

Figure 2 is a hardware configuration diagram of a computing device according to one embodiment of the present invention.

Referring to FIG. 2, a health status information estimation device (100) according to an embodiment of the present invention may include one or more processors (110), a memory (120) for loading a computer program (151) executed by the processor (110), a bus (130), a communication interface (140), and a storage (150) for storing the computer program (151). Here, only components related to the embodiment of the present invention are illustrated in FIG. 2. Therefore, a person skilled in the art to which the present invention pertains may understand that other general components may be included in addition to the components illustrated in FIG. 2.

The processor (110) controls the overall operation of each component of the health status information estimation device (100). The processor (110) may be configured with one or more cores, and may include a processor for data analysis and deep learning, such as a central processing unit (CPU), a general purpose graphics processing unit (GPGPU), or a tensor processing unit (TPU) of a computing device. Alternatively, the processor may be configured to include any type of processor well known in the technical field of the present invention.

Additionally, the processor (110) can perform operations for at least one application or program for executing a method according to embodiments of the present invention, and the health status information estimation device (100) can have one or more processors.

In various embodiments, the processor (110) may further include a Random Access Memory (RAM, not shown) and a Read-Only Memory (ROM, not shown) that temporarily and/or permanently store signals (or data) processed within the processor (110). In addition, the processor (110) may be implemented in the form of a system on chip (SoC) that includes at least one of a graphics processing unit, RAM, and ROM.

The memory (120) stores various data, commands, and/or information. The memory (120) can load a computer program (151) from the storage (150) to execute methods/operations according to various embodiments of the present invention. When the computer program (151) is loaded into the memory (120), the processor (110) can perform the method/operation by executing one or more instructions constituting the computer program (151). The memory (120) may be implemented as a volatile memory such as RAM, but the technical scope of the present invention is not limited thereto.

The bus (130) provides a communication function between components of the health status information estimation device (100). The bus (130) can be implemented as various types of buses, such as an address bus, a data bus, and a control bus.

The communication interface (140) supports wired and wireless Internet communication of the health status information estimation device (100). Furthermore, the communication interface (140) may support various communication methods other than Internet communication. To this end, the communication interface (140) may be configured to include a communication module well known in the technical field of the present invention. In some embodiments, the communication interface (140) may be omitted.

In one embodiment, the communication interface (140) can transmit the user's health status information measured by the health status information estimation device (100) to the user terminal (200). Specifically, the communication interface (140) can transmit data to the user terminal (200) in real time via a wireless communication method such as Wi-Fi, Bluetooth, NFC, etc. For example, if the health status information estimation device (100) is implemented as a wearable device, the communication interface (140) can transmit the health status information to a user terminal such as a smartphone, so that the user can visually display and analyze the data through a dedicated application.

Additionally, the communication interface (140) can be connected to an external server (300) via wired communication to store data or transmit it to an analysis server. Through this, the communication interface (140) can monitor the user's long-term health data or share information with medical staff in real time, thereby supporting precise health management.

Storage (150) can non-temporarily store a computer program (151). When performing a process according to an embodiment of the present invention through a health status information estimation device (100), storage (150) can perform a method according to the disclosed embodiment or store various types of information necessary to provide a service.

Storage (150) may be configured to include non-volatile memory such as ROM (Read Only Memory), EPROM (Erasable Programmable ROM), EEPROM (Electrically Erasable Programmable ROM), flash memory, a hard disk, a removable disk, or any type of computer-readable recording medium well known in the art to which the present invention pertains.

The computer program (151) may include one or more instructions that cause the processor (110) to perform a method/operation according to various embodiments of the present invention when loaded into the memory (120). That is, the processor (110) may perform the method/operation according to various embodiments of the present invention by executing the one or more instructions.

In one embodiment, the computer program (151) may include one or more instructions for performing various methods associated with various tasks related to learning a neural network model.

A plurality of PPG sensors (160) are configured to be in contact with the user's skin and simultaneously measure PPG data from various parts of the body. The plurality of PPG sensors (160) are attached to various body parts, such as the user's wrist, forearm, finger, or ear, and can detect changes in blood flow in the corresponding parts to collect biosignals. The collected PPG data can be used to estimate health status information, such as the user's blood pressure, blood sugar, heart rate, and stress, after undergoing noise removal and signal refinement processes by the processor (110).

The present invention can improve reliability and accuracy compared to a single sensor by integrating data measured at multiple locations through multiple PPG sensors (160).

The indicator (170) is configured to provide information related to the power on/off status of the device, the device battery status, and health status information when the health status information estimation device (100) is implemented as a wearable device. The indicator (170) operates by outputting an LED or various colors, and can intuitively provide the status of the device with different colors depending on each status. For example, the indicator (170) outputs white when the power is on, red when the power is off, yellow when the battery is low, and blue while analyzing health status information, thereby providing a function that allows the user to grasp the operating status of the device at a glance.

When the health status information estimation device (100) is implemented as a stationary device, the display (180) can visually provide the user with health status information based on PPG data. The display (180) can provide the user with real-time measured information such as blood pressure, blood sugar, heart rate, and stress as immediate feedback, thereby providing the user with the ability to monitor their own condition in real time.

The display (180) can be implemented as a touchscreen or LED/LCD screen, and can intuitively visualize health status information using graphs, icons, numeric information, etc. This allows users to easily check data related to complex health conditions and, as needed, directly manipulate the data to view or check additional information.

The steps of a method or algorithm described in connection with an embodiment of the present invention may be implemented directly in hardware, implemented as a software module executed by hardware, or implemented by a combination thereof. The software module may reside in a random access memory (RAM), a read only memory (ROM), an erasable programmable ROM (EPROM), an electrically erasable programmable ROM (EEPROM), a flash memory, a hard disk, a removable disk, a CD-ROM, or any other form of computer-readable recording medium well known in the art to which the present invention pertains.

The components of the present invention may be implemented as programs (or applications) and stored on a medium to be executed in conjunction with a computer, which is hardware. The components of the present invention may be implemented as software programs or software elements. Similarly, the embodiments may be implemented in a programming or scripting language such as C, C++, Java, or an assembler, including various algorithms implemented as a combination of data structures, processes, routines, or other programming components. Functional aspects may be implemented as algorithms that are executed on one or more processors.

FIG. 3 and FIG. 4 are drawings for explaining an example of a wearable health status information estimation device according to one embodiment of the present invention.

Referring to FIG. 3, the outer side (101A) and inner side (101B) of the wearable health status information estimation device (101) are illustrated.

An indicator (170) is provided on the outer surface (101A) of a wearable health status information estimation device (101), which can visually provide information such as power status, battery status, or health status information to the user. Here, the indicator (170) can display status in various colors using an LED, allowing the user to intuitively grasp the current status of the device.

Meanwhile, a plurality of PPG sensors (160) may be provided on the inner side (101B) of the wearable health status information estimation device (101). The plurality of PPG sensors (160) are in close contact with the user's skin, and can collect PPG data in real time from various locations on the body. These PPG sensors detect changes in blood flow to measure biometric information such as heart rate and blood pressure, and by being located on the inner side, they can enable accurate measurements while in close contact with the body.

Specifically, the plurality of PPG sensors (160) may be designed to fit the user's wrist, forearm, or other various areas, and the sensor array may be appropriately positioned to enhance the accuracy of PPG data. For example, the plurality of sensors may be positioned to minimize signal interference or noise caused by user movement.

In addition, the distance between the multiple PPG sensors (160) can be appropriately adjusted to reduce measurement deviation according to the user's skin condition or movement, and accurate and consistent data can be collected through adjustment.

Referring to FIG. 4, an exploded perspective view of a wearable health status information estimation device (101) is illustrated.

As illustrated, the wearable health status information estimation device (101) may include, in addition to a plurality of PPG sensors, a fabric strap, an aluminum strap joint, silicon, and an on/off button.

The fabric strap is a strap portion of a wearable health status information estimation device (101). Made of fabric, it can provide a flexible wearing experience while softly contacting the user's skin. In addition, the fabric strap can be configured to be adjusted in length to fit the user's body.

The aluminum strap joint is a strap connection made of aluminum that secures the strap to the main body. Aluminum is a durable and lightweight material, allowing the device to maintain durability while minimizing weight. However, the strap joint material is not limited to aluminum.

Silicone can be configured to surround the body, PPG sensor, and circuitry. Silicone offers excellent flexibility and durability, preventing damage to the body, PPG sensor, and circuitry caused by sweat or moisture. Additionally, silicone can provide anti-slip properties, ensuring the device adheres closely to the skin.

The On/Off button is a power button for the wearable health status information estimation device (101), allowing the user to turn the device on and off. Furthermore, the On/Off button may be designed to be intuitively usable as a user interface.

According to various embodiments of the present invention, a wearable health status information estimation device (101) can be used to monitor the health status of a patient hospitalized in a hospital by having the device worn by the patient.

Specifically, the wearable health status information estimation device (101) can continuously measure various bio-signals, such as the patient's heart rate, blood pressure, stress level, and respiration rate, through a PPG sensor, and process these into refined PPG data to accurately monitor the patient's condition. In addition, the wearable health status information estimation device (101) can provide medical staff or guardians with information on the patient's health status by linking with a hospital server or a guardian terminal of the patient, thereby enabling a quick response to the patient's health status.

Therefore, the wearable health status information estimation device (101) can monitor the health status of a patient within a hospital and provide the information, thereby enabling immediate detection of an emergency situation occurring to the patient and enabling immediate medical action to be taken.

FIG. 5 and FIG. 6 are drawings for explaining an example of a stationary health status information estimation device according to one embodiment of the present invention.

Referring to FIG. 5, the stationary health status information estimation device (102) may include a display (180) and a PPG sensor (160).

The display (180) visually displays the user's health status information and can provide measurement results, such as blood pressure, heart rate, and stress level, through a graphical interface. This allows the user to check their health status in real time through the display. This display (180) can be implemented as a touchscreen, and the user can use it to operate desired functions or check data.

The PPG sensor (160) may be a key component of a device that measures PPG data by allowing a user to touch a body part to the device. The PPG sensor (160) detects changes in blood flow to collect biosignals, thereby providing accurate PPG data necessary for estimating health status.

For example, as illustrated in FIG. 6, the stationary health status information estimation device (102) can measure the user's PPG data when the user places his or her arm on the PPG sensor (160) portion of the stationary health status information estimation device (102). At this time, the PPG sensor (160) detects changes in blood flow in the arm to collect biosignals, and the stationary health status information estimation device (102) can estimate health status information based on these.

That is, a user can measure his or her health status in a non-contact manner by a simple action of placing his or her arm on a stationary health status information estimation device (102), and the measured health status can be displayed on a display (180).

In this way, the stationary health status information estimation device (102) of the present invention can be designed so that multiple users can easily measure and monitor their health status in public places, medical institutions, or health management centers, and can simultaneously increase user convenience and measurement accuracy.

FIGS. 7 to 10 are diagrams for explaining an example of a method for estimating health status information based on purified PPG data according to one embodiment of the present invention.

Referring to FIG. 7, the health status information estimation device (100) can obtain multiple PPG data measured in multiple areas of the user's skin through multiple PPG sensors (160) (S100).

For example, if the health status information estimation device (100) is implemented as a wearable device, a plurality of PPG sensors (160) can be attached to the user's body in the form of a smartwatch, wristband, or wearable patch. Here, each PPG sensor can be positioned at a different site, such as a different side of the wrist or arm, to measure PPG data at multiple locations simultaneously. In addition, each PPG sensor is composed of a light source and a light receiver, and can detect changes in light absorption according to blood flow to generate a pulse wave signal. Through this configuration, the health status information estimation device (100) can minimize signal distortion due to the user's movement or surrounding environment, and obtain more reliable biosignals.

For another example, if the health status information estimation device (100) is implemented as a stationary type, multiple PPG sensors (160) can be built into a medical device or healthcare kiosk and fixed at a specific location. In this case, the user can simultaneously measure PPG data in multiple areas by placing the forearm, fingers, or palm of the arm on the sensor area. Here, each PPG sensor can be placed at various locations, such as on different sides of the arm, to acquire multi-channel PPG data. In this way, the health status information estimation device (100) implemented as a stationary type can be utilized for group health monitoring or user health measurement in medical facilities.

According to various embodiments, the health status information estimation device (100) can perform preprocessing on a plurality of PPG data.

In one embodiment, data acquired from multiple PPG sensors may have temporal differences. Accordingly, the health status information estimation device (100) may synchronize the scale and time of multiple PPG data (e.g., assign a timestamp to the data from each sensor and align them temporally) prior to acquiring refined PPG data.

For example, data collected from each PPG sensor (160) may not be directly comparable due to slight time delays or differences in sampling rates between sensors. To address this, the health status information estimation device (100) can reconstruct the data by aligning the sampling rates of the sensors and establishing a common time reference. In this process, the health status information estimation device (100) can align the time axis by utilizing an interpolation technique or a dynamic time warping algorithm.

For example, the health status information estimation device (100) can align the time axis using a dynamic time warping (DTW) algorithm and reconstruct data by applying an interpolation method (e.g., linear interpolation, spline interpolation) as needed. In addition, the health status information estimation device (100) can perform resampling (upsampling or downsampling) when the sampling frequency is different.

In one embodiment, the PPG signal may have a baseline that fluctuates over time. Such baseline fluctuations may cause measurement errors, so the health status information estimation device (100) may pass a plurality of synchronized PPG data through a specific bandpass filter (e.g., a moving average filter or a high-pass filter) to remove signal components (i.e., baselines) corresponding to a specific frequency band.

For example, a typical PPG signal may include low-frequency components (e.g., baseline fluctuations of 0.5 Hz or less) and high-frequency noise (e.g., electrical noise of 40 Hz or more). To remove these, the health status information estimation device (100) may apply a bandpass filter that passes a frequency band between 0.5 Hz and 40 Hz. In addition, the health status information estimation device (100) may design a bandpass filter that passes a frequency band of a heart rate signal, and the order and cutoff frequency of the filter may be optimized according to the characteristics of the signal. Here, a Butterworth filter, a Chebyshev filter, a Bessel filter, etc. may be used in the filter design. In addition, the filter may include a Finite Impulse Response (FIR) filter or an Infinite Impulse Response (IIR) filter. In addition, the order and cutoff frequency of the filter may be adjusted according to the characteristics of the signal (e.g., the user's heart rate range, sampling frequency, signal-to-noise ratio, degree of motion artifact, etc.).

Meanwhile, the health status information estimation device (100) can apply a consistent baseline removal technique to all sensor data. Through this, the health status information estimation device (100) can estimate accurate health status information based on biosignals.

In one embodiment, the health status information estimation device (100) can remove linear trends of a plurality of PPG data.

For example, linear changes due to device drift or environmental changes that may be included in signals during long-term measurements can interfere with signal analysis. To compensate for this, the health status information estimation device (100) can estimate and remove linear or low-dimensional polynomial trends from the signal. For example, the health status information estimation device (100) can remove trends by calculating a linear regression line using the Least Squares Method and subtracting it from the original signal.

That is, the health status information estimation device (100) can remove low-frequency components by applying a high-pass filter to each PPG data and reduce noise through a moving average technique. In addition, the health status information estimation device (100) can compensate for scale differences between sensors through signal standardization (e.g., standardization to a mean of 0 and a standard deviation of 1) or normalization (e.g., Min-Max scaling, Z-score scaling, etc.), thereby increasing the accuracy of data fusion that can be performed in a later stage.

The health status information estimation device (100) can obtain at least one refined PPG data from a plurality of PPG data (S200).

In one embodiment, the health status information estimation device (100) can select at least one main PPG data from among a plurality of PPG data and obtain refined PPG data by removing noise from the data.

Specifically, referring to FIG. 8, the health status information estimation device (100) can determine at least one of the first PPG data having the highest quality among the plurality of PPG data, the second PPG data calculated based on the weighted average of the plurality of PPG data, and the third PPG data satisfying a specific measurement condition among the plurality of PPG data as the main PPG data (S211).

For example, the health status information estimation device (100) can evaluate the quality of each PPG data and select the first PPG data with the highest quality as the main PPG data. For example, the health status information estimation device (100) can calculate quality indicators for each PPG data, including a signal-to-noise ratio, an artifact ratio, and signal consistency.

Specifically, the health status information estimation device (100) can calculate a signal-to-noise ratio by comparing the energy of a valid signal and the energy of noise in each PPG data. Furthermore, the health status information estimation device (100) can determine that a higher signal-to-noise ratio indicates better signal quality.

Additionally, the health status information estimation device (100) can calculate the ratio of outlier data due to exercise or external interference. Furthermore, the health status information estimation device (100) can determine that a signal is more stable when the artifact ratio is lower.

Additionally, the health status information estimation device (100) can evaluate the stability of a signal based on the degree of change in the signal pattern over time.

And, the health status information estimation device (100) can select the first PPG data with the highest quality based on the signal-to-noise ratio (SNR), artifact ratio, and consistency of the signal. For example, assuming that there are three PPG data A (SNR = 20 dB, artifact ratio = 5%, consistency score = 90%), data B (SNR = 15 dB, artifact ratio = 10%, consistency score = 85%), and data C (SNR = 22 dB, artifact ratio = 3%, consistency score = 92%), the health status information estimation device (100) can comprehensively evaluate the quality indices of each data to calculate a quality score, and select data C with the highest quality score as the first PPG data.

Meanwhile, the health status information estimation device (100) may determine the first PPG data as the main PPG data if the quality of specific data among the entire PPG data is significantly superior to that of the remaining data (i.e., if the difference between the quality of the specific data and the quality of the remaining data exceeds a threshold value). For example, the health status information estimation device (100) may determine the first PPG data as the main PPG data if a clean signal with almost no noise is received from only one sensor.

In an additional embodiment, the health status information estimation device (100) comprehensively evaluates quality indicators for each data to calculate a quality score, and if there is data with a quality score below a threshold value, the data is excluded from the data for estimating health status information, thereby improving the accuracy and reliability of the estimation result.

In addition, the health status information estimation device (100) comprehensively evaluates quality indicators for each data to calculate a quality score, and then, if the quality score of each data is greater than a threshold value, assigns a weight corresponding to the quality score of each data and selects the data as the main PPG data to be refined through the weighted sum.

As another example, the health status information estimation device (100) can calculate second PPG data by weighting the average of multiple PPG data and determine this as main PPG data.

Specifically, the health status information estimation device (100) can calculate a weighted average using the quality indicator of each PPG data as a weight.

For example, the health status information estimation device (100) may assign a weight of 0.3 to data A, a weight of 0.2 to data B, and a weight of 0.5 to data C based on the quality score of each data in the preceding example. In addition, the health status information estimation device (100) may generate second PPG data by performing a weighted average of the PPG values at each time point using the weights of each data.

For example, assuming that each PPG data value at a specific point in time is data A: 0.8, data B: 0.75, and data C: 0.82, the health status information estimation device (100) can calculate the second PPG data value by reflecting the weight of each data. That is, the second PPG data value at a specific point in time is calculated as (0.8 * 0.3) + (0.75 * 0.2) + (0.82 * 0.5) = 0.806, and the health status information estimation device (100) can obtain the second PPG data by performing a weighted average for the entire time axis.

Meanwhile, the health status information estimation device (100) may select the second PPG data as the main PPG data when the quality of each PPG data is similar or the reliability of the individual data is low. Through this, the health status information estimation device (100) can reduce noise in the individual data and improve signal reliability through weighted averaging.

As another example, the health status information estimation device (100) can determine third PPG data that satisfies specific measurement conditions as main PPG data.

For example, the health status information estimation device (100) can select third PPG data based on specific conditions including the location of the sensor, the user's activity status, and environmental conditions.

For example, the health status information estimation device (100) can select data from a sensor located on the upper side among the sensors located on the upper and lower sides of the wrist as the third PPG data. In addition, the health status information estimation device (100) can select PPG data measured when the user is in a stable state (i.e., a resting state) as the third PPG data. In addition, the health status information estimation device (100) can select PPG data collected under conditions with little ambient light interference as the third PPG data.

Meanwhile, the health status information estimation device (100) may select the third PPG data as the main PPG data if the data collected under specific conditions is suitable for the analysis purpose. For example, when the health status information estimation device (100) requires data measured by the user in a stable state for blood pressure measurement, the health status information estimation device (100) may select the PPG data satisfying the corresponding conditions as the main PPG data.

In one embodiment, when the health status information estimation device (100) determines main PPG data, it can remove noise from the main PPG data to obtain at least one refined PPG data (S212).

Specifically, the health status information estimation device (100) can obtain refined PPG data by removing noise using other PPG data as a reference. Here, the other PPG data used as a reference may contain similar biosignal components as the main PPG data, but may have different noise patterns.

More specifically, the health status information estimation device (100) can adjust filter coefficients using an adaptive algorithm. For example, the health status information estimation device (100) can adjust filter coefficients based on the Least Mean Squares (LMS) algorithm or the Recursive Least Squares (RLS) algorithm. The LMS algorithm has low computational complexity and allows for real-time processing, making it suitable for wearable devices.

The health status information estimation device (100) can obtain refined PPG data by removing noise from the main PPG data based on the reference PPG data after adjusting the filter coefficient.

Specifically, the health status information estimation device (100) can generate a noise estimate from reference PPG data using an adaptive filter and remove it from the main PPG data. For example, if the main PPG data is measured from the upper part of the wrist and the reference PPG data is measured from the lower part of the wrist, the upper part of the wrist PPG data may be sensitive to movement and may contain many motion artifacts, but the lower part of the wrist PPG data may be relatively stable. Accordingly, the health status information estimation device (100) can use the lower part PPG data as a reference and remove motion noise from the upper part PPG data using an adaptive filter. Through this, the health status information estimation device (100) can obtain refined PPG data with minimized noise.

In various embodiments, the health status information estimation device (100) can obtain refined PPG data in a manner of removing noise by comparing independent components of each of a plurality of PPG data in step (S200).

Specifically, referring to FIG. 9, the health status information estimation device (100) can synchronize the scale and time of a plurality of PPG data (S221).

For example, the health status information estimation device (100) can align the sampling frequencies of each PPG data to the same sampling frequency through upsampling or downsampling if the sampling frequencies are different. In addition, the health status information estimation device (100) can align the time axes by applying a synchronization algorithm to compensate for time delays between sensors.

The health status information estimation device (100) can compare the signal components of each of a plurality of synchronized PPG data to compare the interrelationships at a specific point in time or frequency band (S222).

For example, the health status information estimation device (100) can calculate the similarity between two signals using a cross-correlation function and identify common biosignal components. In addition, the health status information estimation device (100) can compare the frequency spectrum of each signal through spectral analysis to find common frequency components, and can also evaluate mutual dependence by calculating covariance between signals through covariance analysis.

In addition, the health status information estimation device (100) can recognize at least one signal component that has a correlation with each other as a biosignal component, and obtain at least one refined PPG data by combining at least one signal component (S223).

For example, the health status information estimation device (100) can utilize independent component analysis (ICA) to separate independent signal components from multiple PPG data and select only those components corresponding to biosignals. Furthermore, the health status information estimation device (100) can remove noise by extracting only components corresponding to biosignals through principal component analysis (PCA).

For example, let's assume that a health status information estimation device (100) applies ICA to two PPG data and obtains two independent components s1 and s2. Here, component s1 has a signal concentrated in the frequency range of 0.5 Hz to 4 Hz, which can be determined to correspond to the frequency range of a typical heartbeat signal. Meanwhile, component s2 can be determined to correspond to high-frequency noise or motion artifact. In this case, the health status information estimation device (100) can select component s1 and reconstruct refined PPG data.

According to an additional embodiment of the present invention, the health status information estimation device (100) can obtain refined PPG data by removing noise based on an accelerometer.

Specifically, the health status information estimation device (100) can remove motion artifacts from a PPG signal by using an acceleration sensor as a reference signal.

For example, the health status information estimation device (100) can receive a PPG signal and a 3-axis acceleration signal as input and extract a frequency component corresponding to a biosignal by applying a 4th-order Butterworth bandpass filter in the range of 0.4 Hz to 4.0 Hz. Subsequently, the health status information estimation device (100) can perform adaptive noise cancellation using the acceleration signal of each axis as reference noise. Through this, the health status information estimation device (100) can remove artifacts caused by exercise and extract accurate biosignals.

Additionally, the health status information estimation device (100) can remove electrocardiogram (ECG) artifacts. Specifically, the health status information estimation device (100) can remove ECG artifacts by analyzing the correlation between the ECG signal and the PPG signal. For example, the health status information estimation device (100) can remove ECG artifacts from the PPG signal using heart rate information detected from the ECG signal.

Additionally, the health status information estimation device (100) can further remove residual noise remaining in the PPG signal by adjusting the weight vector of the Finite Impulse Response (FIR) filter. Here, the weight vector can be adjusted in a direction that minimizes the error signal in real time using the Least Mean Squares (LMS) algorithm. Through this, the weight vector is continuously updated to reduce the difference between the input signal and the reference signal, thereby improving the accuracy of noise removal.

Thereafter, the health status information estimation device (100) can track and correct the heart rate frequency, and estimate the final heart rate based on this. In addition, the health status information estimation device (100) can remove remaining noise components from the PPG signal by performing notch filtering using the estimated heart rate frequency, thereby obtaining a refined PPG signal.

According to a further embodiment of the present invention, the health status information estimation device (100) can obtain refined PPG data by analyzing two or more PPG data in multiple resolutions through wavelet transform to remove noise. Here, wavelet transform is a technique that can simultaneously analyze a signal in the time and frequency domains, and can be used to detect and remove changes or noise occurring in short intervals by examining specific frequency components of the signal in various resolutions.

Specifically, the health status information estimation device (100) can perform wavelet decomposition on two PPG signals and compare wavelet coefficients of each signal to identify noise components. Thereafter, the health status information estimation device (100) can perform threshold processing on coefficients containing noise among the coefficients obtained through wavelet transformation. Here, the threshold processing is performed in a manner of completely removing or reducing small coefficients that mainly contain noise. That is, wavelet coefficients with small values below the threshold value are regarded as noise and are removed or reduced, and important coefficients above the threshold value can be maintained.

In addition, the health status information estimation device (100) can restore PPG data using modified wavelet coefficients after processing noise coefficients. Here, the restored PPG data can be used as refined PPG data, including only clean signals with noise removed.

According to an additional embodiment of the present invention, the health status information estimation device (100) can obtain refined PPG data by removing noise by resolving temporal mismatch between two or more PPG sensors.

Specifically, the health status information estimation device (100) can synchronize data collection through hardware-based shared clocks or simultaneous triggering, or can perform synchronization by assigning timestamps through software. Thereafter, the health status information estimation device (100) can calculate a correlation function between the two signals to find the delay time, compensate for the delay time, and align the heart rate peaks. Furthermore, the health status information estimation device (100) can reduce noise by matching the sampling rate of each PPG data and applying an interpolation technique to align the signals. PPG data with reduced noise in this manner can be used as refined PPG data.

According to an additional embodiment of the present invention, the health status information estimation device (100) can obtain refined PPG data by removing common mode noise using the difference between two PPG signals.

Specifically, the health status information estimation device (100) can collect simultaneous signals from two PPG sensors and synchronize the signals using a correlation function or peak alignment. Thereafter, the health status information estimation device (100) calculates the difference between the two signals to generate a differential signal, thereby removing common mode noise. Finally, the health status information estimation device (100) can extract a PPG signal from which noise has been removed by applying a low-pass filter and smoothing technique, and reconstruct it into refined PPG data.

According to an additional embodiment of the present invention, the health status information estimation device (100) can obtain refined PPG data by removing complex noise from PPG data using a machine learning algorithm.

Specifically, the health status information estimation device (100) can collect and label signals considering noise levels, and preprocess data through normalization and feature extraction. Thereafter, the health status information estimation device (100) can design a noise removal model by selecting a suitable model, such as an autoencoder, CNN, RNN, etc., and can define a loss function to learn model parameters. The performance of the learned model is evaluated using the improvement in signal-to-noise ratio (SNR) as an evaluation index, and the health status information estimation device (100) can perform noise removal on PPG data by applying the learned model. PPG data with noise removed in this way can be used as refined PPG data.

According to an additional embodiment of the present invention, the health status information estimation device (100) can obtain refined PPG data by improving signal quality and removing noise through sensor fusion (i.e., combining two or more PPG data).

Specifically, the health status information estimation device (100) can perform simultaneous data collection and synchronization from multiple sensors, and combine two or more PPG data based on a fusion algorithm such as Bayesian fusion, Kalman filter, and weighted average. In addition, the health status information estimation device (100) can perform additional filtering, such as smoothing and low-pass filtering, on the combined data to remove residual noise. For example, the health status information estimation device (100) can improve the continuity of the signal by alleviating irregular fluctuations of the signal through smoothing. In addition, the health status information estimation device (100) can extract accurate biosignals by removing high-frequency noise through low-pass filtering. Through this, the health status information estimation device (100) can extract accurate biosignals from the combined data, and can use the extracted biosignals as refined PPG data.

In this way, the health status information estimation device (100) can more accurately estimate the user's health status information by effectively removing noise from multiple PPG data and obtaining refined PPG data through various methods.

According to an additional embodiment of the present invention, the health status information estimation device (100) can obtain a plurality of refined PPG data through the various methods described above and perform additional signal processing based on the plurality of refined PPG data.

Specifically, the health status information estimation device (100) can acquire multiple refined PPG data through methods such as accelerometer-based adaptive noise cancellation, multi-resolution analysis through wavelet transform, and noise removal using a machine learning model. In this case, the health status information estimation device (100) can synchronize the time axis of each of the multiple refined PPG data and calculate the average value of the signal components for each time to reconstruct the refined PPG data. Through this, the health status information estimation device (100) can estimate health status information, which will be described later, based on a more accurate biosignal.

For example, the health status information estimation device (100) can perform temporal synchronization between PPG sensors, align the time axes of two or more refined PPG data, and calculate the average value of the PPG signal components for each time interval to generate final refined PPG data with noise removed. Through this, the health status information estimation device (100) can more accurately provide various health status information, such as the user's heart rate, blood pressure, and stress level.

According to a further embodiment of the present invention, the health status information estimation device (100) can obtain refined PPG data by combining filtering techniques. For example, the health status information estimation device (100) can obtain refined PPG data by removing baseline drift with a high-pass filter and removing high-frequency noise with a low-pass filter (i.e., a combination of a high-pass filter and a low-pass filter). In this case, the health status information estimation device (100) can obtain optimal performance by adjusting the order and cutoff frequency of each filter. For example, the health status information estimation device (100) can use various filter types such as Butterworth, Chebyshev, and Bessel, and the design of the filter can be optimized according to the characteristics of the signal.

For another example, the health status information estimation device (100) may obtain refined PPG data by using a bandpass filter that passes a frequency band of a heart rate signal and adding a notch filter to remove noise of a specific frequency caused by power noise or muscle activity (i.e., a combination of a bandpass filter and a notch filter). Here, the center frequency of the notch filter may be adjusted to match the frequency of the noise.

As another example, the health status information estimation device (100) can obtain refined PPG data by removing noise in the time-frequency domain through wavelet transform and adding adaptive filtering to remove residual noise (i.e., a combination of wavelet filtering and adaptive filtering). Here, wavelet transform can remove abnormal signal patterns, and adaptive filtering can remove noise that varies over time.

According to an additional embodiment of the present invention, the health status information estimation device (100) can obtain refined PPG data by combining a time domain technique and a frequency domain technique.

For example, the health status information estimation device (100) can obtain refined PPG data by removing simple noise through moving average filtering and analyzing the signal in the frequency domain through FFT (Fast Fourier Transform) to remove noise in a specific frequency band (i.e., a combination of moving average filtering and FFT). Here, the health status information estimation device (100) can identify the frequency characteristics of the noise through the FFT analysis results and utilize them to design a notch filter or a bandstop filter.

As another example, the health status information estimation device (100) can extract both time domain features (e.g., heartbeat interval, pulse amplitude, rise time, etc.) and frequency domain features (e.g., heart rate, heart rate variability, frequency band energy, etc.) of the PPG signal and input them into a machine learning model to obtain refined PPG data (i.e., a combination of time domain feature extraction and frequency domain feature extraction). In this case, the health status information estimation device (100) can perform more accurate noise removal and biosignal estimation by combining information in the time domain and frequency domain.

According to an additional embodiment of the present invention, the health status information estimation device (100) can obtain refined PPG data by combining AI techniques and existing signal processing techniques.

For example, the health status information estimation device (100) can obtain refined PPG data by removing residual noise using a deep learning model (e.g., CNN, RNN, autoencoder) after undergoing a preprocessing process such as conventional filtering or wavelet transform. In this case, by reducing noise through the preprocessing process, the learning efficiency of the deep learning model can be increased, and the model performance can be improved.

For another example, the health status information estimation device (100) can obtain refined PPG data through multi-PPG signal fusion and AI-based noise removal. That is, the health status information estimation device (100) can first fuse multiple PPG signals using a statistical method or other signal processing technique, and then apply an AI-based noise removal model. Through this fusion process, the health status information estimation device (100) can increase signal reliability and improve the quality of input data for the AI model.

According to an additional embodiment of the present invention, the health status information estimation device (100) can obtain refined PPG data through fusion with other sensor data.

For example, the health status information estimation device (100) can detect the user's movement using accelerometer data measured by an accelerometer sensor, and obtain refined PPG data by removing movement artifacts from the PPG sensor data. In addition, the health status information estimation device (100) can perform adaptive noise cancellation using the accelerometer data as a reference signal.

As another example, the health status information estimation device (100) can remove electrocardiogram artifacts from PPG sensor data using ECG data measured by the ECG sensor. That is, the health status information estimation device (100) can identify and remove artifacts by analyzing the correlation between the ECG signal and the PPG signal.

Referring again to FIG. 7, when the health status information estimation device (100) acquires at least one refined PPG data, it can estimate the user's health status information based on the at least one refined PPG data (S300).

Referring to FIG. 10, the health status information estimation device (100) can obtain arterial blood pressure data based on at least one refined PPG data (S301).

Specifically, the health status information estimation device (100) can obtain arterial blood pressure data by inputting two or more refined PPG data into an arterial blood pressure conversion algorithm.

In the present invention, the arterial blood pressure conversion algorithm may include an algorithm that estimates pulse wave transit time based on a time difference between two or more refined PPG data, calculates pulse wave velocity based on the pulse wave transit time, and calculates arterial blood pressure data based on the pulse wave velocity and the distance between PPG sensors corresponding to each of the two or more refined PPG data.

That is, the health status information estimation device (100) can estimate the pulse transmission time by measuring the difference in the arrival time of the pulse wave in two refined PPG data. For example, the health status information estimation device (100) can identify characteristic points (e.g., rising point, peak, falling point, etc.) of the PPG signal collected from two PPG sensors attached to the upper and lower sides of the wrist, and calculate the time difference between them. In addition, since the health status information estimation device (100) knows the physical distance between the sensors (e.g., straight-line distance between the two sensors), it can calculate (pulse wave velocity = distance between sensors / pulse wave transmission time).

Therefore, the health status information estimation device (100) can calculate arterial blood pressure data by applying a correlation model between pulse wave velocity and arterial blood pressure that has been built in advance.

Meanwhile, the health status information estimation device (100) can estimate arterial elasticity and stiffness based on the calculated pulse wave velocity, which may be directly related to arterial blood pressure. For example, a faster pulse wave velocity indicates a stiffer artery, which may act as a risk factor for hypertension. Such information may be provided to the user terminal (200).

The health status information estimation device (100) can estimate health status information including blood pressure based on arterial blood pressure data (S302).

Specifically, the health status information estimation device (100) can obtain systolic blood pressure and diastolic blood pressure by inputting arterial blood pressure data into a pre-learned blood pressure estimation model.

In the present invention, the blood pressure estimation model can extract features related to pulse wave velocity and pulse transit time from arterial blood pressure data. In this case, the health status information estimation device (100) can utilize machine learning algorithms such as multiple regression analysis, support vector machines, and artificial neural networks to predict systolic blood pressure based on the extracted features. Here, pulse wave velocity has a strong correlation with systolic blood pressure and can therefore be used as a key input variable.

Additionally, the health status information estimation device (100) can predict diastolic blood pressure by combining features related to pulse transit time with the user's biometric information (e.g., age, gender, height, weight, etc.). Here, pulse transit time is related to blood vessel elasticity and may also be correlated with diastolic blood pressure. Therefore, the health status information estimation device (100) can calculate a more accurate diastolic blood pressure value by combining the individual's biometric information and pulse transit time.

That is, the health status information estimation device (100) can estimate systolic and diastolic blood pressure by analyzing the characteristic patterns of pulse wave velocity and pulse wave transit time extracted from arterial blood pressure data. In this process, the health status information estimation device (100) can use the following mathematical models or algorithms.

The health status information estimation device (100) can assess the user's current blood pressure status by synthesizing predicted systolic and diastolic blood pressures and generate health status information. This health status information can be provided to the user via a user terminal (200) or display (180) and can be utilized for long-term health management or consultation with a medical professional.

Accordingly, the health status information estimation device (100) can provide a function to monitor the user's blood pressure non-invasively and in real time, thereby enabling early detection and management of cardiovascular diseases such as hypertension.

According to various embodiments of the present invention, the health status information estimation device (100) can estimate health status information based on a plurality of PPG data before being refined.

Specifically, the health status information estimation device (100) can obtain at least one numerical value corresponding to each item included in the health status information based on a plurality of PPG data before being refined. In addition, the health status information estimation device (100) can integrate the at least one numerical value to determine the final numerical value of each item included in the health status information. In addition, the health status information estimation device (100) can generate health status information including the final numerical value for each item.

For example, the health status information estimation device (100) can calculate a heart rate of 72 bpm from the first PPG data, a heart rate of 74 bpm from the second PPG data, and a heart rate of 73 bpm from the third PPG data. At this time, the health status information estimation device (100) can integrate these numerical values to determine a final numerical value of the heart rate.

The health status information estimation device (100) may use a weighted average, median, or statistical method to integrate at least one numerical value.

For example, the health status information estimation device (100) can calculate the final heart rate by weighting the previously calculated heart rate values. For example, the final heart rate can be calculated by the calculation formula of '(72*w1)+(74*Хw2)+(73*Хw3) / w1+w2+w3'. Here, w1, w2, and w3 may be weights based on the quality or reliability of each PPG data. Specifically, the health status information estimation device (100) can evaluate the signal quality index of each PPG data and assign weights. For example, the health status information estimation device (100) can evaluate the quality of each PPG data based on the signal-to-noise ratio, the artifact ratio, and the consistency of the signal.

For another example, the health status information estimation device (100) can calculate individual blood pressure values from each PPG data and then integrate them to determine a final blood pressure value when estimating blood pressure. That is, the health status information estimation device (100) can calculate pulse transit time (PTT) or pulse wave velocity (PWV) from each PPG data and input these values into a pre-trained model to obtain individual systolic and diastolic blood pressure values. Thereafter, the health status information estimation device (100) can integrate these values to calculate a final blood pressure value.

Through this, the health status information estimation device (100) can effectively utilize multiple unrefined PPG data to estimate reliable health status information.

According to various embodiments of the present invention, the health status information estimation device (100) can estimate blood sugar based on at least one refined PPG data.

Specifically, the health status information estimation device (100) can predict blood sugar levels based on temporal changes and amplitude patterns of blood flow from refined PPG data. To this end, the health status information estimation device (100) can extract various features from the time and frequency domains of the PPG signal.

For example, the health status information estimation device (100) can extract temporal features such as rise time, fall time, pulse interval, and amplitude variation from the waveform of the PPG signal. These features are closely related to changes in blood flow and may vary depending on blood sugar levels.

In addition, the health status information estimation device (100) can analyze the frequency of the PPG signal to extract frequency features such as power spectral density and frequency band energy. These are related to changes in blood viscosity or oxygen saturation, and may be related to blood sugar levels.

The health status information estimation device (100) can utilize the various features extracted in this manner to determine a correlation with blood sugar levels. Specifically, the health status information estimation device (100) can use the extracted features as input to a machine learning model to predict blood sugar levels. In this case, the health status information estimation device (100) can utilize a regression model or a deep learning-based neural network model that has previously learned the relationship between blood sugar levels and PPG signal features.

For example, the health status information estimation device (100) can detect amplitude increases and waveform change patterns in the user's PPG signal after a meal. For example, if blood sugar levels rise after a meal, changes in blood flow dynamics occur, which affect the amplitude and shape of the PPG signal. The health status information estimation device (100) can quantitatively measure these changes and estimate blood sugar levels based on their correlation with blood sugar fluctuation patterns.

Additionally, the health status information estimation device (100) can perform a personalized calibration procedure to improve the model's accuracy for each user. This can be accomplished by collecting the user's actual blood sugar measurements over a period of time to retrain the model or by applying a calibration coefficient. This can improve prediction accuracy by reflecting variability based on individual physiological characteristics.

In this way, the health status information estimation device (100) of the present invention can estimate blood sugar levels in real time through processes such as temporal feature extraction, frequency analysis, and application of a machine learning model based on refined PPG data. Through this, the health status information estimation device (100) enhances the user's convenience and enables early detection of abnormal signs, enabling appropriate responses.

According to various embodiments of the present invention, the health status information estimation device (100) can estimate a stress level based on at least one refined PPG data.

Specifically, the health status information estimation device (100) can extract changes in heart rate variability (HRV) and blood flow patterns from refined PPG data to predict stress levels. Here, stress causes changes in the balance of the autonomic nervous system, which in turn affects heart rate variability, which manifests as specific characteristics in the PPG signal.

More specifically, the health status information estimation device (100) can calculate a pulse interval or a heartbeat interval (RR interval) from a PPG signal and, based on this, produce time domain, frequency domain, and nonlinear indices.

The health status information estimation device (100) can evaluate heart rate variability by calculating the average heart rate, the standard deviation of heartbeat intervals (SDNN), and the root mean square of the difference between adjacent heartbeat intervals (RMSSD) based on the time domain of refined PPG data. Through this, the health status information estimation device (100) can detect changes in the activity of the autonomic nervous system through the variability of heartbeat intervals.

In addition, the health status information estimation device (100) can calculate the power spectral density (PSD) based on the frequency domain of the refined PPG data to obtain the power of the low frequency (LF, 0.04 to 0.15 Hz) component and the high frequency (HF, 0.15 to 0.4 Hz) component. In addition, the health status information estimation device (100) can evaluate the balance between the sympathetic nervous system and the parasympathetic nervous system by calculating the LF/HF ratio. Here, in a stressful situation, the LF component may increase and the HF component may decrease, thereby increasing the LF/HF ratio.

Additionally, the health status information estimation device (100) can evaluate the complexity and randomness of heart rate variability by calculating indices such as fractal dimension and entropy based on nonlinear dynamic analysis. Through this, the health status information estimation device (100) can detect changes in the complexity of heart rate patterns due to stress.

The health status information estimation device (100) can predict stress levels using a pre-trained machine learning model, taking as input various feature values extracted using the aforementioned methods. The machine learning model may be a regression model or classification model built based on the correlation between stress levels and PPG signal features. For example, the health status information estimation device (100) can calculate stress levels by applying algorithms such as support vector machines, artificial neural networks, and decision trees.

Through this, the health status information estimation device (100) can accurately estimate the user's real-time stress level, which can be utilized for stress management and mental health promotion.

According to various embodiments of the present invention, the health status information estimation device (100) can estimate a heart rate based on at least one refined PPG data.

Specifically, the health status information estimation device (100) can detect heartbeat cycles from refined PPG signals and calculate real-time heart rate. For example, the health status information estimation device (100) can use a peak detection algorithm of the PPG signal to calculate the interval between each heartbeat cycle and estimate heartbeats per minute (BPM) based on this interval.

More specifically, the health status information estimation device (100) can estimate the heart rate by identifying the maximum value (peak) of the pulse wave generated from the PPG signal and calculating the interval (R-R interval) between consecutive peaks. Additionally, the health status information estimation device (100) can minimize error detection through noise removal filtering and calculate an accurate heart rate.

For example, the health status information estimation device (100) can accurately estimate the user's heart rate per minute through real-time peak detection of the PPG signal and provide it to the user in real time.

In addition to blood pressure, blood sugar, stress levels, and heart rate, various health status information, including sleep quality and respiration rate, can be estimated and provided to users. This allows users to monitor and manage their comprehensive health in real time based on accurate biosignal-based information.

According to various embodiments of the present invention, the health status information estimation device (100) can measure cardiovascular metabolic indicators using a multi-PPG measurement technique.

The multi-PPG data of the present invention can be measured by a wearable device equipped with multiple PPG sensors.

In one embodiment, the multi-PPG sensor can measure multiple PPG data (i.e., multi-PPG data) by measuring photoplethysmography (PPC) in multiple regions of the user's body. This multiple PPG data can be corrected to produce higher-quality data with greater accuracy through a process of eliminating deviations and errors. Furthermore, by measuring PPC in multiple regions of the user's body, the multi-PPG sensor can minimize the impact of environmental variables, such as exercise, low-temperature environments, and high-temperature environments.

The wearable device transmits multiple PPG data to a user terminal (200), and the user terminal (200) that receives the data can provide the multiple PPG data to a health status information estimation device (100).

A health status information estimation device (100) can analyze multiple PPG data acquired from a user terminal (200) to measure cardiovascular metabolic indicators including blood sugar and blood pressure of the user, and provide the same to the user terminal (200).

The operations of the health status information estimation device (100) to be described later analyzing multiple PPG data and measuring cardiovascular metabolic indicators may also be performed independently in the user terminal (200).

That is, in this specification, the health status information estimation device (100) may also mean a user terminal (200). In this case, the health status information estimation device (100) may obtain multiple PPG data from a wearable device (210).

According to one embodiment of the present invention, the health status information estimation device (100) can obtain multiple PPG data measured simultaneously in multiple areas through multiple PPG (Photoplethysmography) sensors.

Specifically, the health status information estimation device (100) can collect PPG data in real time from each sensor and integrate the collected data to generate multi-region PPG data. For example, the health status information estimation device (100) can obtain multiple PPG data collected through PPG sensors attached to various locations such as the wrist, forearm, and finger. As another example, the health status information estimation device (100) can obtain multiple PPG data collected through PPG sensors in multiple regions of each of the wrist, forearm, and finger.

In various embodiments, the health status information estimation device (100) can obtain high-quality PPG data with deviations and errors removed by refining the multi-PPG data through a pre-learned AI-based data refining model when acquiring multi-PPG data.

In one embodiment, the data cleaning model may include an input layer for receiving time series data for each region of multiple PPG data as input, an intermediate layer including a deep neural network (DNN), a 1D CNN (one-dimensional convolutional neural network), and a long short-term memory (LSTM) network composed of multiple layers to extract temporal features and learn patterns, and an output layer for outputting cleaned high-quality PPG data and generating a signal with deviations and errors removed.

The health status information estimation device (100) can collect multiple PPG data measured in various environments (exercise, rest, low temperature, high temperature) to train a data refinement model. Here, the collected data can be converted from raw data to high-quality refined data through preprocessing processes such as noise removal and signal amplification. Furthermore, the preprocessed data can be input into a model comprised of various networks. This model can learn the time-series characteristics of PPG data, remove deviations and errors, and generate high-quality PPG data.

The health status information estimation device (100) uses Mean Squared Error (MSE) as a loss function to minimize the error between the predicted value and the actual value in order to optimize the performance of the model, and can adjust the learning rate and update the parameters through the Adam Optimizer.

According to one embodiment of the present invention, when the health status information estimation device (100) acquires multiple PPG data, it can acquire ABP (Arterial blood pressure) data based on the multiple PPG data.

Specifically, the health status information estimation device (100) can input high-quality PPG data into a pre-learned AI-based ABP prediction model, and obtain ABP data from the ABP prediction model.

For example, the health status information estimation device (100) can analyze changes in heart rate to derive ABP data, thereby predicting an accurate blood pressure value.

In one embodiment, the ABP prediction model can be constructed by combining a multilayer perceptron (MLP) and an LSTM network. This model analyzes the time-series characteristics of PPG data to generate ABP data. During the training process, PPG and ABP data collected from various environments can be used to implement a highly accurate prediction model.

Meanwhile, the health status information estimation device (100) can estimate cardiovascular metabolic indicators including blood sugar and blood pressure based on at least one of multiple PPG data and ABP data.

Specifically, the health status information estimation device (100) can estimate blood sugar level based on high-quality PPG data.

For example, the health status information estimation device (100) can predict blood sugar levels by analyzing periodic fluctuations in PPG signals. Additionally, the health status information estimation device (100) can obtain more accurate prediction values by analyzing other biosignals (e.g., ECG, skin temperature, etc.) in addition to PPG to more accurately predict blood sugar levels.

Additionally, the health status information estimation device (100) can estimate blood pressure corresponding to each systolic and diastolic phase based on ABP data.

For example, a health status information estimation device (100) can estimate blood pressure corresponding to each systolic and diastolic phase using a blood pressure data generation model based on a U-Net network. Here, the U-Net network is a convolutional neural network (CNN) structure mainly used in image processing, and is advantageous in recognizing and predicting complex patterns.

For example, the health status information estimation device (100) can perform a process of converting PPG data into ABP data by utilizing the encoder and decoder structures of U-Net. The encoder part extracts features of the PPG data, and the decoder part generates ABP data based on the extracted features to predict systolic and diastolic blood pressure.

Additionally, the health status information estimation device (100) can improve performance by applying GAN-based models such as Pix2Pix, CycleGAN, and Diffusion model.

For example, the health status information estimation device (100) may use the Pix2Pix model, an image-to-image conversion method that receives PPG data as input and generates ABP data. Pix2Pix is a conditional GAN that generates a target output based on a conditional input, and can generate ABP data from PPG data.

In addition, based on CycleGAN, which is suitable for learning transformation between two domains, the health status information estimation device (100) learns transformation between PPG data and ABP data, and can perform more sophisticated blood pressure prediction through mapping between each domain.

In addition, the health status information estimation device (100) based on a diffusion model that generates high-resolution output by modeling the data generation process as a normalized stochastic process can achieve high accuracy in the process of generating ABP data from PPG data.

For another example, the health status information estimation device (100) can monitor blood pressure changes in real time based on ABP data or predict systolic and diastolic blood pressure by analyzing the highest and lowest points in the ABP data.

For example, the health status information estimation device (100) can analyze the periodic pattern of ABP data to identify the highest point (systolic blood pressure) and lowest point (diastolic blood pressure) within each cycle. Furthermore, the health status information estimation device (100) can apply a peak detection algorithm to detect peaks within the data. Through this, the health status information estimation device (100) can accurately identify the highest and lowest points of each cycle.

In this process, the health status information estimation device (100) can statistically analyze ABP data from multiple cycles to calculate average systolic blood pressure and average diastolic blood pressure. Furthermore, the health status information estimation device (100) can perform a filtering process to remove outlier data to increase data reliability.

Therefore, the health status information estimation device (100) of the present invention can provide users with accurate and reliable cardiovascular metabolic indicators through the aforementioned processes. This can help users more efficiently manage their health status and recognize early warning signals to take appropriate action.

In an additional embodiment, the health status information estimation device (100) can estimate a user's stress level based on multiple PPG data. Here, the stress level can be calculated by evaluating variability through frequency analysis of the PPG data and calculating an index related to heart rate variability (HRV).

Specifically, the health status information estimation device (100) can help manage stress by analyzing the user's heart rate variability to monitor the user's stress level in real time and notifying the user of the same.

In another additional embodiment, the health status information estimation device (100) can evaluate the quality of a user's sleep using multiple PPG data.

Specifically, the health status information estimation device (100) collects PPG data collected during the user's sleep from a wearable device (210) worn by the user, and analyzes the data to identify sleep stages (e.g., light sleep, deep sleep, REM sleep). This allows the user to recognize his or her sleep patterns.

In another additional embodiment, the health status information estimation device (100) can utilize multiple PPG data to evaluate the recovery status after exercise.

Specifically, the health status information estimation device (100) analyzes PPG data during the post-exercise recovery phase to calculate a heart rate recovery rate, thereby enabling real-time monitoring of the user's recovery status. This information can help the user adjust their exercise intensity and recovery status to achieve optimal exercise effects.

In this way, various embodiments of the present invention can utilize multiple PPG data to provide a method for monitoring and managing various health indicators in real time, including cardiovascular metabolic indicators, stress levels, sleep quality, and post-exercise recovery status. This can provide users with a comprehensive health management solution, contributing to a better quality of life.

While the embodiments of the present invention have been described above with reference to the attached drawings, those skilled in the art will appreciate that the present invention can be implemented in other specific forms without altering the technical concept or essential features thereof. Therefore, the embodiments described above should be understood to be illustrative in all respects and not restrictive.

Claims (10)

A method performed by a health status information estimation device including a plurality of PPG (Photoplethysmography) sensors,
A step of acquiring multiple PPG data measured from multiple areas of a user's skin through multiple PPG sensors;
A step of obtaining at least one refined PPG data from the plurality of PPG data; and
A step of estimating health status information of the user based on at least one purified PPG data;
Including,
The step of obtaining at least one refined PPG data from the plurality of PPG data comprises:
A step of determining at least one of the first PPG data having the highest quality among the plurality of PPG data, the second PPG data calculated based on the weighted average of the plurality of PPG data, and the third PPG data satisfying a specific measurement condition among the plurality of PPG data as the main PPG data; and
A step of removing noise from the main PPG data to obtain at least one refined PPG data;
The step of determining with the above main PPG data is:
A step of calculating quality indicators including signal-to-noise ratio (SNR), artifact rate, and signal consistency;
A step of calculating a quality score of each of the plurality of PPG data based on the above quality indicator;
A step of acquiring data having the highest quality score as the first PPG data;
A step of obtaining the second PPG data by weighting and averaging the weights based on the quality scores of the plurality of PPG data to the values of the plurality of PPG data at any point in time;
A step of acquiring the third PPG data based on a specific condition for at least one of the plurality of PPG sensors, the user's activity status, and environmental conditions for acquiring the plurality of PPG data; and
A step of preferentially applying the first PPG data among the acquired first PPG data to the third PPG data as the main PPG data; including;
The step of obtaining the above refined PPG data is:
A step of removing noise from the main PPG data obtained based on any other PPG data containing a biosignal component similar to the main PPG data,
A method for estimating health status information based on refined PPG data.
In the first paragraph,
The above method,
A step of synchronizing the scale and time of the plurality of PPG data prior to acquiring the refined PPG data;
A step of passing the synchronized plurality of PPG data through a specific bandpass filter to remove a signal component corresponding to a specific frequency band; and
A step of removing a linear trend of the plurality of PPG data;
including more,
A method for estimating health status information based on refined PPG data.
delete In the first paragraph,
The step of obtaining at least one refined PPG data from the plurality of PPG data comprises:
A step of synchronizing the scale and time of the plurality of PPG data;
A step of comparing the signal components of each of the synchronized plurality of PPG data to compare the mutual correlation at a specific time point or frequency band; and
A step of recognizing at least one signal component having the above mutual correlation as a biosignal component, and combining the at least one signal component to obtain the at least one refined PPG data;
including,
A method for estimating health status information based on refined PPG data.
In the first paragraph,
The step of estimating the health status information of the user based on at least one purified PPG data is as follows:
A step of obtaining arterial blood pressure data based on at least one purified PPG data; and
A step of estimating the health status information including blood pressure based on the arterial blood pressure data;
including,
A method for estimating health status information based on refined PPG data.
In paragraph 5,
The step of obtaining arterial blood pressure data based on at least one purified PPG data is as follows:
A step of inputting two or more refined PPG data into an arterial blood pressure conversion algorithm to obtain the arterial blood pressure data;
Including,
The above arterial blood pressure conversion algorithm is,
Estimating the pulse transit time based on the time difference between the two or more refined PPG data,
The pulse wave velocity is calculated based on the pulse wave transmission time,
Calculating the arterial blood pressure data based on the pulse wave velocity and the distance between the PPG sensors corresponding to each of the two or more refined PPG data,
A method for estimating health status information based on refined PPG data.
In paragraph 6,
The step of estimating the health status information including blood pressure based on the arterial blood pressure data is as follows:
A step of obtaining the blood pressure by inputting the arterial blood pressure data into a pre-learned blood pressure estimation model;
Including,
The above blood pressure estimation model is,
Extracting features for the pulse wave velocity and the pulse wave transit time from the above arterial blood pressure data,
Predicting systolic blood pressure based on the characteristics of the pulse wave velocity and pulse wave transit time,
Predicting diastolic blood pressure based on the characteristics of the pulse transmission time and the user's biometric information.
A method for estimating health status information based on refined PPG data.
In the first paragraph,
The step of estimating the health status information of the user based on at least one purified PPG data is as follows:
A step of obtaining at least one numerical value corresponding to each item included in the health status information based on at least one refined PPG data;
A step of integrating at least one numerical value to determine a final numerical value for each item included in the health status information; and
A step of generating the health status information including the final numerical value for each of the above items;
including more,
A method for estimating health status information based on refined PPG data.
In a health status information estimation device,
Multiple PPG (Photoplethysmography) sensors;
Memory that stores one or more instructions; and
A processor that executes one or more instructions related to processing of a plurality of PPG data measured by the plurality of PPG sensors;
Including,
The above processor,
Acquire multiple PPG data measured from multiple areas of the user's skin through the multiple PPG sensors,
Obtaining at least one refined PPG data from the plurality of PPG data,
Estimating health status information of the user based on at least one purified PPG data,
The above processor,
At least one of the first PPG data having the highest quality among the plurality of PPG data, the second PPG data calculated based on the weighted average of the plurality of PPG data, and the third PPG data satisfying a specific measurement condition among the plurality of PPG data is determined as the main PPG data,
By removing noise from the above main PPG data, at least one refined PPG data is obtained,
In order to determine the main PPG data, a quality index including a signal-to-noise ratio (SNR), an artifact rate, and signal consistency is calculated, a quality score of each of the plurality of PPG data based on the quality index is calculated, data having the highest calculated quality score is acquired as the first PPG data, and the second PPG data is acquired by weighting and averaging the plurality of PPG data values at any point in time by a weight based on the quality score of the plurality of PPG data, and the third PPG data is acquired based on a specific condition for at least one of the plurality of PPG sensors, the user's activity state, and environmental conditions for acquiring the plurality of PPG data, and the first PPG data is preferentially applied as the main PPG data among the acquired first PPG data to the third PPG data,
Removing noise from the main PPG data obtained based on any other PPG data containing bio-signal components similar to the main PPG data to obtain the refined PPG data.
Health status information estimation device.
Combined with the hardware, the computer,
A step of acquiring multiple PPG data measured from multiple areas of a user's skin through multiple PPG (Photoplethysmography) sensors;
A step of obtaining at least one refined PPG data from the plurality of PPG data; and
A step of estimating health status information of the user based on at least one purified PPG data;
A computer program stored on a computer-readable recording medium to enable execution of a task.