Electronic device and method for controlling same

Abstract

Disclosed is an electronic device. The electronic device comprises: at least one sensor; memory storing a machine learning model and instructions; and at least one processor. When the instructions are individually or collectively executed by the at least one processor, the electronic device: obtains a PPG signal of a user by using the at least one sensor; performs preprocessing on the PPG signal by using a filter; obtains a power spectrum by converting the preprocessed PPG signal into frequency over time; obtains an average spectrum power for each frequency by calculating an average value of the obtained power spectrum over time; and obtains information about diabetes of the user by inputting the obtained average spectrum power for each frequency to the machine learning model.

Description

Electronic device and control method thereof

[0001] The present disclosure relates to an electronic device and a method for controlling the same, and more specifically, to an electronic device and a method for controlling the same for diagnosing diabetes using photoplethysmography (PPG).

[0002] Diabetes mellitus is a metabolic disease caused by factors such as insufficient insulin secretion or failure of the insulin-secreting organs to function normally. Diabetes mellitus is characterized by hyperglycemia, which is a high concentration of glucose in the blood; when hyperglycemia occurs, it can lead to various complications or cause glucose to be excreted in the urine.

[0003] Diabetes can be diagnosed through blood tests. For example, if blood glucose levels are 126 mg / dL or higher after fasting for more than 8 hours, or 200 mg / dL or higher 2 hours after an oral glucose tolerance test, diabetes may be suspected.

[0004] However, diagnosing diabetes through blood tests is time-consuming or cumbersome, making it difficult for users to utilize. Furthermore, since most diabetic patients do not experience symptoms in mild hyperglycemia, it is difficult to suspect they have diabetes, which may lead to them not undergoing testing.

[0005] According to the present disclosure, an electronic device comprises at least one sensor, a memory for storing machine learning models and instructions, and at least one processor. When the instructions are executed individually or collectively by the at least one processor, the electronic device acquires a user’s Photoplethysmography (PPG) signal using the at least one sensor, performs preprocessing on the PPG signal using a filter, converts the preprocessed PPG signal into a frequency over time to acquire a power spectrum, calculates the average value of the acquired power spectrum over time to acquire the average spectrum power by frequency, and inputs the acquired average spectrum power by frequency into the machine learning model to acquire information about the user’s diabetes.

[0006] According to the present disclosure, the machine learning model may be a model trained to output information about diabetes based on the average spectral power for each frequency obtained using the PPG signals of a plurality of people.

[0007] When the above instructions are executed individually or collectively by the at least one processor, the electronic device may perform preprocessing on the PPG signal using a bandpass filter or a Savitzky-Golay filter.

[0008] The electronic device can convert the preprocessed PPG signal into a time-frequency axis using a wavelet transform and obtain a wavelet power spectrum.

[0009] The electronic device can obtain the sum of the spectral power for each frequency range of blood flow regulation mechanisms based on the average spectral power of the user, and input the obtained sum of the spectral power into the machine learning model to obtain information about the user's diabetes.

[0010] According to the present disclosure, the blood flow regulating mechanism may include at least one of a cardiac mechanism and a myogenic mechanism. Additionally, when the instructions are executed individually or collectively by the at least one processor, the electronic device may obtain the sum of the spectral power for the frequency range of the cardiac mechanism or the myogenic mechanism among the frequency components based on the frequency-averaged spectral power of the user.

[0011] The electronic device can obtain the sum of the spectral powers for the frequency domains of the user's cardiac mechanism and muscle mechanism, and input the sum of the spectral powers into the machine learning model to obtain information about the user's diabetes.

[0012] According to the present disclosure, the machine learning model may include a machine learning model having a plurality of people's ages as variables. The electronic device may obtain information regarding the user's diabetes by inputting the sum of the spectral powers in the frequency domain of the user's cardiac mechanism or muscle mechanism and the user's age information into the machine learning model.

[0013] According to the present disclosure, the blood flow regulating mechanism may further include a respiratory mechanism. Additionally, the machine learning model may include a machine learning model having as a variable the sum of the spectral power in the frequency domain of the cardiac mechanism, the spectral power in the frequency domain of the muscle mechanism, and the spectral power in the frequency domain of the respiratory mechanism. The electronic device may obtain information regarding the user's diabetes by inputting the sum of the spectral power in the frequency domains of the user's cardiac mechanism, muscle mechanism, and respiratory mechanism into the machine learning model.

[0014] Meanwhile, a control method for an electronic device according to one or more embodiments of the present disclosure includes the steps of acquiring a user's PPG (Photoplethysmography) signal, performing preprocessing on the PPG signal using a filter, acquiring a power spectrum by converting the preprocessed PPG signal into a frequency over time, acquiring an average spectrum power by frequency by calculating an average value over time of the acquired power spectrum, and acquiring information about the user's diabetes by inputting the acquired average spectrum power by frequency into a machine learning model.

[0015] According to the present disclosure, the step of acquiring the power spectrum may include the step of converting the preprocessed PPG signal into a time-frequency axis using a wavelet transform and acquiring a wavelet power spectrum.

[0016] The step of obtaining the average spectral power by frequency may include the step of obtaining the sum of the spectral power for each frequency range of blood flow regulation mechanisms based on the average spectral power by frequency.

[0017] According to the present disclosure, the blood flow regulating mechanism may include at least one of a cardiac mechanism and a myogenic mechanism. The step of obtaining the frequency-specific average spectral power may include obtaining the sum of the spectral power for the frequency range of the cardiac mechanism or the myogenic mechanism among the frequency components based on the frequency-specific average spectral power of the user.

[0018] According to the present disclosure, the machine learning model may include a machine learning model having a plurality of people's ages as variables. Additionally, the step of obtaining information regarding the user's diabetes may include the step of obtaining information regarding the user's diabetes by inputting the sum of the spectral powers in the frequency domain of the user's cardiac mechanism or muscle mechanism and the user's age information into the machine learning model.

[0019] The above blood flow regulation mechanism may further include a respiratory mechanism. Additionally, the machine learning model may include a machine learning model that takes as a variable the sum of the spectral power obtained by summing the spectral power in the frequency domain of the cardiac mechanism, the spectral power in the frequency domain of the muscle mechanism, and the spectral power in the frequency domain of the respiratory mechanism. The step of obtaining information about the user's diabetes may include the step of obtaining information about the user's diabetes by inputting the sum of the spectral power obtained by summing the spectral power in the frequency domains of the user's cardiac mechanism, muscle mechanism, and respiratory mechanism into the machine learning model.

[0020] FIG. 1 is a block diagram showing the configuration of an electronic device (100) according to various embodiments of the present disclosure.

[0021] FIG. 2 is a block diagram showing the detailed configuration of an electronic device according to various embodiments of the present disclosure.

[0022] FIGS. 3 to 7 are drawings for explaining the operation of acquiring a frequency-dependent power spectrum based on a PPG signal according to various embodiments of the present disclosure.

[0023] FIGS. 8 to 13 are drawings for explaining the diabetes diagnostic performance of an electronic device according to various embodiments of the present disclosure.

[0024] FIG. 14 is a flowchart illustrating a method for controlling an electronic device according to various embodiments of the present disclosure.

[0025] The terms used in this specification will be briefly explained, and the present disclosure will be described in detail.

[0026] The terms used in the embodiments of this disclosure have been selected to be as widely used as possible, taking into account their functions within this disclosure; however, these terms may vary depending on the intent of those skilled in the art, case law, the emergence of new technologies, etc. Additionally, in specific cases, terms have been arbitrarily selected by the applicant, and in such cases, their meanings will be described in detail in the relevant explanatory section of this disclosure. Therefore, terms used in this disclosure should be defined not merely by their names, but based on their meanings and the overall content of this disclosure.

[0027] In this specification, expressions such as “have,” “may have,” “include,” or “may include” indicate the presence of such features (e.g., numerical values, functions, operations, or components such as parts) and do not exclude the presence of additional features.

[0028] In the present disclosure, expressions such as “A or B,” “at least one of A or / and B,” or “one or more of A or / and B” may include all possible combinations of items listed together. For example, “A or B,” “at least one of A and B,” or “at least one of A or B” may refer to cases including (1) at least one A, (2) at least one B, or (3) both at least one A and at least one B.

[0029] Expressions such as "first," "second," "first," or "second" used in this specification may modify various components regardless of order and / or importance, and are used only to distinguish one component from another and do not limit said components.

[0030] Where it is stated that a component (e.g., Component 1) is "operatively or communicatively coupled with / to" or "connected to" another component (e.g., Component 2), it should be understood that the component may be directly connected to the other component or connected through the other component (e.g., Component 3).

[0031] The singular expression includes the plural expression unless the context clearly indicates otherwise. In this application, terms such as "comprising" or "consisting of" are intended to specify the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0032] In the present disclosure, a "module" or "part" performs at least one function or operation and may be implemented in hardware or software, or a combination of hardware and software. Additionally, a plurality of "modules" or a plurality of "parts" may be integrated into at least one module and implemented by at least one processor (not shown), except for a "module" or "part" that needs to be implemented in specific hardware.

[0033] An embodiment of the present disclosure will be described in more detail below with reference to the attached drawings.

[0034] FIG. 1 is a block diagram showing the configuration of an electronic device (100) according to various embodiments of the present disclosure. According to one embodiment of the present disclosure, the electronic device (100) may be implemented as various types of devices such as a wearable device, a smartphone, a PDA (personal digital assistant), a computer, a laptop PC, a tablet PC, etc. Additionally, wearable devices may include a smart watch, a smart ring, a smart bracelet, a smart band, etc. However, it is not limited thereto, and the electronic device (100) may be implemented as various types of devices equipped with at least one processor capable of collecting biometric data and processing data.

[0035] According to FIG. 1, the electronic device (100) includes at least one sensor (110), memory (120), and processor (130).

[0036] At least one sensor (110) is configured to sense various information in relation to the operation of the electronic device (100). At least one sensor (110) can detect the user's biometric data. For example, at least one sensor (110) may include a PPG sensor for detecting the user's photoplethysmography (PPG).

[0037] Photoplethysmography is a compound word derived from Photo, Plethysmos, and Graphos, and can be abbreviated as PPG using the first letters of the three words. PPG can reveal minute changes in blood vessels that occur during the process of blood circulation caused by the contraction and relaxation of the heart.

[0038] A PPG sensor can measure blood flow through a user's skin. Specifically, as the amount of blood flow within the blood vessels repeatedly increases and decreases whenever the heart contracts and relaxes, the PPG sensor can measure the PPG signal by projecting light onto the skin and detecting the light reflected or transmitted from the skin. For example, the PPG sensor may include an LED for projecting green light or infrared light and a photodiode for receiving reflected light.

[0039] Because oxidized hemoglobin in the blood has the characteristic of absorbing green light well, green light can efficiently detect the distribution of hemoglobin in the blood when measuring blood flow using a PPG sensor. In addition, since the infrared light projected from the PPG sensor does not obstruct the user's vision, it can be used to measure the user's blood flow at night.

[0040] The memory (120) may be implemented as internal memory such as ROM (e.g., EEPROM (electrically erasable programmable read-only memory)) or RAM included in the processor (130), or as memory separate from the processor (130). Depending on the purpose of data storage, the memory (120) may be implemented as a memory embedded in the electronic device (100) or as a memory detachable from the electronic device (100). For example, data for operating the electronic device (100) may be stored in memory embedded in the electronic device (100), and data for the expansion function of the electronic device (100) may be stored in memory detachable from the electronic device (100).

[0041] Meanwhile, the memory embedded in the electronic device (100) may be implemented as at least one of volatile memory (e.g., DRAM (dynamic RAM), SRAM (static RAM), or SDRAM (synchronous dynamic RAM), non-volatile memory (e.g., OTPROM (one time programmable ROM), PROM (programmable ROM), EPROM (erasable and programmable ROM), EEPROM (electrically erasable and programmable ROM), mask ROM, flash ROM, flash memory (e.g., NAND flash or NOR flash), hard drive, or solid state drive (SSD), and the memory that is detachable from the electronic device (100) may be implemented in the form of a memory card (e.g., CF (compact flash), SD (secure digital), Micro-SD (micro secure digital), Mini-SD (mini secure digital), xD (extreme digital), MMC (multi-media card), etc.), or external memory that can be connected to a USB port (e.g., USB memory).

[0042] The memory (120) may store at least one instruction, data, program, etc., required for the operation of the electronic device (100) or the processor (130). For example, the memory (120) may store biometric data or PPG signals obtained using at least one sensor (110). Additionally, an operating system (O / S) for operating the electronic device (100) may be stored in the memory (120). Furthermore, various software programs or applications for the operation of the electronic device (100) may be stored in the memory (120) according to various embodiments of the present disclosure. Also, the memory (120) may include semiconductor memory such as flash memory or magnetic storage media such as hard disk.

[0043] Specifically, various software modules for operating an electronic device (100) according to various embodiments of the present disclosure may be stored in the memory (120), and the processor (130) may control the operation of the electronic device (100) by executing the various software modules stored in the memory (120). That is, the memory (120) is accessed by the processor (130), and reading / writing / modifying / deleting / updating of data by the processor (130) may be performed.

[0044] The memory (120) may be implemented as a single memory that stores data generated in various operations according to the present disclosure, but is not limited thereto, and the memory (120) may be implemented to include a plurality of memories that each store different types of data or each store data generated in different stages.

[0045] Meanwhile, in the present disclosure, the term memory (120) may be used to include a storage unit, a ROM (not shown), a RAM (not shown) within a processor (130), or a memory card (not shown) (e.g., a micro SD card, a memory stick) mounted in an electronic device (100).

[0046] A processor (130) is a component connected to each component of an electronic device (100) to control the overall operation of the electronic device (100). The processor (130) may be implemented as a digital signal processor (DSP) that processes digital signals, a microprocessor, a Graphics Processing Unit (GPU), an Artificial Intelligence (AI) processor, a Neural Processing Unit (NPU), or a Time Controller (TCON). However, it is not limited thereto, and may include or be defined by one or more of a central processing unit (CPU), a Micro Controller Unit (MCU), a micro processing unit (MPU), a controller, an application processor (AP), a communication processor (CP), or an ARM processor. Additionally, the processor (130) may be implemented as a System on Chip (SoC) or Large Scale Integration (LSI) with built-in processing algorithms, or may be implemented in the form of an Application Specific Integrated Circuit (ASIC) or Field Programmable Gate Array (FPGA). In addition, the processor (130) can perform various functions by executing computer executable instructions stored in memory (120).

[0047] Additionally, a processor (130) for executing a machine learning model according to one embodiment can be implemented through a combination of software and a general-purpose processor such as a CPU, AP, DSP (Digital Signal Processor), a graphics-dedicated processor such as a GPU, VPU (Vision Processing Unit), or a machine learning-dedicated processor such as an NPU.

[0048] The processor (130) can be controlled to process input data according to predefined operation rules or machine learning models stored in memory (120). Alternatively, if the processor (130) is a dedicated processor (or a machine learning dedicated processor), it can be designed with a hardware structure specialized for processing a specific machine learning model. For example, hardware specialized for processing a specific machine learning model can be designed as a hardware chip such as an ASIC or FPGA.

[0049] When the processor (130) is implemented as a dedicated processor, it may be implemented to include memory for implementing an embodiment of the present disclosure, or may be implemented to include a memory processing function for using external memory. The processor (130) may be implemented as one or a plurality of processors.

[0050] Meanwhile, the artificial intelligence-related functions according to the present disclosure may be operated through a processor (130) and a memory (120). One or more processors (130) control the processing of input data according to a predefined operation rule or machine learning model stored in the memory (120). Alternatively, if one or more processors are dedicated artificial intelligence processors, the dedicated artificial intelligence processors may be designed with a hardware structure specialized for processing a specific machine learning model. The predefined operation rule or machine learning model is characterized by being created through learning.

[0051] Here, "created through learning" means that a basic machine learning model is trained using multiple learning data by a learning algorithm, thereby creating a predefined behavioral rule or machine learning model configured to perform a desired characteristic (or objective). Such learning may be performed on the device itself where the artificial intelligence according to the present disclosure is executed, or it may be performed through a separate server and / or system. Examples of learning algorithms include, but are not limited to, supervised learning, unsupervised learning, semi-supervised learning, or reinforcement learning.

[0052] The processor (130) can measure the user's biometric data or photoplethysmography (PPG) using at least one sensor (110). For example, if at least one sensor (110) includes a PPG sensor, the processor (130) can measure blood flow from the user's skin for a preset time using the PPG sensor and acquire a PPG (Photoplethysmography) signal.

[0053] The processor (130) can perform preprocessing on the PPG signal using a filter. For example, the processor (130) can perform preprocessing on the PPG signal using a bandpass filter or a Savitzky-Golay filter. The operation of performing preprocessing on the PPG signal using a filter will be described in detail in the following section.

[0054] The processor (130) can obtain a power spectrum by converting the preprocessed PPG signal into a frequency over time. For example, the processor (130) can obtain a wavelet power spectrum by converting the preprocessed PPG signal into a time-frequency axis using a wavelet transform.

[0055] The processor (130) can obtain the average spectrum power by frequency by calculating the average value of the acquired power spectrum over time. Additionally, the processor (130) can obtain information about the user's diabetes based on the acquired average spectrum power by frequency and provide information about the diabetes to the user. For example, the processor (130) can obtain information about the user's diabetes by inputting the acquired average spectrum power by frequency into a machine learning model.

[0056] In this case, a machine learning model may be stored in the memory (120). The machine learning model may be a model trained to output information about diabetes based on the average spectral power for each frequency obtained using the PPG signals of multiple people. For example, the processor (130) may receive the PPG signals of multiple people, obtain the average spectral power for each frequency, and input the obtained average spectral power for each frequency into the machine learning model to train the machine learning model to output information about diabetes.

[0057] The operation of obtaining information about a user's diabetes based on the average spectral power by frequency will be explained in detail again in the following section through FIGS. 3 to 7.

[0058] FIG. 2 is a block diagram showing the detailed configuration of an electronic device according to various embodiments of the present disclosure. According to FIG. 2, the electronic device (100) may include at least one sensor (110), a memory (120), a processor (130), an interface (140), and a display (150). A detailed description of configurations shown in FIG. 2 that overlap with configurations shown in FIG. 1 will be omitted.

[0059] The interface (140) is configured to receive various data from a user, external memory, or external device. For example, the interface (140) can receive the user's biometric data or PPG signals from an external electronic device.

[0060] The interface (140) may include a communication interface (141), an operation interface (142), and an input / output interface (143), etc. The communication interface (141) is configured to perform communication with at least one external device. The communication interface (141) may include at least one wireless communication module, at least one wired communication module, etc. Each communication module may be implemented in the form of at least one hardware chip. The wireless communication module may include at least one module among a Wi-Fi module, a Bluetooth module, an infrared communication module, or other communication modules. In addition, the communication interface (141) may include at least one communication chip that performs communication according to various wireless communication standards such as Zigbee, 3G (3rd Generation), 3GPP (3rd Generation Partnership Project), LTE (Long Term Evolution), LTE-A (LTE Advanced), 4G (4th Generation), 5G (5th Generation), etc. The wired communication module may include, for example, at least one of a LAN (Local Area Network) module, an Ethernet module, a pair cable, a coaxial cable, a fiber optic cable, or an UWB (Ultra Wide-Band) module. The communication interface (141) is implemented in various forms in this way and can receive various sensing data from an external device by communicating with the external device.

[0061] The operation interface (142) is configured to receive user operation input. The operation interface (142) may include various buttons, a touch screen, etc. provided on the main body of the electronic device (100). The user can use the operation interface (142) to control the operation of at least one sensor (110) or directly input various data into the electronic device (100).

[0062] The input / output interface (143) is configured to input and output various external signals. The input / output interface (143) can be connected to various external memory or external sources (e.g., web server, user terminal device, etc.) to receive various data. The input / output interface (143) can be implemented as at least one interface among HDMI (High Definition Multimedia Interface), MHL (Mobile High-Definition Link), USB (Universal Serial Bus), USB C-type, DP (Display Port), Thunderbolt, VGA (Video Graphics Array) port, RGB port, D-SUB (D-subminiature), and DVI (Digital Visual Interface). At least some of the input / output interfaces (143) may be connected to a communication interface (141). For example, the input / output interface (143) can transmit information received from an external device to the communication interface (141) or transmit information received through the communication interface (141) to an external device. The electronic device (100) can directly read or receive voice data or biometric data stored in an external memory or external source connected through an input / output interface (143).

[0063] The display (150) can perform display operations under the control of the processor (130). For example, the processor (130) can control the display (150) to display information about diabetes.

[0064] The display (150) may be implemented as a display including a self-emissive element or as a display including a non-emissive element and a backlight. For example, it may be implemented as various types of displays such as an LCD (Liquid Crystal Display), an OLED (Organic Light Emitting Diodes) display, an LED (Light Emitting Diodes), a micro LED, a Mini LED, a PDP (Plasma Display Panel), a QD (Quantum dot) display, a QLED (Quantum dot light-emitting diodes), etc.

[0065] FIGS. 3 to 7 are drawings for explaining the operation of acquiring a frequency-dependent power spectrum based on a PPG signal according to various embodiments of the present disclosure. FIG. 3 is a drawing for explaining a PPG signal acquired using at least one sensor (110). In FIG. 3, the x-axis may represent time, and the y-axis may represent the amplitude of the PPG signal.

[0066] The electronic device (100) can acquire a user's PPG signal using at least one sensor (110). When the at least one sensor (110) includes a PPG sensor, the electronic device (100) can collect the user's PPG signal for a preset time using the PPG sensor.

[0067] Figure 4 is a diagram illustrating a PPG signal preprocessed using a filter. In Figure 4, the x-axis represents time, and the y-axis represents the amplitude of the preprocessed PPG signal.

[0068] The electronic device (100) can perform preprocessing on the PPG signal using a filter. When the electronic device (100) measures the user's PPG signal using at least one sensor (110), signal distortion may occur due to noise caused by poor blood perfusion, ambient light, user movement, etc. Therefore, the electronic device (100) can perform preprocessing on the PPG signal using a filter to minimize the influence of such noise.

[0069] Meanwhile, a filter can perform the function of blocking unwanted signals from a specific signal or allowing only desired signals to pass through. Measurements that vary over time may consist of combinations of frequency components or harmonics. In this case, unwanted noise may be included in the frequencies or harmonics, potentially distorting or degrading the signal. Filters can selectively remove noise by blocking signals in specific frequency bands. For example, if a signal is represented by frequencies, a filter can selectively pass or block specific frequency components from the signal.

[0070] Filters can be classified in various ways. For example, filters can be classified into linear and nonlinear filters based on their linearity, and into analog and digital filters based on the types of input and output values. Additionally, filters can be classified into continuous-time and discrete-time filters based on the time type of the input and output values, and into passive and active filters based on the type of filter element.

[0071] For example, linear filters can be classified into low-pass filters, high-pass filters, band-pass filters, and band-stop filters depending on their frequency passbands and cutoff bands. A low-pass filter allows low-frequency components below a certain frequency to pass through while blocking frequency components above that frequency. Low-pass filters can be used to remove high-frequency noise or high-frequency components. A high-pass filter allows only high-frequency components above a certain frequency to pass through while blocking frequency components below that frequency. High-pass filters can be used to remove low-frequency noise or DC components.

[0072] A band-pass filter can allow only specific frequency bands to pass through while blocking the rest. Using a band-pass filter allows you to extract only signals within a specific frequency range. A band-stop filter can block only specific frequency bands while allowing the rest to pass through. Band-stop filters can be used to eliminate specific frequency bands.

[0073] The Savitzky-Golay filter can represent a filter for smoothing data. Specifically, the Savitzky-Golay filter can represent a digital filter that can be applied to a set of digital data points to increase the precision of the data without distorting the signal trend. The Savitzky-Golay filter can filter signals by fitting a continuous subset of adjacent data points to a low-order polynomial using linear least squares in the convolution process.

[0074] Figure 5 is a diagram illustrating a power spectrum. Specifically, Figure 5 may represent a wavelet power spectrum obtained by performing a wavelet transform on a preprocessed PPG signal. In Figure 5, the x-axis represents time, the y-axis on the left represents the frequency of the power spectrum, and the y-axis on the right represents the wavelet power levels.

[0075] The electronic device (100) can perform a wavelet transform on the preprocessed PPG signal to decompose it into a time-frequency axis and obtain a power spectrum.

[0076] Meanwhile, the wavelet transform can analyze data where features change at different scales. The wavelet transform can be used to overcome the limitations of the Fourier transform. While the Fourier transform decomposes a signal into sine waves of specific frequencies, the wavelet transform decomposes the signal into various waveforms and can represent them as arbitrary waveforms through scaling, such as enlargement or reduction. For example, if a signal that changes over time is applied to the wavelet transform, features such as frequency changes over time and trends that change excessively or slowly can be analyzed. In the case of images, image boundaries and textures can be analyzed as features of the wavelet transform.

[0077] FIGS. 6 and FIGS. 7 are diagrams illustrating the average spectral power by frequency. Specifically, FIG. 6 may show the average spectral power by frequency for a normal person, and FIG. 7 may show the average spectral power by frequency for a diabetic patient. In FIGS. 6 and FIGS. 7, the x-axis may represent frequency, and the y-axis may represent average wavelet spectral power. When the electronic device (100) acquires a power spectrum, it may acquire an average value on the time axis for the acquired power spectrum and acquire the average spectral power by frequency. Additionally, the electronic device (100) may input the acquired average spectral power by frequency into a machine learning model to obtain information about the user's diabetes.

[0078] According to FIGS. 6 and 7, the frequency of the acquired average spectral power may range from 0.02 Hz to 20.00 Hz. In this case, the electronic device (100) may acquire the sum of the spectral power for each frequency domain of the blood flow control mechanism based on the average spectral power acquired for each frequency domain. The blood flow control mechanism may include at least one of a neurogenic mechanism, a myogenic mechanism, a respiratory mechanism, and a cardiac mechanism. For example, the electronic device (100) may represent the frequency domains for each blood flow control mechanism as shown in Table 1 below.

[0079] Frequency (Hz) Blood Flow Mechanism 0.02 ~ 0.06 Neurogenic: Regulation by sympathetic nerves 0.06 ~ 0.2 Myogenic: Regulation of blood vessel walls by vascular smooth muscle cells 0.2 ~ 0.6 Respiratory: Movement of the thoracic cavity due to respiration 0.6 ~ 1.6 Cardiac: Heartbeat

[0080] Referring to FIGS. 6 and 7, a first frequency range (610, 710) corresponding to a frequency of average spectral power of 0.02 to 0.06 Hz may correspond to blood flow caused by sympathetic nerves (neurogenic). The electronic device (100) can obtain information about diabetes caused by a neurogenic mechanism by using the sum of the spectral power for the first frequency range (610, 710). Additionally, a second frequency range (620, 720) corresponding to a frequency of average spectral power of 0.06 to 0.2 Hz may correspond to blood flow caused by intravascular muscle (myogenic). The electronic device (100) can obtain information about diabetes caused by a myogenic mechanism by using the sum of the spectral power for the second frequency range (620, 720).

[0081] A third frequency range (630, 730) corresponding to an average spectral power frequency of 0.2 to 0.6 Hz can correspond to blood flow caused by respiration. The electronic device (100) can obtain information about diabetes caused by a respiratory mechanism by using the sum of the spectral power for the third frequency range (630, 730). Additionally, a fourth frequency range (640, 740) corresponding to an average spectral power frequency of 0.6 to 1.6 Hz can correspond to blood flow caused by cardiac beats. The electronic device (100) can obtain information about diabetes caused by a cardiac mechanism by using the sum of the spectral power for the fourth frequency range (640, 740).

[0082] Referring to FIGS. 6 and 7, it can be seen that the second frequency range (720) of the average spectral power for diabetic patients is relatively reduced compared to the second frequency range (620) of the average spectral power for normal people, and the fourth frequency range (740) of the average spectral power for diabetic patients is relatively increased compared to the fourth frequency range (640) of the average spectral power for normal people. The electronic device (100) can obtain the sum of the spectral power for the frequency ranges of each blood flow control mechanism based on the average spectral power for each frequency corresponding to the blood flow control mechanism for multiple people, and can train the machine learning model to output information about diabetes by inputting the obtained sum of spectral power into the machine learning model.

[0083] Additionally, when the electronic device (100) obtains the average spectrum power of the user by frequency, it obtains the sum of the spectrum power for each frequency range of the blood flow control mechanism based on the average spectrum power of the user by frequency, and inputs the obtained sum of spectrum power into a machine learning model to obtain information about the user's diabetes.

[0084] For example, the electronic device (100) can obtain the sum of the spectral power for the frequency range of the cardiac or myogenic mechanism among the frequency components based on the average spectral power of the user by frequency, and input the obtained sum of the spectral power into a machine learning model to obtain information about the user's diabetes.

[0085] Alternatively, the electronic device (100) may obtain a sum of spectral powers by summing the spectral powers for the frequency domains of the user's heart mechanism and muscle mechanism based on the user's average spectral power by frequency, and input the sum of the spectral powers into a machine learning model to obtain information about the user's diabetes.

[0086] In this case, the machine learning model may include a logistic regression model with the age of multiple people as a variable.

[0087] A logistic regression model can be described as a machine learning algorithm used to predict the probability that data belongs to a specific class based on binary classification. Unlike linear regression, logistic regression models produce output values ​​between 0 and 1 and can classify data along an S-shaped curve. The logistic regression model calculates the weighted sum of input variables and applies the resulting value to the sigmoid activation function to convert it into a value between 0 and 1. Here, the sigmoid function represents a function capable of predicting probability values ​​by matching input values ​​to a range between 0 and 1. Because logistic regression models demonstrate high performance in classification problems such as binary classification, they can be utilized in various fields, including spam email classification and disease diagnosis.

[0088] When the electronic device (100) obtains the average spectrum power of the user by frequency and the user's age information, it can obtain information about the user's diabetes by inputting the sum of the spectrum power for the frequency domain of the user's heart mechanism or muscle mechanism and the input user's age information into a machine learning model.

[0089] The machine learning model may include a logistic regression model that uses the sum of the spectral powers of the frequency domain of the cardiac mechanism, the frequency domain of the muscle mechanism, and the frequency domain of the respiratory mechanism as variables.

[0090] In this case, when the electronic device (100) obtains the average spectrum power of the user by frequency, it can obtain information about the user's diabetes by inputting the sum of the spectrum powers for the frequency domains of the user's heart mechanism, muscle mechanism, and respiratory mechanism into a machine learning model.

[0091] FIGS. 8 to 13 are drawings for explaining the diabetes diagnostic performance of an electronic device (100) according to various embodiments of the present disclosure. FIG. 8 is a drawing for explaining the case in which diabetes is diagnosed using the sum of the spectral power for the frequency domain of the cardiac mechanism among the frequency components of the PPG signal.

[0092] Referring to FIG. 8, the electronic device (100) can obtain the sum of the spectral power for the frequency domain of the cardiac mechanism among the frequency components of the average spectral power, and can obtain a Receiver Operator Characteristic (ROC) curve (810) for classifying diabetes or normal using the obtained sum of the spectral power. The electronic device (100) can obtain an Area Under Curve (AUC) based on the obtained ROC curve (810), and can obtain the diabetes diagnostic performance of the electronic device (100) by comparing the obtained AUC with a reference AUC. For example, to obtain the diabetes diagnostic performance of the electronic device (100), the electronic device (100) can compare the AUC obtained from the electronic device (100) with the AUC obtained through a skin autofluorescence measurement device (AGE Reader). In FIG. 8, the ROC curve (820) may represent the ROC curve obtained from the skin autofluorescence measurement device (AGE Reader).

[0093] Meanwhile, the ROC curve (Receiver Operator Characteristic Curve) can represent a graph to evaluate the performance of a classification model in machine learning. The ROC curve can be used to evaluate the performance of a classification model at various threshold settings. For example, in an ROC curve, sensitivity can be displayed on the y-axis and specificity on the x-axis; sensitivity indicates how well the classification model identifies actual positives, and specificity indicates the proportion of actual negatives that the classification model classifies as negative.

[0094] The electronic device (100) can obtain an Area Under Curve (AUC) based on an ROC curve and evaluate the performance of a classification model using the obtained AUC. In this case, the AUC can be provided as a single scalar value representing the performance of the classification model. The AUC has a value between 0 and 1, and can indicate that the closer it is to 1, the better the performance of the classification model, and the closer it is to 0.5, the worse the performance of the classification model. Specifically, the AUC can indicate that the larger the area on the graph, the higher the performance of the diagnostic test.

[0095] Referring to FIG. 8, the electronic device (100) can obtain an AUC of approximately 0.707 by using the sum of the spectral power for the frequency range of the cardiac mechanism among the frequency components of the average spectral power. The electronic device (100) can obtain the highest performance at the point (811) corresponding to the coordinates (0.612, 0.820). In this case, the AUC obtained through the skin autofluorescence measurement device (AGE Reader) can be approximately 0.746.

[0096] As such, the electronic device (100) according to various embodiments of the present disclosure can obtain a diagnostic result with performance similar to that of a skin autofluorescence measuring device (AGE Reader) by using the sum of spectral powers in the frequency domain of the heart mechanism.

[0097] Figure 9 is a diagram illustrating a case in which diabetes is diagnosed using the sum of the spectral power for the frequency range of the myogenic mechanism among the frequency components of the PPG signal.

[0098] Referring to FIG. 9, the electronic device (100) can obtain the sum of the spectral power for the frequency range of the muscle mechanism among the frequency components of the average spectral power, and obtain a ROC (Receiver Operator Characteristic) curve (910) for classifying diabetes or normal using the obtained sum of the spectral power. In addition, the electronic device (100) can obtain an AUC (Area under curve) based on the obtained ROC curve (910), and obtain the diabetes diagnostic performance of the electronic device (100) by comparing the obtained AUC with a reference AUC.

[0099] According to FIG. 9, the electronic device (100) can obtain an AUC of approximately 0.709 by using the sum of the spectral powers for the frequency domain of the muscle mechanism. The electronic device (100) can obtain the highest performance at a point (911) corresponding to coordinates (0.571, 0.800).

[0100] As such, the electronic device (100) according to various embodiments of the present disclosure can obtain a diagnostic result with performance similar to that of a skin autofluorescence measuring device (AGE Reader) by using the sum of spectral powers in the frequency domain of the muscle mechanism.

[0101] Figure 10 is a diagram illustrating a case in which diabetes is diagnosed using the sum of the spectral powers for the frequency domains of the muscle mechanism and the heart mechanism among the frequency components of the PPG signal.

[0102] Referring to FIG. 10, the electronic device (100) can obtain a sum of spectral powers by summing the spectral powers for the frequency domains of muscle mechanism and heart mechanism among the frequency components of the average spectral power, and can obtain a Receiver Operator Characteristic (ROC) curve (1010) for classifying diabetes or normal using the sum of the spectral powers. In addition, the electronic device (100) can obtain an AUC based on the obtained ROC curve (1010), and obtain the diabetes diagnostic performance of the electronic device (100) by comparing the obtained AUC with a reference AUC.

[0103] According to FIG. 10, the electronic device (100) can obtain an AUC of about 0.724 by using the sum of the spectral powers for the frequency domains of the muscle mechanism and the heart mechanism.

[0104] Figure 11 is a diagram illustrating a case in which diabetes is diagnosed using the sum of the spectral power for the frequency domain of the cardiac mechanism among the frequency components of the PPG signal and the user's age information.

[0105] Referring to FIG. 11, when an electronic device (100) obtains the sum of the spectral power for the frequency domain of the cardiac mechanism among the frequency components of the average spectral power and the user's age information, it can obtain an ROC curve (1110) for classifying diabetes or normal using the obtained sum of the spectral power for the user's cardiac mechanism and age information. Additionally, the electronic device (100) can obtain an AUC based on the obtained ROC curve (1110) and obtain the diabetes diagnostic performance of the electronic device (100) by comparing the obtained AUC with a reference AUC. According to FIG. 11, the electronic device (100) can obtain an AUC of approximately 0.870 using the sum of the spectral power for the frequency domain of the cardiac mechanism and the user's age information.

[0106] Figure 12 is a diagram illustrating a case in which diabetes is diagnosed using the sum of the spectral power for the frequency domain of the muscle mechanism among the frequency components of the PPG signal and the user's age information.

[0107] Referring to FIG. 12, when an electronic device (100) obtains the sum of the spectral power for the frequency domain of the muscle mechanism among the frequency components of the average spectral power and the user's age information, it can obtain an ROC curve (1210) for classifying diabetes or normal using the obtained sum of the spectral power of the user's muscle mechanism and age information. Additionally, the electronic device (100) can obtain an AUC based on the obtained ROC curve (1210) and obtain the diabetes diagnostic performance of the electronic device (100) by comparing the obtained AUC with a reference AUC. According to FIG. 12, the electronic device (100) can obtain an AUC of approximately 0.902 using the sum of the spectral power for the frequency domain of the muscle mechanism and the user's age information.

[0108] As such, the electronic device (100) according to various embodiments of the present disclosure can obtain a diagnostic result with superior performance to a skin autofluorescence measuring device (AGE Reader) by using the sum of the spectral power in the frequency domain of the cardiac mechanism or muscle mechanism and the user's age information.

[0109] Figure 13 is a diagram illustrating a case in which diabetes is diagnosed using the sum of spectral powers for the frequency domains of the muscular, cardiac, and respiratory mechanisms among the frequency components of the PPG signal.

[0110] Referring to FIG. 13, the electronic device (100) can obtain a sum of spectral powers by summing the spectral powers for the frequency domains of the muscle mechanism, cardiac mechanism, and respiratory mechanism among the frequency components of the average spectral power, and can obtain an ROC curve (1310) for classifying diabetes or normal using the sum of the spectral powers. Additionally, the electronic device (100) can obtain an AUC based on the obtained ROC curve (1310), and obtain the diabetes diagnostic performance of the electronic device (100) by comparing the obtained AUC with a reference AUC. According to FIG. 13, the electronic device (100) can obtain an AUC of approximately 0.739 using the sum of spectral powers for the frequency domains of the muscle mechanism, cardiac mechanism, and respiratory mechanism.

[0111] As such, the electronic device (100) according to various embodiments of the present disclosure can obtain high-performance diabetes diagnostic information by using the sum of spectral powers for frequency domains according to blood flow control mechanisms.

[0112] FIG. 14 is a flowchart illustrating a method for controlling an electronic device according to various embodiments of the present disclosure. In this case, the electronic device may be implemented as various types of devices such as a wearable device, a smartphone, a PDA (personal digital assistant), a computer, a laptop PC, a tablet PC, etc. Additionally, the wearable device may include a smart watch, a smart ring, a smart bracelet, a smart band, etc.

[0113] According to FIG. 14, the electronic device acquires the user's PPG (Photoplethysmography) signal (S1410). For example, the electronic device may irradiate green light or infrared light onto the skin for a certain period of time and measure the PPG signal using the light reflected or transmitted from the skin. In this case, the PPG signal may include changes in blood flow over time.

[0114] The electronic device performs preprocessing on the PPG signal using a filter (S1420). For example, the electronic device may perform preprocessing on the PPG signal using a bandpass filter or a Savitzky-Golay filter.

[0115] When an electronic device measures a user's PPG signal, signal distortion may occur due to the inclusion of noise caused by poor blood perfusion, ambient light, user movement, etc. Therefore, the electronic device may perform preprocessing on the PPG signal using a filter to minimize the impact of such noise.

[0116] The electronic device converts the preprocessed PPG signal into a frequency over time to obtain a power spectrum (S1430). For example, the electronic device can convert the preprocessed PPG signal into a time-frequency axis using a wavelet transform and obtain a wavelet power spectrum.

[0117] The electronic device calculates the average value of the acquired power spectrum over time to obtain the average spectral power by frequency (S1440). For example, the electronic device can obtain the sum of the spectral power for each frequency domain by blood flow control mechanism based on the average spectral power by frequency. In this case, the blood flow control mechanism may include at least one of a neurogenic mechanism, a myogenic mechanism, a respiratory mechanism, and a cardiac mechanism.

[0118] In addition, a first frequency range corresponding to an average spectral power frequency of 0.02 to 0.06 Hz may correspond to blood flow caused by the sympathetic nervous system (neurogenic). The electronic device can obtain information about diabetes caused by a neurogenic mechanism by using the sum of the spectral power for the first frequency range. In addition, a second frequency range corresponding to an average spectral power frequency of 0.06 to 0.2 Hz may correspond to blood flow caused by intravascular muscle (myogenic). The electronic device can obtain information about diabetes caused by a myogenic mechanism by using the sum of the spectral power for the second frequency range.

[0119] A third frequency region, corresponding to a frequency of average spectral power of 0.2 to 0.6 Hz, can correspond to blood flow caused by respiration. An electronic device can obtain information about diabetes caused by a respiratory mechanism by using the sum of the spectral power for the third frequency region. Additionally, a fourth frequency region, corresponding to a frequency of 0.6 to 1.6 Hz, can correspond to blood flow caused by cardiac beats. An electronic device can obtain information about diabetes caused by a cardiac mechanism by using the sum of the spectral power for the fourth frequency region.

[0120] For example, if an electronic device acquires the average spectral power by frequency of a user, the electronic device can acquire the sum of the spectral power for the frequency range of the cardiac or myogenic mechanism among the frequency components based on the average spectral power by frequency of the user.

[0121] The electronic device inputs the acquired frequency-specific average spectral power into a machine learning model to obtain information about the user's diabetes (S1450). For example, when the electronic device acquires the user's frequency-specific average spectral power, the electronic device acquires the sum of the spectral power for the frequency domain of the cardiac mechanism or myogenic mechanism among the frequency components based on the user's frequency-specific average spectral power, and inputs the acquired sum of spectral power into a machine learning model to obtain information about the user's diabetes.

[0122] In this case, the machine learning model may be a model trained to output information about diabetes based on the average spectral power for each frequency obtained using PPG signals from multiple people. For example, an electronic device may receive PPG signals from multiple people, obtain the average spectral power for each frequency, obtain the sum of the spectral power for each frequency domain of the blood flow regulation mechanism based on the obtained average spectral power for each frequency, and input the obtained sum of the spectral power into the machine learning model to train the machine learning model to output information about diabetes.

[0123] The machine learning model may include a logistic regression model with the age of multiple people as a variable.

[0124] When an electronic device acquires the user's average spectral power by frequency and the user's age information, it can obtain information about the user's diabetes by inputting the sum of the spectral power in the frequency domain of the user's cardiac or muscular mechanism and the user's age information into a machine learning model.

[0125] The machine learning model may include a logistic regression model that uses the sum of the spectral powers of the frequency domain of the cardiac mechanism, the frequency domain of the muscle mechanism, and the frequency domain of the respiratory mechanism as variables.

[0126] When the electronic device acquires the average spectral power of the user by frequency, it can obtain information about the user's diabetes by inputting the sum of the spectral power—which is the sum of the spectral powers for the frequency domains of the user's cardiac, muscular, and respiratory mechanisms—into a machine learning model. Additionally, the electronic device can display the acquired information about diabetes to provide it to the user.

[0127] In this case, the electronic device may provide the user with information about diabetes obtained by blood flow regulation mechanism, but it may also provide information about each diabetes by comprehensively aggregating it. For example, the electronic device may assign weights to information about diabetes by each blood flow regulation mechanism and provide summed result information through calculations between the weights.

[0128] As such, the electronic device and the control method according to various embodiments of the present disclosure can obtain information about diabetes in each frequency domain according to a blood flow control mechanism based on a PPG signal measuring the user's blood flow.

[0129] Accordingly, the electronic device and the control method according to various embodiments of the present disclosure can easily obtain information about diabetes by a non-invasive method without using blood tests. In addition, information about diabetes can be easily and quickly obtained without using expensive equipment such as an AGE Reader.

[0130] Meanwhile, according to the embodiments of the present disclosure, the various embodiments described above may be implemented as software containing instructions stored on a machine-readable storage medium (e.g., a computer). The machine may include an electronic device according to the disclosed embodiments, which is a device capable of calling instructions stored from the storage medium and operating according to the called instructions. When instructions are executed by a processor, the processor may perform a function corresponding to the instructions directly or by using other components under the control of the processor. Instructions may include code generated or executed by a compiler or an interpreter. The machine-readable storage medium may be provided in the form of a non-transitory storage medium. Here, "non-transitory" means only that the storage medium does not contain a signal and is tangible, and does not distinguish whether data is stored semi-permanently or temporarily in the storage medium.

[0131] Additionally, according to one embodiment of the present disclosure, the method according to the various embodiments described above may be provided as included in a computer program product. The computer program product may be traded between a seller and a buyer as a product. The computer program product may be distributed in the form of a device-readable storage medium (e.g., compact disc read-only memory (CD-ROM)) or online through an application store (e.g., Play Store™). In the case of online distribution, at least a portion of the computer program product may be temporarily stored or temporarily created in a storage medium such as the memory of a manufacturer's server, an application store's server, or a relay server.

[0132] Additionally, each component (e.g., module or program) according to the various embodiments described above may be composed of a single or multiple entities, and some of the aforementioned sub-components may be omitted, or other sub-components may be further included in the various embodiments. Generally or additionally, some components (e.g., module or program) may be integrated into a single entity to perform the functions performed by each of the respective components prior to integration in the same or similar manner. The operations performed by the module, program, or other components according to the various embodiments may be executed sequentially, in parallel, iteratively, or heuristically, or at least some operations may be executed in a different order, omitted, or other operations added.

[0133] Although preferred embodiments of the present disclosure have been illustrated and described above, the present disclosure is not limited to the specific embodiments described above. It is understood that various modifications can be made by those skilled in the art without departing from the essence of the present disclosure as claimed in the claims, and such modifications should not be understood individually from the technical spirit or perspective of the present disclosure.

Claims

1. In an electronic device, At least one sensor; Memory for storing machine learning models and instructions; and It includes at least one processor; and When the above instructions are executed individually or collectively by the at least one processor, the electronic device, Acquire a user's PPG (Photoplethysmography) signal using at least one of the above sensors, and Preprocessing of the above PPG signal is performed using a filter, and The preprocessed PPG signal is converted into frequency over time to obtain the power spectrum, and Calculate the time-averaged value of the acquired power spectrum to obtain the frequency-specific average spectral power, and An electronic device that inputs the acquired average spectrum power by frequency into the machine learning model to obtain information about the user's diabetes.

2. In Paragraph 1, The above machine learning model is, An electronic device, which is a model trained to output information about diabetes based on the average spectral power for each frequency obtained using PPG signals from multiple people.

3. In Paragraph 1, When the above instructions are executed individually or collectively by the at least one processor, the electronic device, An electronic device that performs preprocessing on the PPG signal using a band-pass filter or a Savitzky-Golay filter.

4. In Paragraph 1, When the above instructions are executed individually or collectively by the at least one processor, the electronic device, An electronic device that converts the above-mentioned preprocessed PPG signal into a time-frequency axis using a wavelet transform and obtains a wavelet power spectrum.

5. In Paragraph 1, When the above instructions are executed individually or collectively by the at least one processor, the electronic device, Based on the average spectral power of the above user by frequency, the sum of the spectral power for each frequency domain of the blood flow regulation mechanism is obtained, and An electronic device that inputs the sum of acquired spectrum powers into the machine learning model to obtain information about the user's diabetes.

6. In Paragraph 5, The above blood flow regulation mechanism includes at least one of a cardiac mechanism and a myogenic mechanism, and When the above instructions are executed individually or collectively by the at least one processor, the electronic device, An electronic device that obtains the sum of spectral power for the frequency range of the cardiac or myogenic mechanism among the frequency components, based on the average spectral power of the above-mentioned user by frequency.

7. In Paragraph 6, When the above instructions are executed individually or collectively by the at least one processor, the electronic device, Obtain the sum of the spectral power for the frequency domains of the heart mechanism and muscle mechanism of the above user, and An electronic device that inputs the sum of the spectral powers into the machine learning model to obtain information about the user's diabetes.

8. In Paragraph 6, The above machine learning model is, It includes a machine learning model that uses the age of multiple people as a variable, and When the above instructions are executed individually or collectively by the at least one processor, the electronic device, An electronic device that obtains information about the user's diabetes by inputting the sum of the spectral powers in the frequency domain of the user's cardiac mechanism or muscle mechanism and the user's age information into the machine learning model.

9. In Paragraph 6, The above blood flow regulating mechanism further includes a respiratory mechanism, and The above machine learning model is, It includes a machine learning model that uses as a variable the sum of the spectral power obtained by summing the spectral power in the frequency domain of the cardiac mechanism, the spectral power in the frequency domain of the muscle mechanism, and the spectral power in the frequency domain of the respiratory mechanism. When the above instructions are executed individually or collectively by the at least one processor, the electronic device, An electronic device that obtains information about the user's diabetes by inputting the sum of the spectral powers for the frequency domains of the user's cardiac mechanism, muscle mechanism, and respiratory mechanism into the machine learning model.

10. In a method for controlling an electronic device, A step of acquiring the user's PPG (Photoplethysmography) signal; A step of performing preprocessing on the PPG signal using a filter; A step of obtaining a power spectrum by converting a preprocessed PPG signal into a frequency over time; A step of obtaining average spectrum power by frequency by calculating the average value of the acquired power spectrum over time; and A control method comprising the step of inputting the acquired average spectrum power by frequency into a machine learning model to obtain information about the user's diabetes.

11. In Paragraph 10, The step of acquiring the power spectrum above is, A control method comprising the step of converting the above-mentioned preprocessed PPG signal into a time-frequency axis using a wavelet transform and obtaining a wavelet power spectrum.

12. In Paragraph 10, The step of obtaining the average spectrum power for each frequency above is, A control method comprising the step of obtaining the sum of spectral powers for frequency domains according to blood flow control mechanisms based on the average spectral power for each frequency.

13. In Paragraph 12, The above blood flow regulation mechanism includes at least one of a cardiac mechanism and a myogenic mechanism, and The step of obtaining the average spectrum power for each frequency above is, A control method comprising the step of obtaining the sum of the spectral power for the frequency range of the cardiac mechanism or myogenic mechanism among the frequency components based on the average spectral power of the above-mentioned user by frequency.

14. In Paragraph 13, The above machine learning model is, It includes a machine learning model that uses the age of multiple people as a variable, and The step of obtaining information about the user's diabetes is, A control method comprising the step of obtaining information about the user's diabetes by inputting the sum of the spectral power in the frequency domain of the user's cardiac mechanism or muscle mechanism and the user's age information into the machine learning model.

15. In Paragraph 13, The above blood flow regulating mechanism further includes a respiratory mechanism, and The above machine learning model is, It includes a machine learning model that uses as a variable the sum of the spectral power obtained by summing the spectral power in the frequency domain of the cardiac mechanism, the spectral power in the frequency domain of the muscle mechanism, and the spectral power in the frequency domain of the respiratory mechanism. The step of obtaining information about the user's diabetes is, A control method comprising the step of inputting the sum of the spectral powers for the frequency domains of the cardiac mechanism, muscle mechanism, and respiratory mechanism of the user into the machine learning model to obtain information about the user's diabetes.

Images