Non-invasive blood glucose measurement device using PPG sensor and artificial neural networks
The non-invasive blood glucose measuring device with a ring-shaped and hand-held design, combined with an artificial neural network, addresses the issues of adhesion and comfort in wristwatch devices, achieving accurate and continuous glucose monitoring by using two PPG sensors to enhance measurement reliability.
Patent Information
- Application Number
- KR1020250057981
- Authority / Receiving Office
- KR · KR
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2025-05-02
- Publication Date
- 2026-11-10
AI Technical Summary
Existing non-invasive blood glucose measuring devices, particularly those worn as wristwatches, suffer from low adhesion to the skin, frequent contact and non-contact issues, leading to measurement inaccuracies and discomfort, and are not worn consistently due to their cumbersome nature, resulting in incomplete glucose monitoring.
A non-invasive blood glucose measuring device using photoplethysmography and an artificial neural network, comprising a ring-shaped first measuring device worn on the finger for continuous use and a second device for convenient measurement, both equipped with PPG sensors, to accurately acquire pulse wave signals and estimate glucose levels.
The device provides accurate and continuous blood glucose monitoring with enhanced user comfort and convenience, increasing measurement reliability by maintaining close contact and reducing variability due to different body parts, thereby improving estimation accuracy.
Smart Images

Figure PAT00001_ABST
Abstract
Description
Technology Field
[0001] The present invention relates to a non-invasive blood glucose measuring device and method using photoplethysmography and an artificial neural network, wherein bio-information is obtained from each pulse wave signal measured using two measuring devices and then input into an artificial neural network to estimate the blood glucose level, thereby enabling a more accurate estimation of the blood glucose level. Background Technology
[0003] Diabetes is a metabolic disease characterized by hyperglycemia resulting from impaired secretion or function of insulin, which is necessary for blood sugar control within the body. Chronic hyperglycemia caused by diabetes leads to damage and dysfunction of various organs in the body; in particular, it causes microvascular complications affecting the retina, kidneys, and nerves, as well as macrovascular complications such as arteriosclerosis, cardiovascular disease, and cerebrovascular disease, thereby increasing mortality rates.
[0004] However, diabetes can worsen or increase the incidence of complications due to blood sugar control, weight loss, and medication. Therefore, diabetic patients frequently monitor their blood glucose levels at home to manage their blood sugar.
[0005] Recently, a technology for estimating blood glucose levels using a non-invasive measuring device designed to be worn on the wrist in the form of a watch utilizing a PPG sensor has been disclosed.
[0006] Patent Publication No. 10-1512076 discloses a clock-type technology that estimates blood glucose using a conventional PPG sensor.
[0007] However, measuring devices designed to be attached to the wrist in the form of a watch, as described in the aforementioned patent, have low adhesion to the wrist skin and frequent contact and non-contact, which causes various error factors during PPG sensor measurement and limits accurate blood glucose measurement.
[0008] In addition, as it is in the form of a watch, there are many users who find it uncomfortable or reluctant to wear it on their wrist, so the frequency of use is relatively low. Furthermore, due to the cumbersome and inconvenient nature of use, it is not often worn all day or for more than 10 hours, which leads to the problem of not being able to continuously measure blood sugar.
[0009] In addition, conventionally, since the user's pulse wave signal is acquired using a single PPG sensor, there is a problem of low measurement accuracy because the pulse wave signal varies depending on the measurement location, even if the user moves or it is the same body part. In particular, since the pulse wave signal differs for each part of the user's body (e.g., wrist, finger, etc.), the accuracy is lowered as there is a difference in the estimated blood glucose value, which is the learning result, due to the different pulse wave signals for each body part. Prior art literature
[0011] Published Patent Application No. 10-2019-0063446 (June 7, 2019, Non-invasive blood glucose prediction system, blood glucose prediction method, and blood glucose sensor) Published Patent Application No. 10-2023-0163968 (December 1, 2023, Artificial intelligence-based blood glucose prediction system and method) The problem to be solved
[0012] The present invention was devised to solve the above problems and provides a non-invasive blood glucose measuring device and method using photoplethysmography and an artificial neural network, wherein two measuring devices are used to measure pulse wave signals, one measuring device is configured to be worn on the user's finger to measure pulse wave signals while being worn at all times, and the other measuring device is configured to consider the user's convenience to measure the user's pulse wave signals at any specific time, thereby acquiring bio-information from the two pulse wave signals and inputting it into an artificial neural network to estimate blood glucose levels.
[0013] Furthermore, the invention provides a non-invasive blood glucose measuring device and method using photoplethysmography and an artificial neural network, wherein the first measuring device is structured to be worn on the finger in a ring shape, thereby providing an aesthetic appearance even when exposed to the outside and maintaining a tight seal to enable accurate measurement and acquisition of pulse wave signals; and the second measuring device is structured such that the user places their hand on it and measures by placing their index finger in the finger placement groove, allowing for convenient and easy measurement of pulse wave signals in a comfortable state without effort, while enabling accurate detection and thereby increasing the accuracy of blood glucose level estimation. means of solving the problem
[0015] The non-invasive blood glucose measuring device using photoplethysmography and an artificial neural network according to the present invention for solving the above problem comprises: a first measuring device (10) formed in a ring shape for constant use on a user's finger and acquiring a user's pulse wave signal in real time through a first PPG sensor (14) provided on the inner surface; a second measuring device (20) which does not constantly wear on a part of the user's body but only raises and holds the user's hand when measuring the pulse wave signal and acquires a pulse wave signal through a second PPG sensor (23); and a user terminal (30) that is wired or wirelessly connected to the first measuring device (10) and the second measuring device (20) to receive the user's pulse wave signal acquired from the first measuring device (10) and the second measuring device (20). It comprises an artificial neural network (45) that is installed on the user terminal (30) or on a server (40) connected to the user terminal (30) via a communication network, and is pre-trained to estimate blood glucose levels as a result by inputting bio-information extracted from the user's pulse wave signal.
[0016] The first measuring device (10) comprises: a first measuring device body (11) formed in the shape of a ring to be worn on a user's finger; a wearing position indicator (12) formed on one side of the outer surface of the first measuring device body (11) to indicate the wearing position when the first measuring device body (11) is worn on a user's finger; a first PPG sensor (13) installed on one side of the inner surface of the first measuring device body (11) facing the inner surface of the finger, which detects the user's pulse wave signal while adhering to the inner skin of the finger when worn on the user's finger; and a first control unit (14) installed on the first measuring device body (11) to receive the pulse wave signal detected from the first PPG sensor (13) and transmit the pulse wave signal to the user terminal (30) via wireless communication.
[0017] At this time, a guide projection (15) is formed on the inner surface of the first measuring body (11) to protrude a certain length at spaced-apart positions on both sides of the first PPG sensor (13), and when the first measuring body (11) is worn on a finger, the guide projection (15) presses on both sides of the inner surface of the finger so that the central part of the inner surface of the finger protrudes and the skin adheres to the first PPG sensor (13).
[0018] The second measuring device (20) comprises: a second measuring device body (21) formed such that the outer surface is curved convexly from top to bottom so that the user’s hand can be placed on it and grasped, and a display part (21a) is formed on the upper surface to display measurement information; a finger seating groove (22) formed to a certain depth so that the user’s index finger is inserted and seated on the upper outer surface of the second measuring device body (21); a second PPG sensor (23) installed so that the inner surface of the user’s index finger contacts the inner surface of the finger seating groove (22) to detect the user’s pulse wave signal; and a second control unit (24) inside the second measuring device body (21) that receives the pulse wave signal detected from the second PPG sensor (23) and transmits the pulse wave information to the user terminal (30) via wired or wireless communication.
[0019] Here, it is preferable that a cover (25) be installed on the outer side of the finger seating groove (22) to cover the outer side of the user's finger seated in the finger seating groove (22) while blocking the external exposure of the second PPG sensor (23).
[0020] When training the artificial neural network described above, pulse wave signals are acquired for each body part of the user, bio-information is extracted from each pulse wave signal, and then training data is generated and the artificial neural network is trained for each body part. Then, bio-signals are extracted from two pulse wave signals acquired from different body parts, and the extracted bio-signals are input into the artificial neural network to estimate each blood glucose level. The estimated blood glucose levels are then summed to determine the average value as the final estimated blood glucose level.
[0021] The non-invasive blood glucose measurement method using photoplethysmography and an artificial neural network according to the present invention comprises: a first pulse wave signal acquisition step of acquiring a user's pulse wave signal in real time through a first PPG sensor of a first measuring device worn on the user's finger; a second pulse wave signal acquisition step of acquiring a user's pulse wave signal at a specific time through a second PPG sensor of a second measuring device configured to raise and grip the user's hand only when measuring the pulse wave signal; a step of transmitting the first pulse wave signal and the second pulse wave signal acquired from different parts of the user's body to a user terminal; a step of searching for and extracting the pulse wave signal acquired at the same time among the received first pulse wave signal and the second pulse wave signal; and a step of generating input data by extracting bio-information from the two extracted pulse wave signals, and then inputting the input data into a pre-trained artificial neural network to estimate the user's blood glucose level and obtain an estimated blood glucose level value. The method comprises the step of calculating a disease prediction probability and an alarm level corresponding to the obtained blood glucose level estimate, and then outputting the alarm level along with the blood glucose level estimate; wherein the alarm level is classified into reassurance, concern, caution, and suspicion, and is characterized by outputting "reassurance" when the disease prediction probability is 0% to 10%, "concern" when the disease prediction probability is 11% to 30%, "caution" when the disease prediction probability is 31% to 50%, and "suspicion" when the disease prediction probability is 51% or higher. Effects of the invention
[0023] According to the present invention, which is configured as described above, the structure of the first measuring instrument is a ring-shaped ring structure, thereby providing an aesthetically pleasing appearance and excellent ease of use. In particular, it can consistently maintain a state of close contact, allowing for accurate measurement of pulse wave signals.
[0024] In addition, the second measuring device is configured with a structure that allows measurement with the user's hand placed on it and the index finger resting on it, which not only enables convenient use in a comfortable posture but also allows for accurate measurement of pulse wave signals, thereby increasing the accuracy of blood glucose level estimation. Brief explanation of the drawing
[0026] FIG. 1 is a configuration diagram of a non-invasive blood glucose measuring device using photoplethysmography and an artificial neural network according to the present invention. FIG. 2 is a perspective view of a first measuring instrument according to the present invention, FIG. 3 is a cross-sectional view of a first measuring instrument according to the present invention, FIG. 4 is a state in which the central part of the inner surface of the finger is in close contact with the first PPG sensor by means of an induction projection while the first measuring device according to the present invention is worn on the finger. FIG. 5 is a front view of a second measuring device according to the present invention, FIG. 6 is a left side view seen from the right side of FIG. 5. FIG. 7 is a plan view of FIG. 5 seen from the upper side. FIG. 8 is a flowchart for explaining the process of estimating blood glucose levels based on pulse wave information measured by two measuring devices according to the present invention. Specific details for implementing the invention
[0027] Hereinafter, a non-invasive blood glucose measuring device and method using photoplethysmography and an artificial neural network according to a preferred embodiment of the present invention will be described in detail with reference to the attached drawings.
[0028] As described above, the non-invasive blood glucose measuring device using photoplethysmography and an artificial neural network according to the present invention acquires a user's pulse wave signal measured by two measuring devices equipped with PPG (photoplethysmography) sensors and transmits it to a user terminal, and the user terminal extracts bio-information from the received pulse wave signal and inputs it into an artificial neural network to more accurately estimate the user's blood glucose level, thereby enabling effective blood glucose management in a non-invasive manner.
[0029] One of the two measuring devices used to measure the pulse wave signal in the present invention is a measuring device that measures the pulse wave signal while being worn on the user's finger at all times, and is a first measuring device (10) that measures the user's pulse wave signal in real time through a first PPG sensor (14).
[0030] In addition, another measuring device is a second measuring device (20) that is used without being worn or attached to the user's body at all times, and is used only at any specific time to measure the user's pulse wave signal.
[0031] The first measuring device (10) is manufactured in the shape of a ring and measures blood sugar while being worn on the wearer's finger, so it has high adhesion to the skin and almost no movement, allowing it to accurately detect pulse wave signals.
[0032] This first measuring device (10) can solve the problem of inaccurately detected pulse signals due to insufficient contact force and frequent non-contact like conventional clock types, thereby increasing measurement accuracy and providing reliable result values to the user.
[0033] The first measuring device (10) is composed of a first measuring device body (11), a wearing position indicator (12), a first PPG sensor (13), and a first control unit (14) that communicates with a user terminal (20) via wired or wireless communication.
[0034] The first measuring body (11) is formed in the shape of a ring so that it can be worn on the wearer's finger.
[0035] The first measuring device body (11) may be provided with a separate storage case and configured to be wirelessly charged in the storage case. Due to its ring-shaped structure, it may be configured so that power is supplied and the device operates by taking it out of the storage case and placing it on the wearer's finger without having a separate power button.
[0036] In addition, the operation of the first control unit (14) can be configured to be turned on / off through a dedicated app installed on the user terminal (30).
[0037] At this time, the first measuring body (11) may be equipped with a lamp that flashes to emit light when turned ON by supplying power.
[0038] The wearing position indicator (12) is formed on one side of the outer surface of the first measuring device body (11) and is a part that indicates the wearing position when worn on a finger.
[0039] The above-mentioned wearing position indicator (12) can be worn so that it faces the outer upper side of the finger when the first measuring device body (11) is worn on the finger.
[0040] The first PPG sensor (13) is installed on one side of the inner surface of the first measuring body (11) facing the inner surface of the finger, and when worn on the user's finger, it comes into close contact with the inner skin of the finger and detects the user's pulse wave signal.
[0041] The pulse wave signal detected from the first PPG sensor (13) is transmitted to the first control unit (14).
[0042] The first control unit (14) is configured to include a control board and a microcontroller (MCU), and is installed inside the first measuring body (11) together with the first PPG sensor (13) to receive pulse wave information detected by the first PPG sensor (13) and transmit a pulse wave signal to the user terminal (30).
[0043] The first control unit (14) may be configured to directly transmit the pulse wave signal detected by the first PPG sensor (13) to the user terminal (30), or it may be configured to provide a separate memory to temporarily store the signal in the memory and then transmit it to the user terminal (30) via wired or wireless communication.
[0044] The control board constituting the first control unit may be formed flexibly due to the characteristics of the first measuring body (11) which is ring-shaped, and may be provided in an arc shape on one side of the internal space of the first measuring body (11) or formed by being rolled up in a ring shape throughout the entire internal space of the first measuring body (11).
[0045] At this time, a guide projection (15) is formed on the inner surface of the first measuring body (11) to protrude a certain length at spaced-apart positions on both sides of the first PPG sensor (13).
[0046] The above-mentioned induction protrusion (15) presses both sides of the inner surface of the finger when the first measuring body (11) is worn on the finger, thereby causing the central part of the inner surface of the finger to protrude and induce the skin to adhere to the first PPG sensor (13).
[0047] Therefore, since the first measuring body (11) does not shake or move while the first PPG sensor (13) is in close contact with the skin, the pulse wave signal can be accurately detected, and due to the increased accuracy of the pulse wave signal, it is possible to estimate blood glucose levels more accurately and increase measurement reliability.
[0048] The second measuring device (20) consists of a second measuring device body (21), a finger seating groove (22), a second PPG sensor (23), and a second control unit (24) that communicates with a user terminal (30) via wired or wireless communication.
[0049] The second measuring body (21) is formed to wrap around the measuring body while the user places their hand on it and holds it, and a display part (21a) is formed on the upper part to display measurement information, etc.
[0050] The display unit (21a) displays whether the on / off state and measurement state are on or off, or measurement information, and may also receive and display blood glucose levels and glycated hemoglobin levels analyzed by the terminal.
[0051] The finger seating groove (22) is formed on the upper outer surface of the second measuring body (21) and is formed with a certain depth and an area slightly larger than the finger area so that the user's index finger can be inserted and seated.
[0052] The outer surface of the second measuring body (21) having the finger seating groove (22) formed therein is formed to curve convexly toward the bottom, allowing the user to grip it while wrapping their hand in the most natural state possible, thereby improving finger contact during the measurement process and increasing ease of use, which can reduce stress.
[0053] On the lower side of one side of the second measuring body (21), an operating part (21b), such as a power button for power supply or a measurement button used when measuring pulse wave information, is formed, and a charging terminal for charging is installed.
[0054] The power button and measurement button are formed as push-type buttons that are easy to press.
[0055] Additionally, the second measuring body (21) may be equipped with a lamp that flashes to emit light when power is supplied due to the operation of the control unit (21b).
[0056] In addition to this, the above-mentioned control unit (21b) may be equipped with a reset button for resetting or initializing in case of an error.
[0057] The second PPG sensor (23) is installed on the inner surface of the finger seating groove (22) to be circuit-connected with the second control unit (24). Of course, to prevent the surface of the user's index finger from directly contacting the second PPG sensor (23), a transparent panel is separately installed on the outside of the second PPG sensor in the part where the second PPG sensor is installed.
[0058] When the user places their index finger on the finger placement groove (22), the second PPG sensor (23) detects the user's pulse signal as the inner surface of the index finger comes into contact, and the detected user's pulse signal is transmitted to the second control unit (24).
[0059] The second control unit (24) is configured to include a control board and a microcontroller (MCU), similar to the first control unit (14), and receives pulse wave information detected by the second PPG sensor (23) and transmits the pulse wave information to the user terminal (30).
[0060] At this time, a cover (25) is installed on the outer side of the finger seating groove (22).
[0061] The above cover (25) is intended to block external exposure of the second PPG sensor (23) installed on the inner surface of the finger seating groove (22), and at the same time, to cover the outer side of the finger when the user's finger is placed in the finger seating groove (22) for measuring the pulse wave signal.
[0062] Therefore, as external exposure is blocked, it is possible to prevent the decrease in measurement accuracy caused by external noise or interference, and thereby increase the measurement accuracy.
[0063] The user terminal (30) receives the acquired pulse wave signal by being connected to the first measuring device (10) and the second measuring device (20) via wired or wireless connection, and measures blood sugar using bio-information extracted from the pulse wave signal.
[0064] A user terminal (30) refers to a terminal equipped with a memory for storing a program, a microprocessor for executing the program to perform calculations and control, etc.
[0065] For example, a user terminal is a communication-capable terminal such as a personal computer (PC), laptop, personal digital assistant (PDA), mobile communication terminal, or tablet terminal, and for convenience of use, a smart pad or smartphone carried by the user can be used as a mobile communication terminal.
[0066] When a portable terminal such as a tablet, smart pad, or smartphone is used as a user terminal, a dedicated app (application) is installed to access the server (30) for the analysis of pulse wave signals, and the user's basic biometric information is entered into the dedicated app, and pulse wave signals detected by the first measuring device (10) and the second measuring device (20) are collected through the dedicated app.
[0067] An artificial neural network (45) is installed in the server (40) as a machine learning algorithm, and the user terminal (30) connects to the server (40) to extract bio-information from pulse wave signals received from the first measuring device (10) and the second measuring device (20), inputs this into the artificial neural network (45) to analyze it through machine learning to estimate blood glucose levels, and transmits measurement information including the estimated values to the user terminal (30) to display on the screen.
[0068] Accordingly, when a pulse wave signal is measured by the first measuring device (10) and the second measuring device (20) while the dedicated app is running on the user terminal (30), bio-information is extracted using the measured pulse wave signal, and the extracted bio-information is input into an artificial neural network (45) to estimate blood sugar levels and determine the alarm level, and the result is displayed on the screen, and the user checks their health status, such as whether they have diabetes, based on the information displayed on the screen of the user terminal.
[0069] Meanwhile, in the present invention, an artificial neural network is installed on a server (40) and configured to be accessed via a dedicated app from a user terminal (30); however, if a server (40) is not separately provided, an artificial neural network program may be installed on the user terminal (30) to extract the user's bio-information using pulse wave signals from the user terminal itself, thereby enabling the measurement of blood sugar levels and the presence of disease (diabetes), etc.
[0070] In the present invention, pulse wave signals are measured by two measuring devices. As previously explained, the first measuring device (10) is worn on the user's finger at all times and measures the user's pulse wave signal in real time through the first PPG sensor.
[0071] The second measuring device (15) is not worn on the user's body at all times, and the user's finger is placed in the finger seating groove (22) and pressed against the second PPG sensor only when measuring the pulse wave signal to acquire the pulse wave signal.
[0072] Therefore, pulse wave signals are measured at the first measuring device and the second measuring device, respectively, in different parts of the body, and accordingly, the measurement accuracy of the pulse wave signal can be increased compared to measuring the pulse wave signal using a single measuring device in the past.
[0073] Since the two pulse wave signals acquired from the first and second measuring instruments contain noise, it would be desirable to undergo a noise removal process.
[0074] When removing noise, the amplitude of the pulse wave signal is first adjusted through a scaling process, and only the pulse wave signal corresponding to a preset range of minimum and maximum amplitude values is extracted from the scaled pulse wave signal.
[0075] The preset minimum and maximum amplitude values are average values extracted from the pulse wave signals of diabetic patients and can be set differently depending on the surrounding environment.
[0076] After setting the noise-removed pulse wave signal as the pulse wave signal reference value, bio-information is extracted based on this to estimate blood glucose levels.
[0077] The process of predicting diabetes by estimating blood glucose levels involves the following steps: pulse wave signal measurement and acquisition -> noise removal -> bio-information extraction -> input and training to an artificial neural network -> blood glucose level estimation -> alarm level output.
[0078] Pulse wave signal measurement acquires a first pulse wave signal continuously measured by a first measuring device and a second pulse wave signal of the corresponding body part arbitrarily measured by a second measuring device, and transmits them to a user terminal.
[0079] Among the two acquired pulse wave signals, each pulse wave signal acquired at the same time is searched, and noise is removed from the two pulse wave signals acquired at the same time.
[0080] The user's biometric information is extracted from the noise-removed pulse signal, and input data is generated based on the extracted biometric information and input into a pre-trained artificial neural network for machine learning.
[0081] Through machine learning, the user's blood glucose level estimate is calculated, and the disease prediction probability and alarm level corresponding to the calculated blood glucose level estimate are calculated and the blood glucose level estimate and alarm level are output, and the user can predict diabetes through their blood glucose level and alarm level.
[0082] Here, when training the artificial neural network for estimating the blood glucose level, a training model is constructed separately for each body part of the user. That is, pulse wave signals are acquired for each body part, and bio-information containing the acquired pulse wave signals is generated as training data for each part, thereby constructing an artificial neural network for each body part.
[0083] The reason for doing this is that pulse wave signals vary by body part, and these differing signals can cause significant errors in the results (blood sugar levels) during machine learning; therefore, this measure aims to reduce such errors and improve the accuracy of the results.
[0084] Biosignals are extracted from each pulse wave signal acquired from each body part by the first and second measuring devices to generate input data, which is then input into the artificial neural network to estimate the blood glucose level of the corresponding body part, and the estimated blood glucose levels are summed to determine the average value as the final estimated blood glucose level.
[0085] The above artificial neural network was trained by collecting 10,000 biological data sets for training from diabetic patients and normal people, respectively.
[0086] The artificial neural network above ultimately outputs blood glucose levels, calculates disease prediction probabilities based on the output values, and determines the alarm level.
[0087] The above alert level can be classified as "Safe," "Concern," "Caution," or "Suspicious," and is determined as "Safe" when the disease prediction probability is 0% to 10%, "Concern" when the disease prediction probability is 11% to 30%, "Caution" when the disease prediction probability is 31% to 50%, and "Suspicious" when the disease prediction probability is 51% or higher.
[0088] Through these steps, the presence of a disease in the user can be diagnosed, and by estimating blood glucose levels with high accuracy and providing them to the user's terminal, the user can effectively manage their blood glucose. Explanation of the symbols
[0090] 10: First measuring instrument 11: First measuring instrument body 12: Wearing position indicator 13: 1st PPG sensor 14: First control unit 15: Guiding projection 20: Second measuring instrument 21: Second measuring instrument body 21a: Display unit 21b: Control section 22: Finger resting groove 23: 2nd PPG sensor 24: Second control unit 25: Cover 30: User terminal 40: Server 45: Artificial Neural Networks
Claims
Claim 1 A non-invasive blood glucose measuring device using photoplethysmography and an artificial neural network, characterized by comprising: a first measuring device (10) formed in a ring shape for constant use on the user's finger and acquiring the user's pulse wave signal in real time through a first PPG sensor (14) provided on the inner surface; a second measuring device (20) which acquires the pulse wave signal by acquiring the user's pulse wave signal through a second PPG sensor (23) while holding the user's hand only when measuring the pulse wave signal, without the device being constantly worn on the user's body part; a user terminal (30) connected wired or wirelessly to the first measuring device (10) and the second measuring device (20) to receive the user's pulse wave signal acquired from the first measuring device (10) and the second measuring device (20); and an artificial neural network (45) that is installed on the user terminal (30) or installed on a server (40) connected to the user terminal (30) via a communication network, and is pre-trained to estimate the blood glucose level as a result by inputting bio-information extracted from the user's pulse wave signal. Claim 2 A non-invasive blood glucose measuring device using photoplethysmography and an artificial neural network, characterized in that, in claim 1, the first measuring device (10) comprises: a first measuring device body (11) formed in the shape of a ring to be worn on a user's finger; a wearing position indicator (12) formed on one side of the outer surface of the first measuring device body (11) to indicate the wearing position when the first measuring device body (11) is worn on a user's finger; a first PPG sensor (13) installed on one side of the inner surface of the first measuring device body (11) facing the inner surface of the finger, which detects the user's pulse wave signal while adhering to the inner skin of the finger when worn on the user's finger; and a first control unit (14) installed on the first measuring device body (11) to receive the pulse wave signal detected from the first PPG sensor (13) and transmit the pulse wave signal to the user terminal (20) via wireless communication. Claim 3 In paragraph 2, a non-invasive blood glucose measuring device using photoplethysmography and an artificial neural network is characterized in that, on the inner surface of the first measuring body (11), a guiding projection (15) is formed to protrude a certain length at a spaced-apart position on both sides of the first PPG sensor (13), and when the first measuring body (11) is worn on a finger, the guiding projection (15) presses on both sides of the inner surface of the finger so that the central part of the inner surface of the finger protrudes and the skin adheres to the first PPG sensor (13). Claim 4 In claim 1, the second measuring device (20) is characterized by comprising: a second measuring device body (21) having an outer surface that is formed to be convexly curved from top to bottom so as to place the user's hand on it for gripping, and a display part (21a) formed on the upper surface to display measurement information; a finger seating groove (22) formed to a certain depth so that the user's index finger is inserted and seated on the upper outer surface of the second measuring device body (21); a second PPG sensor (23) installed so that the inner surface of the user's index finger contacts the inner surface of the finger seating groove (22) to detect the user's pulse wave signal; and a second control unit (24) inside the second measuring device body (21) that receives the pulse wave signal detected from the second PPG sensor (23) and transmits the pulse wave information to the user terminal (30) via wired or wireless communication. Claim 5 A non-invasive blood glucose measuring device using photoplethysmography and an artificial neural network, characterized in that, in claim 4, a cover (25) is installed on the outer side of the finger seating groove (22) to cover the outer side of the user's finger seated in the finger seating groove (22) while blocking the external exposure of the second PPG sensor (23). Claim 6 A non-invasive blood glucose measuring device using photoplethysmography and an artificial neural network, characterized in that, in any one of claims 1 to 5, when training the artificial neural network, the device classifies the user's body parts and acquires pulse wave signals for each body part, extracts bio-information from each pulse wave signal and generates training data, trains the artificial neural network for each body part, extracts bio-signals from two pulse wave signals acquired from different body parts, inputs the extracted bio-signals into the artificial neural network to estimate each blood glucose level, sums the estimated blood glucose levels, and determines the average value as the final estimated blood glucose level. Claim 7 A first pulse wave signal acquisition step of acquiring a user's pulse wave signal in real time through a first PPG sensor of a first measuring device worn on the user's finger; a second pulse wave signal acquisition step of acquiring a user's pulse wave signal at a specific time through a second PPG sensor of a second measuring device configured to raise and grip the user's hand only when measuring the pulse wave signal; a step of transmitting the first pulse wave signal and the second pulse wave signal acquired from different parts of the user's body to a user terminal; a step of searching for and extracting the pulse wave signal acquired at the same time among the received first pulse wave signal and the second pulse wave signal; a step of generating input data by extracting bio-information from the two extracted pulse wave signals, and then inputting the input data into a pre-trained artificial neural network to estimate the user's blood glucose level and obtain an estimated blood glucose level value; a step of calculating a disease prediction probability and an alarm level corresponding to the obtained estimated blood glucose level value, and then outputting the alarm level along with the estimated blood glucose level value; wherein the alarm level is classified into reassurance, concern, caution, and suspicion, and the disease prediction probability is 0% to A non-invasive blood glucose measurement method using photoplethysmography and an artificial neural network, characterized by determining and outputting "Reassured" when the probability of the disease prediction is 10%, "Interest" when the probability of the disease prediction is 11% to 30%, "Caution" when the probability of the disease prediction is 31% to 50%, and "Suspicion" when the probability of the disease prediction is 51% or higher.
