Closed-loop control device and method for optimized PTH amount with blood ionized calcium estimation using non-invasive multi-biosignal-based field edge AI

Abstract

In order to address the problems associated with conventional blood calcium measurement, such as the difficulty of real-time continuous monitoring, significant patient discomfort, difficulty in implementing a wearable form factor, and limitations in establishing a closed-loop control system; the problem that there is virtually no technology to non-invasively estimate blood calcium status and link it with PTH control; the problem of discontinuity caused by intermittent blood sampling, the need for hospital visits, and the inability to predict risks in real time; the problem that while low calcium levels induce various biosignal changes such as QT prolongation, increased nerve excitability, muscle spasms, and autonomic nervous system changes, there is no system to analyze these changes in an integrated manner; and the problem that although an artificial pancreas currently exists, a calcium-PTH homeostasis-based artificial parathyroid system has not been commercialized, the present invention comprises a smart ring-type wearable sensor unit (100), an externally worn wearable patch unit (200), a wrist-attached smartwatch (300), and a field edge type By configuring the AI ​​module (400), the calcium balance smart application module (500), and the photo-stimulated response type body-implantable PTH secreting object (600), the ECG electrode unit, the ECG amplifier unit, and the ECG filter unit are linked through the external wearable patch unit, thereby improving the SNR (Signal to Noise Ratio) of the electrocardiogram signal by 80% compared to the existing one, and by performing HRV analysis based on the RR interval, the ability to analyze the autonomic nervous state can be improved by 2 to 4 times compared to the existing one, and by using the QT interval and QTc analysis function, the ability to detect QT prolongation occurring in hypocalcemia can be improved by 3 to 5 times compared to the existing one, and by integrating the heart rate, RR interval, HRV, QT interval, and QTc analysis, the early prediction performance of hypocalcemia risk can be improved by 80% compared to the existing one, and by configuring the field edge type AI module, the QT interval, HRV, pulse waveform,By simultaneously indexing the neural excitability index and having the blood calcium status estimation control unit analyze it based on AI, the reliability of estimating the blood ionized calcium status can be improved by 80% compared to the simple ECG-based QT prolongation judgment method; by having the low calcium risk prediction control unit analyze not only the current calcium status but also recent biosignal trends, QTc change rate, HRV decrease, and increased neural excitability together, the early prediction performance of low calcium risk can be improved by 2 to 5 times compared to the existing simple notification method; by having the edge AI processor unit execute AI inference, risk prediction, and closed-loop control algorithms on-site, the impact of communication delay can be reduced to less than 70% compared to cloud server-dependent analysis; and by having the edge wireless communication unit link with a wrist-worn smartwatch and the Calcium Balance smart app module, the closed-loop delivery capability of calcium status display, PTH infusion control, and risk notifications can be improved by 1.5 to 3 times compared to the existing unidirectional monitoring method; and with the configuration of the Calcium Balance smart app module, the estimated blood ionized calcium status, low calcium risk, PTH infusion status and It can provide real-time biosignal status, and through risk color display, vibration, warning sounds, voice guidance, and emergency pop-up functions, the user's awareness time for low-calcium risk situations can be reduced to less than 80% compared to existing methods, and emergency responsiveness can be improved by 1.5 to 3 times compared to existing methods. Furthermore, since long-term calcium change trends and PTH administration history can be managed through calcium estimate storage units, risk history storage units, PTH injection history storage units, and biosignal summary storage units, the clinical utility for medical staff can be improved by 80% compared to existing methods. Additionally, by configuring a photostimulation-responsive implantable PTH-secreting object that includes functional cells cultured from human-derived cell lines, an organoid scaffold, and a biocompatible spherical body, it can generate and release implantable PTH in response to photostimulation, and by culturing functional cells into a three-dimensional organoid structure, cell viability can be improved by 80% compared to existing methods.The purpose is to provide a non-invasive, multi-biosignal-based, on-site edge AI-based implantable PTH secretion closed-loop control device and method that can improve PTH generation efficiency by 1.5 to 3 times compared to existing methods, and improve the efficiency of a photostimulation-based closed-loop control sequence consisting of photostimulation -> functional cell activation -> PTH generation -> PTH release -> blood calcium recovery by 80% compared to existing methods by dispersing and implanting multiple photostimulation-responsive implantable PTH-secreting objects within the wrist muscle.

Description

Technology Field

[0001] The present invention relates to a closed-loop control device and method for secreting implantable PTH based on an estimation of blood ionized calcium using a non-invasive multi-biometric signal-based field edge AI, capable of closed-loop control to output a photo-stimulation response signal to secrete implantable PTH toward a photo-stimulation response type body-implantable PTH secreting object using biosignal data sensed by attaching or wearing it on a user's body to estimate the blood calcium status and using the estimated blood calcium status to output a photo-stimulation response signal to secrete implantable PTH. Background Technology

[0003] Blood calcium (Ca) is a key electrolyte essential for nerve transmission, muscle contraction, cardiac conduction, cell signaling, and maintaining bone metabolism.

[0004] In particular, blood ionized calcium (Ca 2+ : ionized calcium) maintains homeostasis within a very narrow range through complex mechanisms such as parathyroid hormone (PTH), vitamin D, intestinal function, bone function, and kidney function.

[0005] Therefore, continuous monitoring of blood calcium is important for patients with hypoparathyroidism, chronic renal failure, post-thyroid and parathyroid surgery, dialysis patients, patients with bone metabolic disorders, and patients with hypocalcific arrhythmia.

[0006] Currently, blood calcium measurement relies mostly on blood sampling-based blood tests, invasive electrode measurements, and hospital-based analytical equipment.

[0007] However, there were problems such as difficulty in real-time continuous monitoring, significant patient discomfort, difficulty in implementing a wearable form, and limitations in establishing a closed-loop control system.

[0008] On the other hand, while smart rings, wearable ECGs, PPGs, bioimpedance, and AI-based digital biomarker technologies have recently been advancing, there is virtually no technology to non-invasively estimate blood calcium levels and link this to PTH control.

[0009] The reason is that direct non-invasive measurement of blood Ca is very difficult because it has a weak optical signal, a protein-bound form, and is greatly affected by pH.

[0010] Furthermore, the current method had the problem of discontinuity due to issues such as intermittent blood sampling, the need for hospital visits, and the inability to predict risks in real time.

[0011] Furthermore, while low calcium levels induce various changes in biosignals such as QT prolongation, increased neuronal excitability, muscle spasms, and autonomic nervous system alterations, there was a problem in the absence of a system to comprehensively analyze these changes.

[0012] Furthermore, while artificial pancreases currently exist, there was a problem in that calcium-PTH homeostasis-based artificial parathyroid systems were not commercialized. Prior art literature

[0014] Korean Patent Publication No. 10-2026-0054764 The problem to be solved

[0015] To solve the above problems, the present invention senses first non-invasive biosignal data regarding PPG (Photoplethysmography), blood oxygen saturation (SpO2), bioimpedance, skin temperature, tremor, and electromyography (EMG) through a smart ring-type wearable sensor unit, senses second non-invasive biosignal data regarding ECG (Electrocardiography) signals, heart rate, and heart rate variability flowing from the surface of the chest skin by attaching to the surface of the chest skin through an externally worn wearable patch unit, and uses the first non-invasive biosignal data, second non-invasive biosignal data, and 3-axis acceleration data through a field edge-type AI module to calculate the QT (Q wave to T wave) interval, HRV (Heart Rate Variability), pulse waveform, and neural excitability index, estimates the blood calcium status, and the estimated The purpose is to provide a closed-loop control device and method for the secretion of a body-implantable PTH based on an on-site edge AI that estimates blood ionized calcium, which can transmit the blood calcium status to a calcium balance smart application module and output a photo-stimulation response signal to secrete body-implantable PTH using the estimated blood calcium status to a photo-stimulation response type body-implantable PTH secreting object, and through the calcium balance smart application module, display the estimated blood calcium status, risk level, PTH injection status, and biosignal status on a smart device screen and perform an emergency notification role in case of low calcium risk, and through the photo-stimulation response type body-implantable PTH secreting object, automatically secrete parathyroid hormone (PTH) into the user's body according to the control signal of the on-site edge AI module. means of solving the problem

[0017] To achieve the above objective, the body-implantable PTH secretion closed-loop control device of the blood ionized calcium estimation formula of the non-invasive multi-biosignal-based field edge AI according to the present invention

[0018] This is achieved by configuring it to control a closed loop by using biosignal data sensed by attaching or wearing it on the user's body to estimate the blood calcium status, and outputting a photo-stimulation response signal to secrete body-implantable PTH towards a photo-stimulation response type body-implantable PTH secreting object using the estimated blood calcium status.

[0020] More specifically, the above-mentioned non-invasive multi-biosignal-based field edge AI blood ionized calcium estimation formula implantable PTH secretion closed-loop controller,

[0021] A smart ring-type wearable sensor unit (100) formed in a ring structure, worn on a finger to sense first non-invasive biosignal data regarding PPG (Photoplethysmography), blood oxygen saturation (SpO2), bioimpedance, skin temperature, tremor, and electromyography (EMG), and to transmit the sensed first non-invasive biosignal data to a wrist-mounted smartwatch, and

[0022] An external wearable patch part (200) that attaches to the surface of the chest skin to sense second non-invasive biosignal data regarding ECG (Electrocardiography) signals, heart rate, and heart rate variability flowing from the surface of the chest skin, and transmits the sensed second non-invasive biosignal data to a wrist-attached smartwatch, and

[0023] A wrist-mounted smartwatch (300) that is worn on the user's wrist to measure 3-axis acceleration data regarding the user's moving posture, collects the first non-invasive biosignal data transmitted from the smart ring-type wearable sensor unit and the second non-invasive biosignal data transmitted from the external wearable patch unit, transmits them to the field edge-type AI module via a wireless communication network, receives a photo-stimulation response signal from the field edge-type AI module, and photostimulates the photo-stimulation response type body-implantable PTH-secreting object implanted inside the PTH minor neck muscle;

[0024] A field edge type AI module (400) located at a hospital or public institution within a short distance of 1m to 5000m relative to a wrist-worn smartwatch, connected to the wrist-worn smartwatch via a wireless communication network, and using the first non-invasive biosignal data, second non-invasive biosignal data, and 3-axis acceleration data transmitted from the wrist-worn smartwatch to calculate the QT (Q wave to T wave) interval, HRV (Heart Rate Variability), Pulse waveform, and neural excitability index, then estimates the blood calcium state, transmits the estimated blood calcium state to the calcium balance smart app module, and uses the estimated blood calcium state to output a photostimulation response signal that causes the body-implantable PTH to be secreted toward a photostimulation response type body-implantable PTH secreting object located in the wrist muscle, thereby controlling a closed loop;

[0025] A calcium balance smart app module (500) that is activated on a smart device app for patients and a smart device app for guardians, is connected to a field edge-type AI module and a wireless network, displays the estimated blood calcium status, risk level, the state of being automatically secreted into the body of the PTH user, and the biosignal status on the smart device screen, and performs the role of an emergency notification when there is a risk of low calcium, and

[0026] It is characterized by being composed of a photo-stimulation-responsive body-implantable PTH-secreting object (600) that is implanted in the user's wrist muscle and secretes body-implantable PTH (Parathyroid Hormone) according to the photo-stimulation signal of a wrist-attached smartwatch. Effects of the invention

[0028] As explained above, in the present invention

[0030] First, through an external wearable patch, the ECG electrode, ECG amplifier, and ECG filter are linked to improve the Signal to Noise Ratio (SNR) of the electrocardiogram signal by 80% compared to the existing method, and the ability to analyze the autonomic nervous state can be improved by 2 to 4 times compared to the existing method by performing HRV analysis based on the RR interval, and the ability to detect QT prolongation occurring in hypocalcemia can be improved by 3 to 5 times compared to the existing method through QT interval and QTc analysis functions, and the early prediction performance of hypocalcemia risk can be improved by 80% compared to the existing method by integrating the analysis of heart rate, RR interval, HRV, QT interval, and QTc.

[0032] Second, by configuring a wrist-mounted smartwatch, the time axis data of the smart ring-type wearable sensor unit, the externally worn wearable patch unit, and the wrist-mounted smartwatch's own sensor are aligned to the same reference time through the time synchronization unit, thereby reducing the time error between sensors to less than 70% compared to the existing method. Additionally, by generating the synchronized data into an integrated data packet containing a sensor ID, timestamp, error detection code, and compression information through the data packet generation unit, the data interpretation speed of the field edge AI module can be improved by 1.5 to 3 times compared to the existing method. Furthermore, by the photo-stimulation unit stimulating the photo-stimulation-responsive body-implantable PTH-secreting object inside the wrist muscle according to the photo-stimulation response signal received from the field edge AI module, the PTH secretion induction response speed can be improved by 80% compared to the existing drug injection-centered method.

[0034] Third, by configuring an on-site edge AI module, the QT interval, HRV, pulse waveform, and neural excitability index are simultaneously indexed. Since the blood calcium status estimation control unit analyzes this based on AI, the reliability of estimating the blood ionized calcium status can be improved by 80% compared to the simple ECG-based QT prolongation judgment method. Furthermore, by the low calcium risk prediction control unit analyzing not only the current calcium status but also recent biosignal trends, QTc change rates, HRV decrease, and increased neural excitability, the early prediction performance of low calcium risk can be improved by 2 to 5 times compared to the existing simple notification method. Additionally, by the edge AI processor unit executing AI inference, risk prediction, and closed-loop control algorithms on-site, the impact of communication delay can be reduced to less than 70% compared to cloud server-dependent analysis. Moreover, by the edge wireless communication unit linking with wrist-worn smartwatches and the Calcium Balance smart app module, the closed-loop delivery capability for calcium status display, PTH injection control, and risk notifications can be improved by 1.5 to 3 times compared to the existing unidirectional monitoring method. Based on the low calcium risk prediction results... By generating a photo-stimulation response signal to stimulate a photo-stimulation-responsive body-implantable PTH-secreting object inside the wrist muscle, the PTH secretion induction response speed can be improved by 1.5 to 3 times compared to the existing manual treatment method.

[0036] Fourth, by configuring the Calcium Balance smart application module, the estimated blood ionized calcium status, low calcium risk, PTH infusion status, and biosignal status can be provided in real time. Through risk color display, vibration, warning sound, voice guidance, and emergency pop-up functions, the user's awareness time regarding low calcium risk situations can be reduced to less than 80% compared to the existing system, and emergency responsiveness can be improved by 1.5 to 3 times compared to the existing system. Additionally, since long-term calcium change trends and PTH administration history can be managed through the calcium estimate value storage unit, risk history storage unit, PTH infusion history storage unit, and biosignal summary storage unit, the clinical utility for medical staff can be improved by 80% compared to the existing system.

[0038] Fifth, by configuring a photo-responsive body-implantable PTH-secreting object, including functional cells cultured from human-derived cell lines, an organoid scaffold, and a biocompatible spherical body, it is possible to generate and release body-implantable PTH in response to photo-stimulation. By culturing the functional cells into a three-dimensional organoid structure, the cell viability can be improved by 80% compared to the existing method, and the PTH generation efficiency can be improved by 1.5 to 3 times compared to the existing method. Furthermore, by dispersing and implanting multiple photo-responsive body-implantable PTH-secreting objects within the wrist muscle, the photo-stimulation-based closed-loop control efficiency, consisting of photo-stimulation -> functional cell activation -> PTH generation -> PTH release -> blood calcium recovery, can be improved by 80% compared to the existing method. Brief explanation of the drawing

[0040] FIG. 1 is a block diagram illustrating the components of a body-implantable PTH secretion closed-loop control device (1) with a blood ionized calcium estimation formula of a non-invasive multi-biosignal-based field edge AI according to the present invention. FIG. 2 is a configuration diagram illustrating the components of a body-implantable PTH secretion closed-loop control device (1) with a blood ionized calcium estimation formula of a non-invasive multi-biosignal-based field edge AI according to the present invention. FIG. 3 is a block diagram illustrating the components of a smart ring-type wearable sensor unit according to the present invention, FIG. 4 is a perspective view illustrating the components of a smart ring-type wearable sensor unit according to the present invention. FIG. 5 is an exploded perspective view illustrating the components of a smart ring-type wearable sensor unit according to the present invention. FIG. 6 is an exemplary embodiment illustrating the components of a smart ring-type wearable sensor unit according to the present invention. FIG. 7 is a block diagram illustrating the components of an externally worn wearable patch according to the present invention. FIG. 8 is a configuration diagram illustrating the components of an externally worn wearable patch portion according to the present invention. FIG. 9 is an internal configuration diagram illustrating the components of an externally worn wearable patch portion according to the present invention. FIG. 10 is an exemplary illustration showing the components of an externally worn wearable patch portion according to the present invention. FIG. 11 is a block diagram illustrating the components of a wrist-mounted smartwatch according to the present invention, FIG. 12 is a perspective view illustrating the components of a wrist-mounted smartwatch according to the present invention. FIG. 13 is an internal configuration diagram illustrating the components of a wrist-mounted smartwatch according to the present invention. FIG. 14 is a rear perspective view illustrating the components of a wrist-mounted smartwatch according to the present invention, FIG. 15 is an exemplary embodiment illustrating the components of a wrist-mounted smartwatch according to the present invention. FIG. 16 is a block diagram illustrating the components of a field edge-type AI module according to the present invention, FIG. 17 is a perspective view illustrating the components of a field edge-type AI module according to the present invention. FIG. 18 is a rear perspective view illustrating the components of a field edge-type AI module according to the present invention, FIG. 19 is an exploded perspective view illustrating the components of a field edge-type AI module according to the present invention. FIG. 20 is a block diagram illustrating the components of a feature value operation control unit according to the present invention, FIG. 21 is a block diagram illustrating the components of a QT interval operation control unit according to the present invention, FIG. 22 is a block diagram illustrating the components of an HRV operation control unit according to the present invention, FIG. 23 is a block diagram illustrating the components of a pulse waveform analysis control unit according to the present invention. FIG. 24 is a block diagram illustrating the components of a neural excitability analysis control unit according to the present invention, FIG. 25 is a block diagram illustrating the components of a blood calcium status estimation control unit according to the present invention, FIG. 26 is a block diagram illustrating the components of a calcium balance smart application module according to the present invention, FIG. 27 is an exemplary embodiment illustrating the components of a calcium balance smart application module according to the present invention. FIG. 28 is an exemplary embodiment illustrating the components of the main activity unit according to the present invention, FIG. 29 is an exemplary embodiment illustrating the components of a calcium state activity section according to the present invention. FIG. 30 is an exemplary embodiment illustrating the components of a biosignal monitoring activity unit according to the present invention. FIG. 31 is an exemplary embodiment illustrating the components of the PTH injection state activity section according to the present invention. FIG. 32 is an exemplary embodiment illustrating the components of a risk notification activity unit according to the present invention, FIG. 33 is an exemplary embodiment illustrating the components of a guardian monitoring activity unit according to the present invention. FIG. 34 is an exemplary embodiment illustrating the components of a setting activity unit according to the present invention, FIG. 35 is a configuration diagram illustrating the components of a calcium balance type service control unit according to the present invention. FIG. 36 is a configuration diagram illustrating the components of a calcium balance type broadcast receiver control unit according to the present invention. FIG. 37 is a configuration diagram illustrating the components of a calcium balance type content provider control unit according to the present invention. FIG. 38 is a configuration diagram illustrating the components of a calcium balance type UI smart control unit according to the present invention. FIG. 39 is a configuration diagram illustrating the components of a photo-stimulation-responsive body-implantable PTH-secreting object according to the present invention. FIG. 40 is an embodiment illustrating a photo-stimulation-responsive body-implantable PTH-secreting object according to the present invention being positioned in the wrist muscle area, which is the location of a wrist-attached smartwatch. FIG. 41 is an exemplary diagram illustrating the operation process of a body-implantable PTH secretion closed-loop control device with a blood ionized calcium estimation formula of a non-invasive multi-biosignal-based field edge AI according to the present invention. FIG. 42 is a flowchart illustrating a body-implantable PTH secretion closed-loop control method using a blood ionized calcium estimation formula of a non-invasive multi-biosignal-based field edge AI according to the present invention. FIG. 43 is a flowchart illustrating a specific process of sensing first non-invasive biosignal data regarding PPG (Photoplethysmography), blood oxygen saturation (SpO2), bioimpedance, skin temperature, tremor, and electromyography (EMG) by wearing a smart ring-type wearable sensor unit according to the present invention on a finger, and transmitting the sensed first non-invasive biosignal data to a wrist-mounted smartwatch. FIG. 44 is a flowchart illustrating a specific process of sensing second non-invasive biosignal data regarding ECG (Electrocardiography) signals, heart rate, and heart rate variability flowing from the surface of the chest skin by attaching to the surface of the chest skin through an external wearable patch portion according to the present invention, and transmitting the sensed second non-invasive biosignal data to a wrist-mounted smartwatch. FIG. 45 is a flowchart illustrating a specific process of measuring 3-axis acceleration data regarding a user's moving posture by wearing a wrist-mounted smartwatch according to the present invention on the user's wrist, collecting first non-invasive biosignal data transmitted from a smart ring-type wearable sensor unit and second non-invasive biosignal data transmitted from an external wearable patch unit, and transmitting them to a field edge-type AI module via a wireless communication network. FIG. 46 is a flowchart illustrating a specific process for estimating blood calcium status after calculating QT (Q wave to T wave) interval, HRV (Heart Rate Variability), Pulse waveform, and neural excitability index using first non-invasive biosignal data, second non-invasive biosignal data, and 3-axis acceleration data transmitted from a wrist-worn smartwatch through a field edge-type AI module according to the present invention. FIG. 47 is a flowchart illustrating a specific process of transmitting an estimated blood calcium status to a calcium balance smart application module through a field edge-type AI module according to the present invention. FIG. 48 is a flowchart illustrating a specific process of outputting a photostimulation response signal to secrete body-implantable PTH toward a wrist-attached smartwatch located in the wrist muscle, using the estimated blood calcium status through a field edge-type AI module according to the present invention. Specific details for implementing the invention

[0042] First, the closed-loop control described in this invention is a control method in which a device continuously measures the user's condition, makes an automatic judgment based on the measured values, and then adjusts the treatment action. It measures first non-invasive biosignal data, second non-invasive biosignal data, and 3-axis acceleration data through a smart ring-type wearable sensor unit + an externally worn wearable patch unit + a wrist-mounted smartwatch, and through a field edge-type AI module, blood Ca 2+ Estimating the state and determining the risk of hypocalcemia, outputting a photostimulation response signal to secrete body-implantable PTH toward the wrist-worn smartwatch, and causing the body-implantable PTH secreting object to secrete body-implantable PTH through the photostimulation signal, Ca 2+ It refers to measuring the first non-invasive biosignal data, the second non-invasive biosignal data, and the 3-axis acceleration data again through the smart ring-type wearable sensor unit + the external wearable patch unit + the wrist-attached smartwatch after inducing recovery.

[0044] In addition, the ideal artificial parathyroid gland described in the present invention, without external light stimulation, blood calcium (Ca 2+ Although it autonomously secretes PTH by detecting only the concentration, in actual treatment settings, photostimulation is used as an "auxiliary control measure" to prepare for stabilization immediately after transplantation, acute hypocalcemia, rapid calcium drop after surgery, and the risk of seizures, rather than being "essential."

[0045] Accordingly, the photo-stimulation-responsive body-implantable PTH-secreting object according to the present invention performs the role of an "auxiliary control means" for secreting body-implantable PTH in response to a photo-stimulation signal from a wrist-attached smartwatch.

[0047] Next, regarding the first and second non-invasive biosignal data described in the present invention, the non-invasive biosignal refers to a physiological signal that can be measured on the skin surface of a finger, wrist, or chest without inserting a needle or electrode catheter into the human body, and is bio-information that can be obtained via skin contact or non-contact methods without blood sampling, surgery, or the insertion of sensors into the body.

[0049] Next, in the present invention, blood calcium is not directly collected; instead, the first non-invasive biosignal data from a smart ring-type wearable sensor unit, the second non-invasive biosignal data from an externally worn wearable patch unit, and the 3-axis acceleration data from a wrist-mounted smartwatch are integrated and analyzed by a field edge-type AI module to obtain blood ionized calcium (Ca 2+ Estimates the state.

[0050] After that, the estimated blood calcium status is transmitted to the calcium balance smart application module, and if the estimated calcium status is determined to be at risk of low calcium or a trend of decreasing calcium, the field edge AI module outputs a PTH automatic control signal to the photo-stimulation-responsive body-implantable PTH secreting object to regulate the amount of PTH injected.

[0051] In other words, blood ionized calcium (Ca 2+ The present invention relates to a closed-loop control device and method for secreting implantable PTH by observing the condition and outputting a photostimulation response signal that causes the secretion of implantable PTH.

[0053] Next, the first difference between the present invention and existing technology is that it is non-invasive.

[0054] That is, without blood sampling, without surgery, and without inserting sensors into the body, the first non-invasive biosignal data, the second non-invasive biosignal data, and the 3-axis acceleration data are sensed using only a smart ring-type wearable sensor part, an externally worn wearable patch part, and a wrist-attached smartwatch.

[0056] Next, the second difference between the present invention and existing technology is multimodal biosignal fusion.

[0057] That is, PPG (Photoplethysmography), blood oxygen saturation (SpO2), bioimpedance, skin temperature, tremor, electromyography (EMG), ECG (Electrocardiography) signals, heart rate and heart rate variability, and 3-axis acceleration are analyzed together.

[0059] Next, the third difference between the present invention and the existing technology is that, without cloud delay, a field edge-type AI module located at a close distance of 1m to 5000m relative to a wrist-worn smartwatch estimates and determines the calcium status in real time, outputs an automatic PTH control signal toward a photo-stimulated response-type body-implantable PTH-secreting object, and controls the PTH injection amount through closed-loop control.

[0061] Hereinafter, preferred embodiments according to the present invention will be described with reference to the accompanying drawings.

[0062] FIG. 1 is a block diagram illustrating the components of a body-implantable PTH secretion closed-loop control device (1) with a blood ionized calcium estimation formula based on a non-invasive multi-biosignal-based field edge AI according to the present invention, and FIG. 2 is a configuration diagram illustrating the components of a body-implantable PTH secretion closed-loop control device (1) with a blood ionized calcium estimation formula based on a non-invasive multi-biosignal-based field edge AI according to the present invention. This device is configured to attach to or wear a user's body, use sensed biosignal data to estimate the blood calcium state, and output a photo-stimulation response signal to secrete body-implantable PTH toward a photo-stimulation response type body-implantable PTH secretion object using the estimated blood calcium state to perform closed-loop control.

[0064] More specifically, the body-implantable PTH secretion closed-loop control device (1) of the above-mentioned non-invasive multi-biosignal-based field edge AI blood ionized calcium estimation formula is

[0065] It consists of a smart ring-type wearable sensor part (100), an external wearable patch part (200), a wrist-attached smartwatch (300), an on-site edge-type AI module (400), a calcium balance smart app module (500), and a photo-stimulating response-type body-implantable PTH secreting object (600).

[0067] First, a smart ring-type wearable sensor unit (100) according to the present invention will be described.

[0068] The smart ring-type wearable sensor unit (100) is formed in a ring structure and is worn on a finger to sense first non-invasive biosignal data regarding PPG (Photoplethysmography), blood oxygen saturation (SpO2: Saturation of Peripheral Oxygen), bioimpedance, skin temperature, tremor, and electromyography (EMG), and to transmit the sensed first non-invasive biosignal data to a wrist-attached smartwatch.

[0069] As shown in FIG. 3, this consists of a ring-shaped main body (110), a PPG photoplethysmography sensor (120), an SpO₂ oxygen saturation measurement unit (130), a bioimpedance measurement unit (140), a skin temperature sensor (150), a tremor detection sensor (160), a finger EMG electrode (170), a first signal processing unit (180), a ring-shaped wireless communication unit (190), and a ring-shaped battery unit (190a).

[0071] First, the ring-shaped main body (110) according to the present invention will be described.

[0072] The above ring-shaped main body (110) is formed in a ring shape for inserting a finger, and serves to protect and support each device from external pressure.

[0073] As illustrated in FIGS. 4 to 6, a PPG photoplethysmography sensor unit is formed on one side of the inner surface, an SpO₂ oxygen saturation measurement unit is formed on one side of the PPG photoplethysmography sensor unit, a bioimpedance measurement unit is formed on one side of the SpO₂ oxygen saturation measurement unit, a skin temperature sensor unit is formed on the inner skin contact surface, a tremor detection sensor unit is formed on a ring-shaped inner PCB, a finger EMG electrode unit is formed on the other side of the PPG photoplethysmography sensor unit, a first signal processing unit is formed on a ring-shaped inner PCB, a ring-shaped wireless communication unit is formed on one side of the first signal processing unit, and a ring-shaped battery unit is formed on one side of the ring-shaped inner PCB.

[0074] In addition, it is configured with an inner diameter of 16mm to 25mm, an outer diameter of 22mm to 32mm, a ring width of 10mm, a ring thickness of 3mm, a circumference of 55mm to 64mm, a sensor placement space of 30 mm², a PPG sensor of 4 mm × 3 mm, a SpO₂ sensor of 3 mm × 3 mm, an EMG electrode of 2 mm × 8 mm, and a Bio-Z electrode of 2 mm × 8 mm.

[0076] Second, the PPG photoplethysmography sensor unit (120) according to the present invention will be described.

[0077] The above PPG photoplethysmography sensor unit (120) is located on one side of the inner surface of the ring-shaped main body and irradiates LED light onto the skin and detects reflected light to sense changes in peripheral blood flow and pulse waves in the finger.

[0078] As shown in FIG. 6, this consists of a green LED, a red LED, an infrared LED, a photodiode, an optical lens, a light blocking wall, and an analog front-end circuit.

[0079] In other words, green LEDs, red LEDs, and infrared LEDs are irradiated.

[0080] Next, reflected light reflected from the finger tissue and blood vessels is detected by a photodiode.

[0081] Next, the change in light intensity is analyzed through the analog front-end circuit.

[0082] Next, generate a PPG waveform.

[0084] Thus, by configuring the PPG photoplethysmography sensor unit according to the present invention, it is possible to measure pulse waves, estimate heart rate, detect changes in peripheral blood flow, and generate a PPG waveform to provide an indirect blood flow indicator for estimating blood calcium.

[0086] Third, the SpO₂ oxygen saturation measuring unit (130) according to the present invention will be described.

[0087] The above SpO₂ oxygen saturation measuring unit (130) is located on one side of the PPG photoplethysmography sensor unit and serves to calculate blood oxygen saturation using the difference in absorbance between red light and infrared light.

[0088] This consists of a red LED, an infrared LED, a photodetector, an absorbance comparison unit, an SpO₂ calculation unit, and a signal correction unit.

[0090] That is, red LEDs and infrared LEDs are irradiated alternately.

[0091] Next, the absorbance of oxygenated and deoxygenated hemoglobin in the blood is measured.

[0092] Next, the absorbance ratio is calculated.

[0093] Next, SpO₂ oxygen saturation data is generated.

[0095] Thus, by configuring the SpO₂ oxygen saturation measuring unit according to the present invention, oxygen saturation can be measured, peripheral circulation status can be determined, PPG reliability can be corrected, and calcium estimation errors can be reduced in hypoxic and blood flow reduction situations.

[0097] Fourth, the bioimpedance measuring unit (140) according to the present invention will be described.

[0098] The above bioimpedance measuring unit (140) is located on one side of the SpO₂ oxygen saturation measuring unit and applies a micro alternating current to the finger tissue and measures the voltage change to calculate indicators related to tissue impedance, body fluid change, and electrolyte change.

[0099] This consists of a first current applying electrode, a second current applying electrode, a first voltage measuring electrode, a second voltage measuring electrode, an alternating current generating unit, a voltage detection unit, and an impedance calculation unit.

[0101] Fifth, the skin temperature sensor part (150) according to the present invention will be described.

[0102] The above skin temperature sensor part (150) is located on the inner skin contact surface of the ring-shaped main body part and serves to sense the surface temperature of the finger skin.

[0104] Sixth, the Tremor detection sensor unit (160) according to the present invention will be described.

[0105] The above-mentioned tremor detection sensor unit (160) is located on a ring-shaped internal PCB and serves to sense minute tremors, vibrations, and abnormal movements of the finger.

[0106] This senses fine tremors, vibrations, and abnormal movements of the fingers to generate a neuromuscular excitability indicator that determines whether there is an increase in hypocalcic neuromuscular excitability.

[0107] And, MEMS 3-axis accelerometer, MEMS gyroscope, vibration frequency analysis unit,

[0108] It consists of a Tremor index calculation unit.

[0110] In other words, when micro-vibrations occur in the finger, acceleration and angular velocity are measured.

[0111] Next, the amplitude and frequency are analyzed.

[0112] Next, create a Tremor index.

[0114] Through this configuration, finger tremors are detected, hypocalcific muscle spasm precursors are detected, and neuromuscular excitability indicators are generated, which can be utilized as auxiliary data for distinguishing between active tremors and pathological tremors.

[0116] Seventh, the finger EMG electrode part (170) according to the present invention will be described.

[0117] The finger EMG electrode part (170) is located on the other side of the PPG photoplethysmography sensor part and serves to measure fine electromyogram signals according to finger muscle or neuromuscular excitation.

[0118] This consists of a first surface EMG electrode, a second surface EMG electrode, a reference electrode, a differential amplifier, high-pass and low-pass filters, and an EMG signal detection unit.

[0119] In other words, when micro-contraction of the finger muscles occurs, a change in skin surface potential occurs.

[0120] Next, the EMG electrode detects the potential difference.

[0121] Next, amplification and filtering are performed.

[0122] Next, generate an EMG index.

[0124] In this way, by configuring the finger EMG electrode, muscle spasms can be detected, increased neuromuscular excitability can be detected, premonitory signs of hypocalcific tetany can be detected, and it can be utilized as auxiliary data for calculating the Neural Excitability Index.

[0126] Eighth, the first signal processing unit (180) according to the present invention will be described.

[0127] The first signal processing unit (180) is located on a ring-shaped internal PCB and functions to amplify, filter, digitally convert, and extract features from raw signals collected from a PPG photoplethysmography sensor unit, an SpO₂ oxygen saturation measurement unit, a bioimpedance measurement unit, a skin temperature sensor unit, a tremor detection sensor unit, and a finger EMG electrode unit, and convert them into first non-invasive biosignal data.

[0128] This consists of an MCU, DSP, ADC, sensor interface circuit, noise removal filter, feature extraction unit, data packet generation unit, and storage memory unit.

[0130] Ninth, a ring-shaped wireless communication unit (190) according to the present invention will be described.

[0131] The above ring-shaped wireless communication unit (190) is located on one side of the first signal processing unit and serves to wirelessly transmit the first non-invasive biosignal data generated by the first signal processing unit to a wrist-worn smartwatch.

[0132] This consists of a BLE communication chip, an RF transceiver, an antenna pattern, and a pairing control unit.

[0134] Tenth, a ring-shaped battery part (190a) according to the present invention will be described.

[0135] The above-mentioned ring-shaped battery section (190a) is formed in an arc-shaped structure on one side of the ring-shaped internal PCB and serves to supply power to each device.

[0136] Here, each device is composed of a PPG photoplethysmography sensor unit, an SpO₂ oxygen saturation measurement unit, a bioimpedance measurement unit, a skin temperature sensor unit, a tremor detection sensor unit, a finger EMG electrode unit, a first signal processing unit, and a ring-type wireless communication unit.

[0137] This consists of a small lithium polymer battery, an overcharge protection circuit, an over-discharge protection circuit, and a battery level detection unit.

[0139] That is, it is worn on the user's finger to sense first non-invasive biosignal data regarding PPG (Photoplethysmography), blood oxygen saturation (SpO2), bioimpedance, skin temperature, tremor, and electromyography (EMG).

[0140] Next, remove the noise.

[0141] Next, the first non-invasive biosignal data is generated.

[0142] Next, the generated first non-invasive biosignal data is transmitted to the wrist-worn smartwatch.

[0144] Thus, by configuring a smart ring-type wearable sensor unit consisting of a ring-type main body, a PPG photoplethysmography sensor unit, an SpO₂ oxygen saturation measurement unit, a bioimpedance measurement unit, a skin temperature sensor unit, a tremor detection sensor unit, a finger EMG electrode unit, a first signal processing unit, a ring-type wireless communication unit, and a ring-type battery unit, the PPG photoplethysmography sensor unit and the SpO₂ oxygen saturation measurement unit enable simultaneous correction of peripheral blood flow status and oxygen saturation compared to simple pulse wave measurement, thereby improving pulse wave signal reliability by 40% compared to existing methods. Furthermore, by acquiring auxiliary indicators related to tissue water content, body fluid changes, and electrolyte changes through the bioimpedance measurement unit, the amount of indirect information required for estimating blood calcium status can be increased by 1.5 to 2.5 times compared to existing smart rings using only PPG and SpO₂. Additionally, it is possible to simultaneously detect increased neuromuscular excitability, fine tremors, and muscle spasm precursors that occur in hypocalcium conditions, thereby providing neuromuscular excitability indicators compared to tremor or EMG measurements alone. The calculation accuracy can be improved by 1.5 to 3 times, and by linking the ring-type wireless communication unit and the first signal processing unit to transmit feature value-centered data packets, the amount of wireless communication data can be reduced to 70% or less compared to the raw data transmission method.

[0146] Next, an external wearable patch part (200) according to the present invention will be described.

[0147] The above-mentioned external wearable patch (200) is attached to the surface of the chest skin to sense second non-invasive biosignal data regarding ECG (Electrocardiography) signals, heart rate, and heart rate variability flowing from the surface of the chest skin, and serves to transmit the sensed second non-invasive biosignal data to a wrist-attached smartwatch.

[0148] As illustrated in FIGS. 7 and 8, this consists of a patch-type adhesive substrate (210), an ECG electrode (220), an ECG amplifier (230), an ECG filter (240), a heart rate calculation unit (250), an HRV calculation unit (260), a raw ECG storage unit (270) for QT·QTc extraction, a wireless communication unit for the patch (280), and a battery unit for the patch (290).

[0150] First, a patch-type adhesive substrate part (210) according to the present invention will be described.

[0151] The above patch-type adhesive substrate (210) is positioned on the lower surface of the entire patch that is in direct contact with the surface of the chest skin, and serves to adhere the ECG electrode and electronic circuit to the skin and support the patch so that it does not deviate from its position.

[0152] This consists of a flexible patch substrate, a skin-contact hydrogel layer, a breathable adhesive film, a waterproof protective film, an electrode fixing groove, a circuit housing, and a skin irritation relief layer.

[0153] And, the substrate layer is composed of 60-90 wt% TPU and 10-40 wt% silicone elastomer, and the skin contact layer is composed of 40-70 wt% medical hydrogel and 20-50 wt% silicone adhesive, and

[0154] It consists of 5 to 15 weight percent of a moisturizer, and the outer protective film is made of a TPU film.

[0155] In addition, the width is formed to be 40 to 80 cm, the length to be 30 to 80 cm, the thickness to be 0.5 to 4.0 mm, the elasticity to be 150 to 400%, the bending radius for close contact with the chest curve to be 20 to 150 mm, and the weight to be 5 to 15 g.

[0157] Second, the ECG electrode part (220) according to the present invention will be described.

[0158] As shown in FIGS. 9 and 10, the ECG electrode portion (220) is positioned on the skin contact surface of the patch-type adhesive substrate portion and serves to detect microelectric signals generated from the heart and transmitted to the surface of the chest skin.

[0159] This consists of a first ECG electrode, a second ECG electrode, a reference electrode, an Ag·AgCl electrode layer, an electrode gel layer, an electrode lead pattern, and a skin contact stabilization part.

[0160] In other words, when cardiac electrical activity occurs, a potential difference is formed on the surface of the chest skin, and the ECG electrode detects this potential difference.

[0161] At this time, a raw ECG signal is generated.

[0163] Third, the ECG amplification unit (230) according to the present invention will be described.

[0164] As shown in FIG. 9, the ECG amplifier (230) is located on one side of the ECG electrode and serves to amplify the minute electrical signal measured by the ECG electrode to an analyzable size.

[0165] This consists of an instrumentation amplifier, a differential amplifier circuit, a common-mode rejection circuit, an input protection circuit, and a reference voltage generation unit.

[0167] Fourth, the ECG filter unit (240) according to the present invention will be described.

[0168] As shown in FIG. 9, the above ECG filter unit (240) is located at one side of the rear end of the ECG amplifier unit and serves to filter power noise, electromyography noise, motion noise, and baseline fluctuations from the ECG signal amplified by the ECG amplifier unit.

[0169] This consists of a high-pass filter, a low-pass filter, a notch filter, a baseline correction unit, a motion artifact removal unit, and a digital filter unit.

[0170] In other words, when an amplified ECG signal is input, baseline fluctuations are eliminated.

[0171] Next, power noise is removed.

[0172] Next, high-frequency noise is removed.

[0173] Next, a filtered ECG signal is generated.

[0175] Fifth, the heart rate calculation unit (250) according to the present invention will be described.

[0176] As shown in FIG. 9, the heart rate calculation unit (250) is located at one side of the rear end of the ECG filter unit and plays the role of detecting the R wave of the ECG signal and calculating the heart rate.

[0177] This consists of an R-wave detection unit, an RR interval calculation unit, a heart rate calculation unit, an abnormal beat removal unit, and an average heart rate calculation unit.

[0178] In other words, it receives a filtered ECG signal and detects the position of the R wave.

[0179] Next, calculate the RR interval.

[0180] Next, calculate the heart rate per minute.

[0181] Next, generate heart rate data.

[0183] Sixth, the HRV calculation unit (260) according to the present invention will be described.

[0184] As shown in FIG. 9, the above HRV calculation unit (260) is located on one side of the heart rate calculation unit and analyzes changes in the RR interval to calculate heart rate variability data.

[0185] Here, the RR interval is one of the most important time indicators in an electrocardiogram (ECG), referring to the time interval between two consecutive R-waves.

[0186] This consists of an RR interval storage unit, a time domain analysis unit, a frequency domain analysis unit, an autonomic nervous system index calculation unit, and an abnormal RR removal unit.

[0187] In other words, collect RR interval data.

[0188] Next, remove the abnormal RR.

[0189] Next, the time-domain HRV is calculated.

[0190] Next, the frequency domain HRV is calculated.

[0191] Next, generate HRV data.

[0193] Seventh, a raw ECG storage unit (270) for extracting QT·QTc according to the present invention will be described.

[0194] As shown in FIG. 9, the raw ECG storage unit (270) for extracting QT and QTc is located on one side of the memory unit inside the tooth-shaped adhesive substrate and serves to store raw ECG or high-quality filtered ECG data of a certain time interval for calculating the QT interval and QTc.

[0196] Eighth, a wireless communication unit (280) for a patch according to the present invention will be described.

[0197] As shown in FIG. 9, the above patch wireless communication unit (280) is positioned on the flexible PCB of the patch-type adhesive substrate and serves to wirelessly transmit second non-invasive biosignal data regarding ECG (Electrocardiography) signals, heart rate, and heart rate variability to a wrist-worn smartwatch.

[0198] This consists of a BLE communication chip, an RF transceiver, an antenna pattern, and a pairing control unit.

[0200] Ninth, a battery part (290) for a patch according to the present invention will be described.

[0201] As shown in FIG. 9, the above-mentioned patch battery portion (290) is located on one side of the upper part of the patch-type adhesive substrate portion and is formed with a curved structure to supply power to each device.

[0202] Here, each device refers to an ECG electrode unit, an ECG amplifier unit, an ECG filter unit, a heart rate calculation unit, an HRV calculation unit, a raw ECG storage unit for QT·QTc extraction, and a wireless communication unit for a patch.

[0203] This consists of a thin-film lithium polymer battery, a power management circuit, an overcharge protection circuit, an over-discharge protection circuit, and a battery remaining capacity detection unit.

[0205] That is, a patch-type adhesive substrate is attached to the skin of the chest.

[0206] Next, the ECG signal is measured.

[0207] Next, heart rate, RR interval, and HRV data are generated.

[0208] Next, second non-invasive biosignal data is generated.

[0209] Next, the second non-invasive biosignal data sensed by the smartwatch is transmitted.

[0211] Here, an external wearable patch is configured comprising a patch-type adhesive substrate, an ECG electrode, an ECG amplifier, an ECG filter, a heart rate calculation unit, an HRV calculation unit, a raw ECG storage unit for QT and QTc extraction, a wireless communication unit for the patch, and a battery unit for the patch. By linking the ECG electrode, ECG amplifier, and ECG filter, the Signal to Noise Ratio (SNR) of the electrocardiogram signal can be improved by 80% compared to the existing method, the ability to analyze the autonomic nervous state can be improved by 2 to 4 times compared to the existing method by performing HRV analysis based on the RR interval, the ability to detect QT prolongation occurring in hypocalcemia can be improved by 3 to 5 times compared to the existing method through QT interval and QTc analysis functions, and the early prediction performance of hypocalcemia risk can be improved by 80% compared to the existing method by integrating heart rate, RR interval, HRV, QT interval, and QTc analysis.

[0213] Next, a wrist-mounted smartwatch (300) according to the present invention will be described.

[0214] The above wrist-mounted smartwatch (300) is worn on the user's wrist to measure 3-axis acceleration data regarding the user's moving posture, collects the first non-invasive biosignal data transmitted from the smart ring-type wearable sensor unit and the second non-invasive biosignal data transmitted from the external wearable patch unit, transmits them to the field edge-type AI module via a wireless communication network, receives a photo-stimulation response signal from the field edge-type AI module, and performs the role of photo-stimulating the photo-stimulation response type body-implantable PTH secreting object implanted inside the PTH minor neck muscle.

[0215] As illustrated in FIG. 11, this consists of a watch-type main body (310), a 3-axis accelerometer (320), a gyroscope (330), a watch-type data receiver (340), a time synchronization unit (350), a data packet generation unit (360), a watch-type wireless communication unit (370), a watch-type memory unit (380), a user notification unit (390), a watch-type battery unit (390a), and a photo-stimulation unit (390b).

[0217] First, the watch-type main body (310) according to the present invention will be described.

[0218] The above watch-type main body (310) is formed in the shape of a wristwatch worn on the user's wrist and serves to protect and support each device from external pressure.

[0219] As illustrated in FIGS. 12 to 15, a 3-axis accelerometer is formed on an internal PCB, a gyroscope is formed on one side of the 3-axis accelerometer, a watch-type data receiver is formed on one side of the gyroscope, a time synchronization unit is formed on one side of the watch-type data receiver, a data packet generation unit is formed on one side of the time synchronization unit, a watch-type wireless communication unit is formed on one side of the data packet generation unit, a watch-type memory unit is formed on one side of the watch-type wireless communication unit, a user notification unit is formed on one side of the front display window, a watch-type battery unit is formed on one side of the lower part of the internal space, and a photo-stimulation unit is formed on one side of the bottom surface that contacts the wrist muscle.

[0220] In addition, the watch housing, wristband portion, front display window, internal PCB storage portion, waterproof and dustproof sealing portion, sensor fixing frame, shock absorbing portion, and skin contact surface are formed and configured.

[0222] Second, the 3-axis acceleration sensor unit (320) according to the present invention will be described.

[0223] The above 3-axis acceleration sensor unit (320) is located on the internal PCB of the watch-type main body and performs the role of measuring linear acceleration changes in the X-axis, Y-axis, and Z-axis directions according to the movement of the user's wrist.

[0224] This consists of a MEMS 3-axis acceleration sensor, an X-axis acceleration detector, a Y-axis acceleration detector, a Z-axis acceleration detector, an acceleration correction circuit, and a sampling control unit.

[0226] In other words, when wrist movement occurs, it detects changes in acceleration along the X, Y, and Z axes.

[0227] Next, 3-axis acceleration data is generated.

[0228] Next, it is passed to the time synchronization unit.

[0230] Third, the gyro sensor unit (330) according to the present invention will be described.

[0231] The above-mentioned gyroscope sensor unit (330) is located on one side of the 3-axis accelerometer sensor unit and performs the role of measuring angular velocity data according to the user's wrist rotation, twisting, and change of direction.

[0232] This includes a MEMS 3-axis gyro sensor, roll detector, pitch detector, yaw detector, angular velocity correction unit,

[0233] It consists of a pose estimation operation unit.

[0234] In other words, when wrist rotation occurs, roll, pitch, and yaw angular velocities are detected.

[0235] Next, generate gyroscope data.

[0236] Next, it is fused with 3-axis acceleration data.

[0237] Next, posture and activity status information is generated.

[0239] Fourth, a watch-type data receiver (340) according to the present invention will be described.

[0240] The watch-type data receiver (340) is located on one side of the gyro sensor unit and serves to receive the first non-invasive biosignal data transmitted from the smart ring-type wearable sensor unit and the second non-invasive biosignal data transmitted from the external wearable patch unit.

[0241] This consists of a first data receiving channel, a second data receiving channel, and a BLE receiving interface.

[0243] Fifth, the time synchronization unit (350) according to the present invention will be described.

[0244] The time synchronization unit (350) is located on one side of the watch-type data receiver and serves to synchronize the first non-invasive biosignal data, the second non-invasive biosignal data, and the 3-axis acceleration data of different time axes generated by the smart ring-type wearable sensor unit, the externally worn wearable patch unit, and the wrist-attached smartwatch itself to the same reference time.

[0245] This consists of a real-time clock unit, a timestamp assignment unit, a sampling period correction unit, a sensor-specific time correction unit, a synchronization queue, and a delay time correction unit.

[0246] That is, the first non-invasive biosignal data is received.

[0247] Next, the second non-invasive biosignal data is received.

[0248] Next, 3-axis acceleration data is collected.

[0249] Next, a timestamp is assigned to each data.

[0250] Next, align the time axis.

[0251] Next, generate synchronization data.

[0253] Sixth, a data packet generation unit (360) according to the present invention will be described.

[0254] The above data packet generation unit (360) is located on one side of the time synchronization unit and plays the role of generating synchronized first non-invasive biosignal data, second non-invasive biosignal data and 3-axis acceleration data into data packets that can be analyzed by the field edge AI module.

[0255] This consists of a data format conversion unit, a packet header generation unit, a sensor ID assignment unit, a timestamp insertion unit, and a compression unit.

[0257] Seventh, a watch-type wireless communication unit (370) according to the present invention will be described.

[0258] The above watch-type wireless communication unit (370) is located on one side of the data packet generation unit and plays the role of transmitting integrated data packets to the field edge-type AI module through a wireless communication network.

[0259] This consists of a BLE communication unit, a Wi-Fi communication unit, an LTE-M / NB-IoT communication unit, an RF transceiver, and an antenna unit.

[0261] Eighth, a watch-type memory unit (380) according to the present invention will be described.

[0262] The above watch-type memory unit (380) is located on one side of the watch-type wireless communication unit and serves to store first non-invasive biosignal data, second non-invasive biosignal data, 3-axis acceleration data, communication status, notification history, and transmission failure data.

[0264] Ninth, the user notification unit (390) according to the present invention will be described.

[0265] The above user notification unit (390) is located on one side of the front display window of the watch-type main body and performs the role of notifying the user of the sensor connection status, data transmission status, low battery, improper wearing, and low calcium risk.

[0266] This consists of a buzzer or a speaker.

[0268] Tenth, the watch-type battery part (390a) according to the present invention will be described.

[0269] As shown in FIG. 14, the watch-type battery unit (390a) is located on one side of the lower part of the internal space and serves to supply power to each device of the wrist-mounted smartwatch.

[0270] Here, each device refers to a 3-axis accelerometer, a gyroscope, a watch-type data receiver, a time synchronization unit, a data packet generation unit, a watch-type wireless communication unit, a watch-type memory unit, and a user notification unit.

[0271] This consists of a lithium polymer battery, a power management circuit, a charging circuit, a wireless charging coil, an overcharge protection circuit, an over-discharge protection circuit, and a battery level detection unit.

[0273] Eleventh, the photo-stimulating part (390b) according to the present invention will be described.

[0274] The above-mentioned photo-stimulation unit (390b) is located on the skin contact surface, which is one side of the bottom surface of the watch-type main body that contacts the wrist muscle, and serves to photo-stimulate a photo-stimulation-responsive body-implantable PTH-secreting object implanted inside the wrist muscle according to a photo-stimulation response signal received from the field edge-type AI module.

[0275] This consists of a micro LED array, a near-infrared LED, a light output control unit, a pulse width control unit, a light irradiation lens unit, a skin contact light window unit, a temperature sensing unit, and an overheat cutoff unit.

[0277] That is, the field edge AI module determines low blood calcium levels and generates a photostimulation response signal.

[0278] Next, the watch-type wireless communication unit receives it.

[0279] Next, the photostimulation unit irradiates light at a specified wavelength, intensity, and time.

[0280] Next, the photo-stimulated body-implantable PTH-secreting object responds.

[0281] Next, the photo-stimulated body-implantable PTH-secreting object induces body-implantable PTH secretion.

[0282] Next, the first non-invasive biosignal data, the second non-invasive biosignal data, and the 3-axis acceleration data are remeasured through the smart ring-type wearable sensor unit + the external wearable patch unit + the wrist-attached smartwatch.

[0284] Thus, by configuring a wrist-mounted smartwatch according to the present invention comprising a watch-type main body, a 3-axis accelerometer sensor, a gyroscope sensor, a watch-type data receiver, a time synchronization unit, a data packet generation unit, a watch-type wireless communication unit, a watch-type memory unit, a user notification unit, a watch-type battery unit, and a photo-stimulation unit, the time-axis data of the smart ring-type wearable sensor, the externally worn wearable patch, and the wrist-mounted smartwatch's own sensor are aligned to the same reference time through the time synchronization unit, thereby reducing the time error between sensors to 70% or less compared to the existing method; and by generating an integrated data packet containing a sensor ID, timestamp, error detection code, and compression information from the synchronized data through the data packet generation unit, the data interpretation speed of the field edge-type AI module can be improved by 1.5 to 3 times compared to the existing method, and the photo-stimulation unit [determines] a photo-stimulation-responsive body-implantable PTH-secreting object inside the wrist muscle according to the photo-stimulation response signal received from the field edge-type AI module By stimulating, the response rate of PTH secretion induction can be improved by 80% compared to the existing drug infusion-centered method.

[0286] Next, a field edge AI module (400) according to the present invention will be described.

[0287] The above-mentioned field edge AI module (400) is located in a hospital or public institution within a distance of 1m to 5000m relative to the wrist-worn smartwatch and is connected to the wrist-worn smartwatch via a wireless communication network. It calculates the QT (Q wave to T wave) interval, HRV (Heart Rate Variability), Pulse waveform, and neural excitability index using the first non-invasive biosignal data, second non-invasive biosignal data, and 3-axis acceleration data transmitted from the wrist-worn smartwatch, estimates the blood calcium state, transmits the estimated blood calcium state to the calcium balance smart app module, and uses the estimated blood calcium state to output a photo-stimulation response signal to secrete body-implantable PTH towards a photo-stimulation response type body-implantable PTH secreting object located in the wrist muscle, thereby performing a closed-loop control function.

[0288] Here, it is connected to a wrist-worn smartwatch via a wireless communication network or in the form of a cloud.

[0289] As illustrated in FIGS. 16 to 19, this consists of a data receiving unit (410), a preprocessing unit (420), a motion artifact removal unit (430), a feature value calculation control unit (440), a QT interval calculation control unit (450), an HRV calculation control unit (460), a pulse waveform analysis control unit (470), a neural excitability analysis control unit (480), an index calculation unit (490), a blood calcium status estimation control unit (490a), a low calcium risk prediction control unit (490b), a photostimulation response signal generation control unit (490c), an edge wireless communication unit (490d), and an edge AI processor unit (490e).

[0290] In addition, the on-site edge AI modules are installed in locations such as hospitals and government offices where a rapid response can be achieved in emergency situations.

[0292] First, the data receiving unit (410) according to the present invention will be described.

[0293] The above data receiving unit (410) serves to receive integrated data packets from a wrist-worn smartwatch.

[0294] This consists of a data reception interface, a packet interpretation unit, a sensor ID verification unit, a timestamp verification unit, a data integrity check unit, and an error detection unit.

[0296] Second, the pretreatment unit (420) according to the present invention will be described.

[0297] The above preprocessing unit (420) serves to correct noise, abnormal values, and missing values ​​of the biosignal received through the data receiving unit.

[0298] This consists of a noise removal unit, a missing value interpolation unit, an outlier removal unit, a signal normalization unit, a sampling correction unit, and a data scaling unit.

[0300] Third, the motion artifact removal unit (430) according to the present invention will be described.

[0301] The above motion artifact removal unit (430) uses 3-axis acceleration data received through the data receiving unit to remove PPG, ECG, and EMG signal distortions caused by movement.

[0302] This consists of an acceleration-based motion judgment unit, a gyroscope-based rotation correction unit, an adaptive filter unit, a motion segment masking unit, and a signal quality index calculation unit.

[0304] Fourth, the feature value operation control unit (440) according to the present invention will be described.

[0305] The above feature value calculation control unit (440) plays the role of controlling the calculation of integrated feature values ​​necessary for estimating blood calcium status from the first non-invasive biosignal data, the second non-invasive biosignal data, and the 3-axis acceleration data.

[0306] As shown in FIG. 20, this consists of an ECG feature extraction unit (441), a PPG feature extraction unit (442), an EMG feature extraction unit (443), a Tremor feature extraction unit (444), a Bio-Z feature extraction unit (445), a skin temperature feature extraction unit (446), and a feature value integration unit (447).

[0308] The above ECG feature extraction unit (441) serves to extract cardiac electrical conduction-related feature values ​​from the ECG signal transmitted from the external wearable patch.

[0309] This consists of an R-wave detection unit, an RR interval calculation unit, a QRS detection unit, a QT interval extraction unit, a QTc correction unit, an ST-T waveform analysis unit, and an arrhythmia suspected pattern detection unit.

[0311] The above PPG feature extraction unit (442) plays the role of extracting peripheral blood flow and pulse wave-related feature values ​​from the PPG signal transmitted from the smart ring-type wearable sensor unit.

[0312] This consists of a PPG peak detection unit, a pulse wave amplitude calculation unit, a pulse wave width calculation unit, a rise time calculation unit, a fall time calculation unit, a pulse wave variability analysis unit, and a peripheral blood flow index calculation unit.

[0314] The above EMG feature extraction unit (443) serves to extract neuromuscular excitability-related feature values ​​from the electromyogram signal measured at the finger EMG electrode unit.

[0315] This consists of an EMG amplification data receiver, an electromyography burst detector, an RMS calculation unit, an average frequency calculation unit, a muscle contraction duration calculation unit, and a microspasm detection unit.

[0317] The above Tremor feature extraction unit (444) is responsible for extracting fine tremor feature values ​​of the finger or wrist from the Tremor detection sensor unit and 3-axis acceleration data.

[0318] This consists of an acceleration amplitude analysis unit, a frequency spectrum analysis unit, a tremor duration calculation unit, an abnormal vibration detection unit, an active tremor removal unit, and a tremor index calculation unit.

[0320] The above Bio-Z feature extraction unit (445) serves to extract feature values ​​related to tissue conductivity, body fluid change, and electrolyte change from the Bio-Z signal measured by the bioimpedance measurement unit.

[0321] This consists of an impedance raw value receiving unit, a resistance component calculation unit, a reactance component calculation unit, a phase angle calculation unit, a frequency-specific impedance analysis unit, and a body fluid change index calculation unit.

[0323] The above skin temperature feature extraction unit (446) serves to extract peripheral circulation and external environment correction feature values ​​from finger skin temperature data measured by the skin temperature sensor unit.

[0324] This consists of a skin temperature data receiving unit, a temperature change rate calculation unit, a reference temperature comparison unit, a hypothermia / hyperthermia correction unit, and a peripheral blood flow temperature index calculation unit.

[0326] The above feature value integration unit (447) serves to form a single integrated feature vector of feature values ​​calculated from each feature extraction unit.

[0327] This consists of a feature value normalization unit, a feature value weighting unit, a sensor reliability reflection unit, a missing feature value correction unit, a feature vector generation unit, and an AI input format conversion unit.

[0329] In other words, the corrected biosignal is input.

[0330] Next, feature values ​​for each sensor are extracted.

[0331] Next, the feature values ​​are integrated.

[0332] Next, it is transmitted to the QT interval operation control unit, HRV operation control unit, pulse waveform analysis control unit, and neural excitability analysis control unit.

[0334] Thus, by configuring a feature value calculation control unit consisting of an ECG feature extraction unit, a PPG feature extraction unit, an EMG feature extraction unit, a Tremor feature extraction unit, a Bio-Z feature extraction unit, a skin temperature feature extraction unit, and a feature value integration unit, raw signals can be converted into meaningful bio-feature values, and the accuracy of blood calcium estimation can be increased by 1.5 to 3 times compared to the existing method. Additionally, multiple bio-feature values ​​required for estimating the blood ionized calcium state can be integrated and calculated from the first non-invasive bio-signal data, the second non-invasive bio-signal data, and 3-axis acceleration data. Furthermore, by converting QT interval, QTc, HRV, pulse waveform, electromyographic excitability, microtremor, bioimpedance change, and skin temperature correction value into a single integrated feature vector, the AI ​​inference accuracy of the blood calcium state estimation control unit and the low calcium risk prediction performance can be improved by 80% compared to the existing method.

[0336] Fifth, the QT interval operation control unit (450) according to the present invention will be described.

[0337] The above QT interval operation control unit (450) plays the role of operating control to calculate the QT interval from the Q wave start point to the T wave end point and QTc in the ECG signal.

[0338] As illustrated in FIG. 21, this consists of a QRS detection unit (451), a Q-wave start point detection unit (452), a T-wave end point detection unit (453), a QT interval calculation unit (454), a QTc correction unit (455), and a QT extension determination unit (456).

[0340] The above QRS detection unit (451) detects the QRS complex in the ECG signal and sets the R-wave position as a reference point.

[0341] This consists of an ECG input unit, an R-wave peak detection unit, a QRS energy detection unit, a slope detection unit, a noise removal unit, and a QRS candidate determination unit.

[0343] The above Q wave start point detection unit (452) detects the point where the Q wave starts within the QRS complex and determines the start point of the QT interval.

[0344] This consists of a QRS start point search unit, an equipotential reference unit, a voice deviation detection unit, a slope change detection unit, and a Q-wave start point correction unit.

[0346] The above T-wave end point detection unit (453) detects the point where the T-wave ends in the ECG signal and determines the end point of the QT interval.

[0347] This consists of a T-wave candidate segment search unit, a T-wave peak detection unit, a slope tangent analysis unit, an equipotential line return determination unit, a T-wave endpoint correction unit, and a noise segment exclusion unit.

[0349] The above QT interval calculation unit (454) calculates the time interval between the Q wave start point and the T wave end point to calculate the QT interval.

[0350] This consists of a Q-wave start time input unit, a T-wave end time input unit, a time difference calculation unit, a beat-by-beat QT storage unit, an average QT calculation unit, and an abnormal QT removal unit.

[0352] The above QTc correction unit (455) corrects the QT interval that varies according to the heart rate and QT c It plays the role of producing.

[0353] This consists of an RR interval input section, a Bazett correction section, a Fridericia correction section, a Framingham correction section, a Hodges correction section, a correction formula selection section, and a QTc calculation section.

[0355] The above QT extension determination unit (456) has a calculated QT interval and QT c It plays a role in determining whether to extend QT and the risk level by comparing it with the reference value.

[0356] This consists of a gender and age reference value storage unit, an individual reference QTc storage unit, a QTc threshold comparison unit, a QT change rate analysis unit, a hypocalcific QT prolongation suspicion unit, a risk grade calculation unit, and an alert signal generation unit.

[0358] In other words, when an ECG signal is input, the QRS complex is detected.

[0359] Next, the starting point of the Q wave is detected.

[0360] Next, the T-wave endpoint is detected.

[0361] Next, the QT interval is calculated.

[0362] Next, the heart rate-corrected QTc is calculated.

[0364] Thus, by configuring a QT interval calculation control unit according to the present invention, comprising a QRS detection unit, a Q-wave start point detection unit, a T-wave end point detection unit, a QT interval calculation unit, a QTc correction unit, and a QT extension determination unit, the time interval from the Q-wave start point to the T-wave end point can be calculated, and QT interval extension and QTc increase that may occur in hypocalcemia can be detected non-invasively.

[0366] Sixth, the HRV operation control unit (460) according to the present invention will be described.

[0367] The above HRV calculation control unit (460) plays the role of controlling to calculate a heart rate variability index by analyzing changes in the RR interval.

[0368] As illustrated in FIG. 22, this consists of an R-wave detection unit (461), an RR interval calculation unit (462), a time domain HRV calculation unit (463), a frequency domain HRV calculation unit (464), a non-linear HRV calculation unit (465), and an autonomic nervous balance judgment unit (466).

[0370] The above R-wave detection unit (461) detects an R-wave from an ECG signal and generates a reference point for calculating the RR interval.

[0371] This consists of an ECG input unit, a digital filter unit, a peak detection unit, an adaptive threshold setting unit, a noise removal unit, and an R-wave position storage unit.

[0373] The above RR interval calculation unit (462) calculates the time interval (RR interval) between two consecutive R waves.

[0374] This consists of an R-wave time storage unit, a time difference calculation unit, an RR interval array storage unit, an abnormal RR removal unit, and an average RR calculation unit.

[0376] The above time domain HRV calculation unit (463) calculates the time domain HRV index using an RR interval array.

[0377] This consists of an RR array input unit, an average RR calculation unit, a standard deviation calculation unit, a variability calculation unit, and a time domain HRV storage unit.

[0379] The above frequency domain HRV calculation unit (464) converts the RR interval fluctuation into a frequency spectrum and analyzes the autonomic nervous system balance state.

[0380] This consists of an RR resampling unit, an FFT analysis unit, a power spectrum calculation unit, an LF calculation unit, an HF calculation unit, and an LF / HF calculation unit.

[0382] The above non-linear HRV calculation unit (465) serves to analyze the complexity and unpredictability of heart rate variability.

[0383] This consists of a Poincare Plot analysis unit, an Entropy analysis unit, a DFA analysis unit, and a non-linear HRV storage unit.

[0385] The above autonomic nervous system balance judgment unit (466) performs the role of determining the state of the autonomic nervous system by integrating the results of time domain HRV, frequency domain HRV, and non-linear HRV.

[0387] In other words, it detects ECG R waves.

[0388] Next, the RR interval is calculated.

[0389] Next, the time domain is analyzed using SDNN, RMSSD, etc.

[0390] Next, the frequency domains such as LF and HF are analyzed.

[0391] Next, generate the HRV indicator.

[0393] Thus, by configuring the R-wave detection unit, RR interval calculation unit, time domain HRV calculation unit, frequency domain HRV calculation unit, non-linear HRV calculation unit, and autonomic nervous balance judgment unit according to the present invention, it is possible to calculate the RR interval from the ECG signal and generate time domain, frequency domain, and non-linear heart rate variability indices, and to determine the user's sympathetic nervous system activity, parasympathetic nervous system activity, and autonomic nervous system balance state using the calculated HRV indices, and to quantify ECG-based autonomic nervous system information to increase the accuracy of estimating the blood ionized calcium state by 1.5 to 3 times compared to the existing method, and to improve the risk judgment performance of the low-calcium risk prediction control unit by 80% compared to the existing method.

[0395] Seventh, a pulse waveform analysis control unit (470) according to the present invention will be described.

[0396] The above pulse waveform analysis control unit (470) analyzes the PPG-based pulse waveform and calculates peripheral blood flow, vascular tone, and pulse wave changes.

[0397] As illustrated in FIG. 23, this consists of a PPG waveform detection unit (471), a peak detection unit (472), a pulse wave amplitude calculation unit (473), a pulse wave width calculation unit (474), a rise time calculation unit (475), a peripheral blood flow index calculation unit (476), and a pulse wave quality judgment unit (477).

[0399] The above PPG waveform detection unit (471) serves to detect a valid pulse waveform from the raw PPG signal collected from the PPG photoplethysmography sensor unit.

[0400] This consists of a PPG raw signal input unit, a light intensity correction unit, a baseline correction unit, a low-frequency noise removal unit, a high-frequency noise removal unit, and an effective waveform extraction unit.

[0402] The peak detection unit (472) above serves to detect the systolic peak, diastolic peak, or auxiliary peak of the pulse in the PPG waveform.

[0403] This consists of a systolic peak detection unit, a diastolic peak detection unit, a peak candidate search unit, an adaptive threshold setting unit, and an incorrect peak removal unit.

[0405] The above pulse amplitude calculation unit (473) calculates the peak height relative to the baseline in the PPG waveform and calculates the pulse amplitude.

[0406] This consists of a baseline setting unit, a peak amplitude calculation unit, a beat-by-beat amplitude storage unit, an average amplitude calculation unit, and an amplitude change rate calculation unit.

[0408] The above pulse wave width calculation unit (474) calculates the time width of the PPG pulse wave and analyzes changes in vascular elasticity and pulse wave shape.

[0409] This consists of a pulse wave start point detection unit, a pulse wave end point detection unit, a half-width calculation unit, a width storage unit per beat, and a width change rate calculation unit.

[0411] The above rise time calculation unit (475) serves to calculate the time from the starting point of the PPG pulse wave until it reaches the systole peak.

[0412] This consists of a pulse wave start point detection unit, a systolic peak time input unit, a time difference calculation unit, an upward slope calculation unit, and a rise time change rate calculation unit.

[0414] The above peripheral blood flow index calculation unit (476) calculates the peripheral blood flow index by combining pulse wave amplitude, pulse wave width, rise time, waveform stability, etc.

[0415] This consists of a pulse amplitude input unit, a pulse width input unit, a rise time input unit, a skin temperature correction unit, an SpO₂ correction unit, a blood flow index calculation unit, and a peripheral blood flow grading unit.

[0417] The above pulse wave quality judgment unit (477) performs the role of determining whether the collected PPG waveform is reliable enough to be used for estimating blood calcium status.

[0418] This consists of a signal-to-noise ratio calculation unit, a motion artifact detection unit, a skin contact status determination unit, a light intensity saturation determination unit, a waveform regularity determination unit, and a quality grade calculation unit.

[0420] In other words, input the PPG signal.

[0421] Next, the pulse wave peak is detected.

[0422] Next, the amplitude, width, and rise time are calculated.

[0423] Next, the peripheral blood flow index is generated.

[0424] Next, the pulse waveform feature value is output.

[0426] Thus, by configuring a pulse waveform analysis control unit according to the present invention, comprising a PPG waveform detection unit, a peak detection unit, a pulse amplitude calculation unit, a pulse width calculation unit, a rise time calculation unit, a peripheral blood flow index calculation unit, and a pulse wave quality judgment unit, the pulse amplitude, pulse width, rise time, and peripheral blood flow index can be calculated from a PPG-based pulse waveform obtained from a smart ring-type wearable sensor unit. Furthermore, by excluding or correcting low-quality PPG data caused by motion artifacts, poor skin contact, and light intensity saturation through the pulse wave quality judgment unit, the reliability of the pulse waveform feature value input to the blood ionized calcium state estimation control unit can be improved by 80% compared to the existing method, and by quantifying peripheral blood flow, vascular tone, and pulse wave changes, the accuracy of estimating the blood ionized calcium state can be increased by 1.5 to 3 times compared to the existing method in conjunction with ECG-based QT interval and HRV analysis.

[0428] Eighth, a neural excitability analysis control unit (480) according to the present invention will be described.

[0429] The above neural excitability analysis control unit (480) plays a role in analyzing changes in neuromuscular excitability using EMG, tremor, bioimpedance, and skin temperature data.

[0430] As illustrated in FIG. 24, this consists of an EMG burst analysis unit (481), a tremor frequency analysis unit (482), a micro-spasm detection unit (483), a Bio-Z change analysis unit (484), a skin temperature correction unit (485), and a neuromuscular response judgment unit (486).

[0432] The above EMG burst analysis unit (481) is responsible for detecting and analyzing abnormal muscle activation signals in the form of a burst from the electromyography (EMG) signal.

[0433] Here, Burst refers to a short muscle activation period (muscle firing period) in which the electromyogram (EMG) signal suddenly increases strongly for a certain period of time.

[0434] This consists of an EMG signal input unit, an EMG amplified signal storage unit, a burst detection unit, an RMS calculation unit, a burst duration calculation unit, a burst frequency calculation unit, and a burst energy calculation unit.

[0436] The above-mentioned Tremor frequency analysis unit (482) converts finger tremor data into the frequency domain and analyzes abnormal tremor components.

[0437] This consists of a Tremor signal input unit, an FFT analysis unit, a frequency spectrum analysis unit, a Tremor peak detection unit, a Tremor frequency calculation unit, and a Tremor amplitude calculation unit.

[0439] The above micro-spasm detection unit (483) serves to detect micro-spasms that are not visible to the naked eye using EMG Burst and Tremor data.

[0440] This consists of an EMG Burst input unit, a Tremor input unit, a convulsion pattern database, a microconvulsion detection unit, a convulsion duration calculation unit, and a convulsion intensity calculation unit.

[0442] The above Bio-Z change analysis unit (484) analyzes changes in bioimpedance to evaluate changes in tissue conductivity and extracellular fluid.

[0443] This consists of a Bio-Z input section, an impedance change calculation section, a resistance component analysis section, a reactance analysis section, a phase angle analysis section, and a tissue conductivity calculation section.

[0445] The above skin temperature correction unit (485) serves to correct the EMG, Tremor, and Bio-Z analysis results using changes in skin temperature.

[0446] This consists of a skin temperature input unit, a reference temperature storage unit, a temperature change rate calculation unit, a temperature correction coefficient calculation unit, and an environmental temperature correction unit.

[0448] The above neuromuscular response judgment unit (486) serves to determine the neuromuscular excitability state by integrating the results of EMG Burst, Tremor, microspasm, Bio-Z, and skin temperature correction.

[0449] This consists of an EMG analysis result input section, a Tremor analysis result input section, a microconvulsion input section, a Bio-Z input section, an integrated judgment algorithm engine section, a risk grade calculation section, and a Neural Excitability Index calculation section.

[0451] In other words, EMG, Tremor, Bio-Z, and skin temperature are input.

[0452] Next, an electromyogram burst is detected.

[0453] Next, the vibration frequency is analyzed.

[0454] Next, the change in tissue impedance is analyzed.

[0455] Next, neuromuscular excitability changes are assessed.

[0457] Thus, by configuring a neural excitability analysis control unit according to the present invention, comprising an EMG burst analysis unit, a tremor frequency analysis unit, a microconvulsion detection unit, a Bio-Z change analysis unit, a skin temperature correction unit, and a neuromuscular response judgment unit, a Neural Excitability Index can be calculated by integrating and analyzing EMG signals, tremor signals, bioimpedance signals, and skin temperature signals. Additionally, neuromuscular hyperexcitation, microconvulsions, muscle rigidity, and tetani precursors occurring in a hypocalcium state can be quantitatively reflected, and provided to a blood ionized calcium state estimation control unit, the accuracy of predicting hypocalcium risk can be increased by 1.5 to 3 times compared to the existing method. Furthermore, by additionally reflecting neuromuscular function changes that are difficult to identify solely through ECG-based QT interval analysis and HRV analysis, the reliability of estimating the blood ionized calcium state can be improved by 80%.

[0459] Ninth, an index calculation unit (490) according to the present invention will be described.

[0460] The index calculation unit (490) above performs the role of quantifying and indexing the results of QT, HRV, Pulse waveform, and Neural excitability.

[0461] This consists of a QT risk index calculation unit, an HRV instability index calculation unit, a pulse waveform index calculation unit, a neural excitability index calculation unit, and a calcium risk composite index calculation unit.

[0463] Tenth, the blood calcium status estimation control unit (490a) according to the present invention will be described.

[0464] The above blood calcium state estimation control unit (490a) uses data calculated from the index calculation unit to control the estimation of the blood ionized calcium state based on AI.

[0465] As illustrated in FIG. 25, this consists of an AI estimation model algorithm engine unit (490a-1), a personal reference value storage unit (490a-2), a calcium state classification unit (490a-3), an estimation value correction unit (490a-4), a trend analysis control unit (490a-5), and a reliability calculation control unit (490a-6).

[0467] [AI Estimation Model Algorithm Engine Unit (490a-1)]

[0469] The above AI estimation model algorithm engine unit (490a-1) receives multiple biosignal-based indices and calculates the blood ionized calcium status and estimated values.

[0470] This consists of a machine learning model unit, a deep learning inference unit, a time series analysis unit, a feature vector input unit, a model parameter storage unit, and a personalized model application unit.

[0472] In other words, it is configured to receive the QT risk index, HRV instability index, Pulse waveform index, Neural excitability index, Bio-Z variation index, skin temperature correction index, motion artifact score, and sensor quality score from the index calculation unit.

[0473] Then, after normalizing each input index to a range of 0 to 100, the Calcium risk score is calculated by assigning weights to the QT risk index, Neural excitability index, HRV instability index, Pulse waveform index, and Bio-Z variation index, respectively.

[0474] In addition, it consists of an ensemble model including an XGBoost-based current state classification model, an LSTM-based time series trend model, and a Bayesian correction model.

[0475] Here, the ensemble model outputs the estimated blood ionized calcium value, calcium status classification value, calcium decline trend, and estimation confidence at the current time point.

[0477] For example, when the QT risk index is 72, the HRV instability index is 58, the Pulse waveform index is 43, the Neural excitability index is 81, the Bio-Z variation index is 55, the motion artifact score is 18, and the sensor quality score is 91, the AI ​​estimation model algorithm engine calculates a Calcium risk score of 59.8 points, uses this to calculate an estimated blood ionized calcium value of 1.07 mmol / L, and classifies the user's calcium condition as a suspected mild low-calcium state.

[0479] Accordingly, the AI ​​estimation model algorithm engine unit according to the present invention estimates the state of ionized calcium in the blood using only a multi-non-invasive biosignal-based index without blood sampling, and transmits the result to the low calcium risk prediction control unit and the PTH automatic control signal generation unit.

[0481] [Personal Reference Value Storage Unit (490a-2)]

[0483] The above-mentioned individual reference value storage unit (490a-2) serves to store the normal biosignal pattern and individual reference calcium status for each user.

[0485] [Calcium State Classification Section (490a-3)]

[0487] The above calcium status classification unit (490a-3) serves to classify the AI ​​estimation results into clinically understandable status grades.

[0488] This consists of a normal state classification unit, a suspected low-calcium classification unit, a low-calcium risk classification unit, a high-calcium risk classification unit, an emergency risk classification unit, and a status grade output unit.

[0490] [Estimated Value Correction Section (490a-4)]

[0492] The above-mentioned estimation value correction unit (490a-4) serves to correct the AI ​​estimation value by reflecting movement, skin temperature, sensor quality, personal reference value, and recent test value.

[0493] This consists of a motion correction value input unit, a skin temperature correction value input unit, a sensor quality correction unit, an individual reference value correction unit, a recent blood test correction unit, and an abnormal value mitigation unit.

[0495] [Trend Analysis Control Unit (490a-5)]

[0497] The above trend analysis control unit (490a-5) plays the role of controlling not only the current calcium state but also the direction and speed of calcium change over time.

[0498] This consists of a time series storage unit, a moving average calculation unit, a rate of change calculation unit, a downward trend determination unit, a recovery trend determination unit, and a prediction input generation unit.

[0500] [Reliability Calculation Control Unit (490a-6)]

[0502] The above reliability calculation control unit (490a-6) plays the role of controlling the calculation of the reliability of the blood calcium estimation result.

[0503] This consists of a sensor quality input unit, a data omission determination unit, a model uncertainty calculation unit, an index agreement determination unit, a reliability score calculation unit, and a re-measurement request unit.

[0505] Thus, by configuring a blood calcium state estimation control unit comprising an AI estimation model algorithm engine unit, a personal reference value storage unit, a calcium state classification unit, an estimation value correction unit, a trend analysis control unit, and a reliability calculation control unit according to the present invention, the blood ionized calcium state can be estimated based on AI using the QT risk index, HRV instability index, Pulse waveform index, and Neural excitability index calculated by the index calculation unit, and by reflecting user-specific biosignal characteristics, movement state, skin temperature, sensor quality, and recent test values, the personalization accuracy of blood calcium state estimation can be increased by 1.5 to 3 times compared to the existing method, and by calculating the direction of decline, rate of decline, and estimation reliability of the blood calcium state together through the trend analysis control unit and the reliability calculation control unit, the low calcium risk prediction control unit and the PTH automatic control signal generation unit can improve safe operation by 80% compared to the existing method.

[0507] Eleventh, the low-calcium risk prediction control unit (490b) according to the present invention will be described.

[0508] The above low calcium risk prediction control unit (490b) plays a role in controlling the current calcium state and predicting future low calcium risk.

[0509] This consists of a time series prediction model, a hypocalcemia progression rate determination unit, a QT prolongation risk prediction unit, a neuromuscular spasm risk prediction unit, a risk grade calculation unit, and an alarm condition determination unit.

[0511] In other words, currently Ca 2+ Input the estimated value.

[0512] Next, the recent biosignal trend is input.

[0513] Next, time series forecasting is performed.

[0514] Next, the risk of low calcium is calculated.

[0515] Next, determine the risk level.

[0517] Twelfth, the photo-stimulation response signal generation control unit (490c) according to the present invention will be described.

[0518] The above photo-stimulation response signal generation control unit (490c) plays the role of controlling the generation of a photo-stimulation response signal to be sent to a photo-stimulation response body-implantable PTH secreting object located in the wrist muscle according to the blood calcium status and the risk of low calcium.

[0520] Thirteenth, an edge-type wireless communication unit (490d) according to the present invention will be described.

[0521] The above edge-type wireless communication unit (490d) is connected to a wrist-mounted smartwatch and a calcium balance smart app module via a wireless communication network and plays the role of transmitting and receiving data.

[0522] This consists of a BLE communication unit, a Wi-Fi communication unit, an LTE-M communication unit, and an RF transceiver.

[0523] In addition, it is located in a hospital or public institution within a range of 1m to 5000m and is configured to enable wireless connectivity, real-time transmission of photostimulation response signals, provision of app status display data, and response to communication failures.

[0524] The above edge-type wireless communication unit (490d) is configured to output a photo-stimulation response signal that causes the body-implantable PTH to be secreted toward a wrist-attached smartwatch and toward a photo-stimulation response body-implantable PTH secreting object located in the wrist muscle.

[0526] Fourteenth, the edge AI processor unit (490e) according to the present invention will be described.

[0527] The edge AI processor unit (490e) above plays the role of controlling AI inference, feature value calculation, risk prediction algorithm, and closed-loop control algorithm to be executed on-site.

[0528] This consists of an NPU, MCU, DSP, AI accelerator, model storage memory unit, security execution area unit, and power management unit.

[0530] Thus, by configuring a field edge-type AI module according to the present invention, comprising a data receiving unit, a preprocessing unit, a motion artifact removal unit, a feature value calculation control unit, a QT interval calculation control unit, an HRV calculation control unit, a pulse waveform analysis control unit, a neural excitability analysis control unit, an index calculation unit, a blood calcium status estimation control unit, a low calcium risk prediction control unit, a PTH automatic control signal generation control unit, an edge-type wireless communication unit, and an edge AI processor unit, the QT interval, HRV, pulse waveform, and neural excitability index are simultaneously indexed, and the blood calcium status estimation control unit analyzes them based on AI, thereby improving the reliability of estimating the blood ionized calcium status by 80% compared to a simple ECG-based QT prolongation judgment method; additionally, by the low calcium risk prediction control unit analyzing not only the current calcium status but also recent biosignal trends, QTc change rate, HRV decrease, and neural excitability increase together, the early prediction performance of low calcium risk can be improved by 2 to 5 times compared to the existing simple notification method. In addition, the edge AI processor unit executes AI inference, risk prediction, and closed-loop control algorithms on-site, thereby reducing the impact of communication delay to 70% or less compared to cloud server-dependent analysis. The edge wireless communication unit is linked with a wrist-worn smartwatch and a calcium balance smart app module, thereby improving the closed-loop transmission of calcium status display, PTH injection control, and risk notifications by 1.5 to 3 times compared to the existing unidirectional monitoring method. Furthermore, by generating a photo-stimulation response signal to stimulate a photo-stimulation-responsive body-implantable PTH-secreting object inside the wrist muscle based on the low-calcium risk prediction result, the PTH secretion induction response speed can be improved by 1.5 to 3 times compared to the existing manual treatment method.

[0532] Next, the calcium balance smart app module (500) according to the present invention will be described.

[0533] The above calcium balance smart app module (500) is activated on a smart device app for patients and a smart device app for guardians, and is connected to a field edge AI module via a wireless network, and displays the estimated blood calcium status, risk level, PTH infusion status, and biosignal status on the smart device screen, and performs the role of an emergency notification when there is a risk of low calcium.

[0534] As illustrated in FIGS. 26 and 27, this consists of a calcium balance type activity control unit (510), a calcium balance type service control unit (520), a calcium balance type broadcast receiver control unit (530), a calcium balance type content provider control unit (540), and a calcium balance type UI smart control unit (550).

[0536] First, the calcium balance type activity control unit (510) according to the present invention will be described.

[0537] The above calcium balance type activity control unit (510) plays the role of controlling the main screen, calcium status screen, PTH injection status screen, notification screen, and history inquiry screen that are displayed when the user runs the app, by activating and managing them.

[0538] As illustrated in FIGS. 27 to 34, this consists of a main activity section (511), a calcium status activity section (512), a biosignal monitoring activity section (513), a PTH injection status activity section (514), a danger notification activity section (515), a guardian monitoring activity section (516), and a setting activity section (517).

[0540] As illustrated in FIG. 28, the main activity unit (511) performs the role of a dashboard screen control unit that allows the user to check the current status at a glance when opening the app.

[0541] This is configured to display the current estimated blood calcium status, a summary of the risk of low calcium, a summary of the PTH infusion status, the sensor connection status, and the emergency alert status, and to control movement to each detailed screen.

[0543] As shown in FIG. 29, the calcium status activity unit (512) performs the role of a calcium status detail screen control unit that explains the blood calcium status estimated by the field edge-type AI module to the user.

[0544] This is Ca 2+ It consists of displaying estimated values, normal / caution / danger / emergency status, calcium upward / downward trends, low calcium risk scores, estimate reliability, and a graph of calcium changes over time.

[0546] As shown in FIG. 30, the above biosignal monitoring activity unit (513) performs the role of a sensor / biosignal monitoring screen control unit that checks whether the input biosignal used for calcium estimation is being measured normally.

[0547] This consists of ECG status display, HRV status display, PPG pulse wave status display, SpO₂ status display, EMG status display, Tremor status display, Bio-Z status display, skin temperature status display, and sensor quality status display.

[0549] As shown in FIG. 31, the above PTH injection status activity unit (514) performs the role of verifying whether PTH injection is being performed normally through a photo-stimulating response type body-implantable PTH secreting object in an emergency situation.

[0551] The above-mentioned danger notification activity unit (515) performs the role of an emergency alarm screen control unit that induces immediate response action from the user in a dangerous situation, as illustrated in FIG. 32.

[0552] This consists of a low calcium risk warning indicator, a QT prolongation warning indicator, an arrhythmia risk warning indicator, a PTH overdose blockage notification indicator, a sensor outage notification indicator, an emergency contact button indicator, and a guardian notification transmission status indicator.

[0554] As illustrated in FIG. 33, the above guardian monitoring activity unit (516) performs the role of a remote guardian monitoring screen control unit that allows the guardian to check the patient's calcium status and dangerous situation in real time.

[0555] This is configured to display the patient's current calcium status, risk of hypocalcemia, PTH infusion status, recent emergency alert history, sensor wearing status, provide a guardian confirmation button, and offer an emergency contact function.

[0557] As shown in FIG. 34, the above setting activity unit (517) performs the role of a setting screen control unit that sets the operating environment of the calcium balance smart app module.

[0558] This is configured to set up user profile settings, guardian contact settings, medical staff contact settings, sensor pairing settings, notification sound and vibration settings, risk alert criteria settings, and data sharing permissions.

[0560] Second, the calcium balance type service control unit (520) according to the present invention will be described.

[0561] The above calcium balance type service control unit (520) continues to communicate with the field edge type AI module even when the app is in the background state and plays the role of monitoring the calcium status and risk level.

[0562] As illustrated in FIG. 35, this consists of a background communication service unit (521), a calcium status receiving service unit (522), a risk level determination service unit (523), a guardian notification transmission service unit (524), and a data synchronization service unit (525).

[0564] The above background communication service unit (521) plays the role of maintaining wireless communication with the field edge AI module even when the app is turned off or the screen is locked.

[0565] This is configured to maintain a connection with the field edge AI module, monitor BLE / Wi-Fi / LTE communication status, detect communication disconnections, perform automatic reconnection, wait for data reception, and maintain background monitoring.

[0567] The above calcium status receiving service unit (522) plays the role of receiving the estimated blood ionized calcium status transmitted from the field edge type AI module.

[0568] This is Ca 2+ It consists of the estimated value, calcium status grade, normal / caution / danger / emergency status, low calcium risk score, calcium change trend, and estimate reliability.

[0570] The above risk judgment service unit (523) performs the role of determining whether to provide a notification by judging the risk level based on the received calcium status and biosignal status.

[0571] This consists of hypocalcium risk, hypercalcium risk, QT prolongation risk, arrhythmia risk, sensor out-of-sense risk, and communication failure risk.

[0573] The above guardian notification transmission service unit (524) plays the role of transmitting a notification to a guardian's smart device app when a risk event of low calcium risk or safety blockage is confirmed by the risk judgment service unit.

[0574] This consists of the patient's name or identification information, current calcium status, risk level, time of occurrence, emergency procedure guidance, optional provision of patient location information, and a confirmation request message.

[0576] The above data synchronization service unit (525) plays the role of synchronizing data between the patient smart device app, the guardian smart device app, and the medical staff viewing system.

[0577] This consists of calcium estimation history, risk history, PTH infusion history, vital sign summary information, notification occurrence history, guardian confirmation history, and untransmitted data after communication failure.

[0579] Third, the calcium balance type broadcast receiver control unit (530) according to the present invention will be described.

[0580] The above calcium balance type broadcast receiver control unit (530) detects system events and external events of the smart device and automatically controls the operation of the app.

[0581] As illustrated in FIG. 36, this consists of a boot completion receiver (531), a network change receiver (532), a low battery receiver (533), a Bluetooth connection status receiver (534), an emergency notification receiver (535), and a guardian call receiver (536).

[0583] The above boot completion receiver (531) detects the boot completion event of the operating system after the smart device power is turned on or rebooted, and automatically executes the Calcium Balance smart application module.

[0584] This is configured to enable app auto-launch, background communication service re-launch, edge AI module reconnection, and guardian integration recovery.

[0586] The above network change receiving unit (532) performs the role of detecting network status changes such as Wi-Fi, LTE, 5G, and Bluetooth.

[0587] This is configured to check communication status, reconnect edge AI modules, transmit undelivered data, and perform data synchronization.

[0589] The above battery low receiver (533) performs the role of monitoring the battery status of the smart device and detecting a battery low event.

[0590] This is configured to provide user alerts, parental notifications, switch to sleep mode, and maintain essential services.

[0592] The above Bluetooth connection status receiver (534) performs the role of monitoring the Bluetooth connection status with the smart ring type wearable sensor unit, the externally worn wearable patch unit, and the photo-stimulating response type body-implantable PTH secreting object.

[0593] This is configured to check the sensor connection, attempt to reconnect, and alert the user.

[0595] The above emergency notification receiving unit (535) performs the role of receiving emergency events transmitted from the field edge-type AI module.

[0596] This is configured to enable emergency pop-up display, warning sound output, vibration output, guardian notification, and emergency contact functions.

[0598] The above guardian call receiving unit (536) serves to receive call events transmitted from a guardian smart device app or a medical staff system.

[0599] This is configured to display the patient screen, output call notifications, provide location information, request a response, and connect with a guardian.

[0601] Fourth, the calcium balance type content provider control unit (540) according to the present invention will be described.

[0602] The above calcium balance type content provider control unit (540) performs the role of safely storing and providing app internal data and external linked data.

[0603] As illustrated in FIG. 37, this consists of a calcium estimate storage unit (541), a risk history storage unit (542), a biosignal summary storage unit (543), a guardian shared data management unit (544), a data provision unit for medical staff viewing (545), an access control unit (546), and an encryption storage unit (547).

[0605] The above calcium estimate storage unit (541) serves to store the blood ionized calcium estimate transmitted from the field edge type AI module.

[0606] This is Ca 2+ It consists of the estimated value, calcium status grade, normal / caution / danger / emergency status, estimate reliability, and measurement time.

[0608] The above risk history storage unit (542) serves to store the history of risk events such as low calcium risk, high calcium risk, QT prolongation risk, and arrhythmia risk.

[0609] This consists of the time of risk occurrence, type of risk, risk level, duration of risk, whether a notification was generated, user confirmation, and guardian confirmation.

[0611] The above biosignal summary storage unit (543) serves to summarize and store the first non-invasive biosignal data, the second non-invasive biosignal data, and the 3-axis acceleration data transmitted through the field edge AI module.

[0612] This consists of QT interval, QTc, HRV, heart rate, PPG pulse wave index, neural excitability index, EMG index, tremor index, Bio-Z index, skin temperature, and sensor quality score.

[0614] The above guardian shared data management unit (544) plays the role of selecting and managing information to be shared with the guardian's smart device app.

[0615] This consists of current calcium status, risk level, emergency alerts, PTH infusion status, sensor wearing status, and guardian verification history.

[0617] The above-mentioned data provision unit (545) for medical staff viewing serves to provide data so that medical staff can check the patient's calcium status, risk history, PTH injection history, and biosignal trends.

[0618] This is daily Ca 2+ It consists of estimated trends, number of hypocalcemia risk occurrences, QTc change trends, HRV change trends, total PTH injection volume, PTH response patterns, and sensor quality history.

[0620] The above access control unit (546) serves to control the range of data that patients, guardians, medical staff, and system administrators can access.

[0622] The above-mentioned encrypted storage unit (547) encrypts and stores health information and external linked data stored inside the app.

[0624] Fifth, the calcium balance type UI smart control unit (550) according to the present invention will be described.

[0625] The above calcium balance type UI smart control unit (550) controls the display of the estimated blood calcium status, risk level, PTH injection status, and biosignal status intuitively on the smart device screen, and controls the color, vibration, voice, and pop-up notifications according to the risk level.

[0626] As illustrated in FIG. 38, this consists of a calcium status display control unit (551), a risk color display control unit (552), a biosignal status display control unit (553), a real-time graph display control unit (554), a guardian sharing screen control unit (555), an emergency notification pop-up control unit (556), and a vibration, sound, and voice guidance control unit (557).

[0628] The above calcium status indicator control unit (551) is blood ionized calcium (Ca 2+ It plays a role in controlling the display of estimation results in a form that is easy for users and guardians to understand.

[0629] This consists of a calcium estimate display section, a calcium status grade display section, a calcium change trend display section, an estimate reliability display section, and a reference range display section.

[0631] The above risk level color display control unit (552) serves to control the low-calcium risk level to be intuitively displayed by color.

[0632] This consists of a risk judgment value input unit, a color mapping unit, a warning level display unit, and a risk color output unit.

[0634] The above biosignal state display control unit (553) serves to control the display of the biosignal state used for blood calcium estimation.

[0635] This consists of an ECG status display, a QT status display, an HRV status display, a PPG status display, an EMG status display, a Tremor status display, a Bio-Z status display, and a skin temperature display.

[0637] The above real-time graph display control unit (554) plays the role of controlling the display of changes in calcium status and biosignals over time as a graph.

[0638] This consists of a time series graph generation unit, a zoom in / out control unit, an interval selection unit, an event marking unit, and a trend line generation unit.

[0640] The above guardian sharing screen control unit (555) serves to control the display of the patient's condition on the guardian app screen.

[0641] This consists of a patient status display, a risk level display, an emergency notification display, a location information display, and a guardian confirmation button.

[0643] The above emergency notification popup control unit (556) plays the role of controlling the display of a popup by force when an emergency situation occurs.

[0644] This consists of an emergency event input section, a popup generation section, a full-screen display section, a response button generation section, and an emergency contact execution section.

[0646] The above vibration, sound, and voice guidance control unit (557) plays the role of controlling to output vibration, warning sound, and voice guidance according to the level of danger.

[0647] This consists of a vibration output unit, a warning sound output unit, a voice synthesis output unit, a volume control unit, and a notification pattern control unit.

[0649] Thus, by configuring a calcium balance smart application module according to the present invention, which consists of a calcium balance type activity control unit, a calcium balance type service control unit, a calcium balance type broadcast receiver control unit, a calcium balance type content provider control unit, and a calcium balance type UI smart control unit, it is possible to provide the estimated blood ionized calcium status, low calcium risk level, PTH infusion status, and biosignal status in real time. Furthermore, through risk level color display, vibration, warning sound, voice guidance, and emergency pop-up functions, the user's awareness time regarding a low calcium risk situation can be reduced to 80% or less compared to the existing method, and emergency responsiveness can be improved by 1.5 to 3 times compared to the existing method. Additionally, since long-term calcium change trends and PTH administration history can be managed through a calcium estimate value storage unit, a risk level history storage unit, a PTH infusion history storage unit, and a biosignal summary storage unit, the clinical utility for medical staff can be improved by 80% compared to the existing method.

[0651] Next, a photo-stimulating response body-implantable PTH-secreting object (600) according to the present invention will be described.

[0652] The above-mentioned photo-stimulation-responsive body-implantable PTH-secreting object (600) is implanted into the user's wrist muscle and plays the role of secreting body-implantable PTH (Parathyroid Hormone) in response to a photo-stimulation signal from a wrist-attached smartwatch.

[0653] Here, in addition to the user's wrist muscles, photosensitive body-implantable PTH-secreting objects are formed on skin areas located on the user's body, specifically the abdomen, upper arm, and thigh.

[0654] In other words, in the case of normal parathyroid glands, blood Ca 2+ If it decreases, it stimulates the secretion of PTH, and Ca 2+ Increases.

[0655] The above-mentioned photo-stimulated body-implantable PTH-secreting object is an application of body-implantable PTH (Parathyroid Hormone) utilizing a body-responsive organoid as described in Patent Registration No. 10-2917672, which was filed and registered by the applicant, and is an artificial mini-organ that responds organically and appropriately with the body. Here, "mini-organ" refers to a three-dimensional (3D) mini-cluster capable of performing the biological functions of an organ (organ) within the human body, and it is not necessary to reproduce the structure of an actual organ or organ within the human body that performs the intended biological functions. Specifically, the "photostimulation-responsive body-implantable PTH-secreting object" according to the present invention is a three-dimensionally cultured artificial mini-organ containing functional cells that secrete body-implantable PTH (Parathyroid Hormone) by being cultured from a human-derived cell line, and is designed so that when it receives photostimulation, the body-implantable PTH (Parathyroid Hormone) located in the wrist muscle is delivered to various parts of the body, and the body-implantable PTH (Parathyroid Hormone) produced by the functional cells contained therein is released through an organically compatible reaction with the body, thereby enabling the administration (delivery) of body-implantable PTH (Parathyroid Hormone) into the body.

[0657] The above-mentioned photo-stimulating type body-implantable PTH secreting object (600) is formed as a component of a spherical body (610) made of biocompatible materials such as polymer and silica, with a diameter of 0.01 μm to 100 μm, and functional cells of organoid scaffolded hormone particles are embedded in the internal space based on the spherical axis, and is composed of a triple membrane structure (Bioshield).

[0658] The photo-stimulated body-implantable PTH-secreting object according to the present invention secretes body-implantable PTH (Parathyroid Hormone) in response to a photo-stimulated signal. As shown in FIG. 39, functional cells (620) are created along the circumference based on a central empty space, and after forming them into an organoid scaffold (630), a spherical body that acts as a bioshield surrounding it is formed. Based on a tissue engineering-based hormone evaluation platform, it is introduced into an endocrine area to undergo a bio-response and is used for the purpose of treatment and continuous secretion of body-implantable PTH (Parathyroid Hormone).

[0659] Here, the photo-stimulated body-implantable PTH-secreting object is an artificial mini-organ cultured in three dimensions (3D) from a human-derived cell line, and refers to a functional cell that acts as an organ for secreting body-implantable PTH (Parathyroid Hormone).

[0660] More specifically, the functional cell according to the present invention is a cell designed to perform a specific pharmacological or biotherapeutic function, and is generated through the process of securing an initial cell source, gene editing and modification, and 3D cell culture.

[0661] In securing the initial cell source mentioned above, a source is obtained by selecting one of adult stem cells, induced pluripotent stem cells, or cells derived from specific tissues. The adult stem cells utilized are those harvested from internal tissues. The induced pluripotent stem cells (iPSCs) are used after adult cells have been reverted to a differentiation-capable state through genetic reprogramming. The cells derived from specific tissues are converted from pancreatic cells (insulin-producing), parathyroid hormone-secreting cells, and liver cells (detoxification function) into specific cells.

[0662] The aforementioned gene editing and modification utilize gene editing technology to manipulate genes within cells to induce the production of specific hormones or proteins (e.g., hormones, anticancer drugs). Additionally, retrovirus / lentivirus-based genes are inserted to confer desired characteristics upon the cells.

[0663] The above 3D cell culture cultures cells in a three-dimensional environment to integrate functional cells with a photosensitive body-implantable PTH-secreting object, and controls oxygen, pH, growth factors, and nutrients to mimic the environment within an adult.

[0665] The above organoid scaffold is a structure designed to physically support functional cells having a 3D structure and to reproduce a biological environment, and consists of a substrate (material), a 3D structure, and a bioactive agent.

[0666] The substrate (material) is composed of, for example, biocompatible materials such as hydrogels (collagen, Matrigel), nanofibers, or biodegradable polymers (PCL, PLGA, polymers, silica), and the 3D structure is designed to have a porous (high surface area) and flexible structure so that cells can attach, grow, and differentiate. It is also designed to be delivered into the body where blood Ca2+ is needed in response to a photostimulation reaction.

[0667] In addition, bioactivators contain growth factors (VEGF, FGF), proteins, and cell adhesion factors, which promote cell growth and function.

[0668] The functional cells generated in this way efficiently reach the target site, receive oxygen or nutrients from the body site, and undergo a biological response to release implantable PTH (Parathyroid Hormone), thereby administering implantable PTH.

[0670] And, the photo-stimulating response type body-implantable PTH secreting object according to the present invention is positioned in the wrist muscle area, which is the location of a wrist-attached smartwatch, through a syringe needle as shown in FIG. 40.

[0671] In addition, multiple photo-stimulated body-implantable PTH-secreting objects with a diameter of 0.01 μm to 100 μm are formed.

[0673] The tissue engineering-based hormone evaluation platform according to the present invention consists of a system that verifies and completes the hormone efficacy testing, toxicity evaluation, and human biological response of organoid-scaffolded functional cells.

[0675] The photo-stimulated implantable PTH-secreting object according to the present invention performs photo-stimulated implantable PTH (Parathyroid Hormone) to overcome the limitations of conventional mechanically regulating PTH hormone, and is utilized as an "auxiliary control means" to prepare for stabilization immediately after implantation, acute hypocalcemia, rapid calcium drop after surgery, and the risk of seizures.

[0677] Thus, by configuring a photo-responsive body-implantable PTH-secreting object, including functional cells cultured from human-derived cell lines, an organoid scaffold, and a biocompatible spherical body, it is possible to generate and release body-implantable PTH in response to photo-stimulation. By culturing the functional cells into a three-dimensional organoid structure, the cell viability can be improved by 80% compared to the existing method, and the PTH generation efficiency can be improved by 1.5 to 3 times compared to the existing method. Furthermore, by dispersing and implanting multiple photo-responsive body-implantable PTH-secreting objects within the wrist muscle, the efficiency of a photo-stimulation-based closed-loop control consisting of photo-stimulation -> functional cell activation -> PTH generation -> PTH release -> blood calcium recovery can be improved by 80% compared to the existing method.

[0679] Hereinafter, a specific process of the body-implantable PTH secretion closed-loop control method using a blood ionized calcium estimation formula of a non-invasive multi-biosignal-based field edge AI according to the present invention will be described.

[0681] FIG. 42 is a flowchart illustrating a body-implantable PTH secretion closed-loop control method for estimating blood ionized calcium using a non-invasive multi-biometric signal-based field edge AI according to the present invention.

[0683] First, through a smart ring-type wearable sensor unit, first non-invasive biosignal data regarding PPG (Photoplethysmography), blood oxygen saturation (SpO2), bioimpedance, skin temperature, tremor, and electromyography (EMG) is sensed by wearing it on a finger, and the sensed first non-invasive biosignal data is transmitted to a wrist-mounted smartwatch (S10).

[0684] That is, as illustrated in FIG. 43, LED light is irradiated onto the skin through the PPG photoplethysmography sensor unit, and reflected light is detected to sense changes in peripheral blood flow and pulse waves in the finger (S11).

[0685] Next, blood oxygen saturation is calculated using the difference in absorbance between red light and infrared light through the SpO₂ oxygen saturation measuring unit (S12).

[0686] Next, a micro alternating current is applied to the finger tissue through a bioimpedance measuring unit, and voltage changes are measured to calculate indicators related to tissue impedance, body fluid changes, and electrolyte changes (S13).

[0687] Next, the surface temperature of the finger skin is sensed through the skin temperature sensor (S14).

[0688] Next, the fine tremor, vibration, and abnormal movement of the finger are sensed through the tremor detection sensor (S15).

[0689] Next, a fine electromyogram signal corresponding to finger muscle or neuromuscular excitation is measured through the finger EMG electrode (S16).

[0690] Next, through the first signal processing unit, the raw signals collected from the PPG photoplethysmography sensor unit, SpO₂ oxygen saturation measurement unit, bioimpedance measurement unit, skin temperature sensor unit, tremor detection sensor unit, and finger EMG electrode unit are amplified, filtered, digitally converted, and feature extracted to be converted into first non-invasive biosignal data (S17).

[0691] Next, the first non-invasive biosignal data generated by the first signal processing unit is wirelessly transmitted to the wrist-worn smartwatch through the ring-type wireless communication unit (S18).

[0693] Next, through an external wearable patch, a second non-invasive biosignal data regarding ECG (Electrocardiography) signals, heart rate, and heart rate variability flowing from the surface of the chest skin is sensed, and the sensed second non-invasive biosignal data is transmitted to a wrist-mounted smartwatch (S20).

[0694] That is, as shown in FIG. 44, the patch-type adhesive substrate part is placed on the lower surface of the entire patch that is in direct contact with the surface of the chest skin, thereby adhering the ECG electrode part and the electronic circuit to the skin and supporting the patch so that the patch position does not deviate (S21).

[0695] Next, a minute electrical signal generated from the heart and transmitted to the surface of the chest skin is detected through the ECG electrode (S22).

[0696] Next, the minute electrical signal measured at the ECG electrode is amplified to an analyzable size through the ECG amplification unit (S23).

[0697] Next, power noise, electromyography noise, motion noise, and baseline fluctuations are filtered from the ECG signal amplified by the ECG amplifier through the ECG filter section (S24).

[0698] Next, the heart rate is calculated by detecting the R wave of the ECG signal through the heart rate calculation unit (S25).

[0699] Next, heart rate variability data is calculated by analyzing changes in RR intervals through the HRV calculation unit (S26).

[0700] Next, raw ECG data or high-quality filtered ECG data for a certain time interval is stored through the raw ECG storage unit for QT·QTc extraction (S27).

[0701] Next, second non-invasive biosignal data regarding ECG (Electrocardiography) signals, heart rate, and heart rate variability is wirelessly transmitted to a wrist-worn smartwatch through a patch wireless communication unit (S28).

[0703] Next, as illustrated in FIG. 41, a wrist-mounted smartwatch is worn on the user's wrist to measure 3-axis acceleration data regarding the user's moving posture, and then the first non-invasive biosignal data transmitted from the smart ring-type wearable sensor unit and the second non-invasive biosignal data transmitted from the external wearable patch unit are collected and transmitted to the field edge-type AI module via a wireless communication network (S30).

[0704] That is, as illustrated in FIG. 45, the linear acceleration change in the X-axis, Y-axis, and Z-axis directions according to the user's wrist movement is measured through the 3-axis acceleration sensor unit (S31).

[0705] Next, angular velocity data corresponding to the user's wrist rotation, twisting, and change of direction is measured through the gyro sensor (S32).

[0706] Next, through the watch-type data receiver, the first non-invasive biosignal data transmitted from the smart ring-type wearable sensor unit and the second non-invasive biosignal data transmitted from the external wearable patch unit are received (S33).

[0707] Next, through the time synchronization unit, the first non-invasive biosignal data, the second non-invasive biosignal data, and the 3-axis acceleration data of different time axes generated from the smart ring type wearable sensor unit, the external wearable patch unit, and the wrist-attached smartwatch itself are synchronized to the same reference time (S34).

[0708] Next, through the data packet generation unit, the synchronized first non-invasive biosignal data, second non-invasive biosignal data, and 3-axis acceleration data are generated into a data packet that can be analyzed by the field edge AI module (S35).

[0709] Next, through the watch-type wireless communication unit, the integrated data packet is transmitted to the field edge-type AI module via the wireless communication network (S36).

[0711] Next, using the first non-invasive biosignal data, the second non-invasive biosignal data, and the 3-axis acceleration data transmitted from the wrist-worn smartwatch via the field edge AI module, the QT (Q wave to T wave) interval, HRV (Heart Rate Variability), Pulse waveform, and neural excitability index are calculated, and then the blood calcium status is estimated (S40).

[0712] That is, as illustrated in FIG. 46, an integrated data packet is received from a wrist-worn smartwatch through a data receiving unit (S41).

[0713] Next, in the preprocessing unit, noise, abnormal values, and missing values ​​of the biosignal received through the data receiving unit are corrected (S42).

[0714] Next, in the motion artifact removal unit, PPG, ECG, and EMG signal distortions caused by motion are removed using 3-axis acceleration data received through the data receiving unit (S43).

[0715] Next, the feature value calculation control unit controls the integration of feature values ​​necessary for estimating blood calcium status from the first non-invasive biosignal data, the second non-invasive biosignal data, and the 3-axis acceleration data (S44).

[0716] Next, the QT interval operation control unit is used to calculate the QT interval from the Q wave start point to the T wave end point and QTc from the ECG signal (S45).

[0717] Next, the HRV calculation control unit is controlled to analyze changes in the RR interval and calculate a heart rate variability index (S46).

[0718] Next, the pulse waveform analysis control unit analyzes the PPG-based pulse waveform to calculate peripheral blood flow, vascular tone, and pulse wave changes (S47).

[0719] Next, changes in neuromuscular excitability are analyzed using EMG, Tremor, bioimpedance, and skin temperature data through the neural excitability analysis control unit (S48).

[0720] Next, through the index calculation unit, the results of QT, HRV, Pulse waveform, and Neural excitability are quantified and indexed (S49).

[0721] Next, through the blood calcium state estimation control unit, the blood ionized calcium state is controlled to be estimated based on AI using the data calculated from the index calculation unit (S49a).

[0723] Next, as illustrated in FIG. 41, the estimated blood calcium status is transmitted to the calcium balance smart application module through the field edge AI module (S50).

[0724] That is, as illustrated in FIG. 47, the low calcium risk prediction control unit controls the current calcium state risk and the future low calcium risk to be predicted (S51).

[0725] Next, through the edge-type wireless communication unit, it is connected to the Calcium Balance smart app module via a wireless communication network to transmit and receive data (S52).

[0727] Next, as illustrated in FIG. 41, the estimated blood calcium status, risk level, PTH infusion status, and biosignal status are displayed on the smart device screen through the calcium balance smart application module, and an emergency notification is performed in case of low calcium risk (S60).

[0729] Next, as illustrated in FIG. 41, a photo-stimulation response signal is output to secrete body-implantable PTH toward a wrist-attached smartwatch located in the wrist muscle using the estimated blood calcium status through the field edge-type AI module (S70).

[0730] That is, as illustrated in FIG. 48, the photo-stimulation response signal generation control unit controls the generation of a photo-stimulation response signal to be sent to a wrist-worn smartwatch according to the blood calcium status and the risk of low calcium (S71).

[0731] Next, through the edge-type wireless communication unit, the wrist-worn smartwatch is connected to a wireless communication network to transmit and receive data (S72).

[0733] Finally, through a photo-stimulating response body-implantable PTH-secreting object, body-implantable PTH (Parathyroid Hormone) is secreted in response to a photo-stimulating signal from a wrist-attached smartwatch (S80).

[0734] And, through the field edge-type AI module, a closed-loop control is performed to repeat the above steps (S10) to (S80). Explanation of the symbols

[0736] 1: Non-invasive multi-biosignal-based field edge AI blood ionized calcium estimation formula implantable PTH secretion closed-loop controller 100 : Smart ring-type wearable sensor unit 110 : Ring-shaped main body 120 : PPG Photoplethysmography Sensor Unit 130 : SpO₂ oxygen saturation measuring unit 140: Bioimpedance measurement unit 150 : Skin temperature sensor part 160 : Tremor detection sensor unit 170: Finger EMG electrode 180: 1st signal processing unit 190 : Ring-type wireless communication unit 190a: Ring-type battery section 200 : External wearable patch part 210 : Patch-type adhesive substrate part 220 : ECG electrode section 230 : ECG amplifier 240 : ECG filter section 250 : Heart rate calculation unit 260 : HRV Calculation Section 270 : Raw ECG storage unit for QT·QTc extraction 280 : Wireless communication unit for patch 290: Battery for patch 300: Wrist-mounted smartwatch 310: Watch-type main body 320: 3-axis accelerometer sensor unit 330: Gyro sensor unit 340: Watch-type data receiver 350 : Time Synchronizer 360: Data packet generation section 370: Watch-type wireless communication unit 380: Watch-type memory unit 390 : User Notification Section 390a: Watch-type battery unit 390b: Photostimulation section 400: On-site edge AI module 410 : Data receiving unit 420 : Preprocessing section 430 : Movement Artifact Removal Section 440 : Feature value operation control unit 450 : QT interval operation control unit 460 : HRV Operation Control Unit 470 : Pulse waveform analysis control unit 480: Neural excitability analysis and control unit 490 : Index calculation unit 490a : Blood calcium status estimation control unit 490b : Low-calcium risk prediction and control unit 490c: Photostimulation response signal generation control unit 490d: Edge-type wireless communication unit 490e: Edge AI processor unit 500: Calcium Balance Smart App Module 510: Calcium Balance Type Activity Control Unit 520 : Calcium Balance Type Service Control Unit 530: Calcium Balance Type Broadcast Receiver Control Unit 540: Calcium Balance Type Content Provider Control Unit 550: Calcium Balance Type UI Smart Control Unit 600 : Photostimulation-responsive body-implantable PTH-secreting object

Claims

Claim 1 A non-invasive multi-biometric signal-based field edge AI blood ionized calcium estimation type implantable PTH secretion closed-loop control device characterized by attaching or wearing it to a user's body, using sensed biosignal data to estimate the blood calcium status, and using the estimated blood calcium status to output a photostimulation response signal that secretes implantable PTH toward a photostimulation response type implantable PTH secreting object, thereby controlling the closed loop. Claim 2 In claim 1, the blood ionized calcium estimation type body-implantable PTH secretion closed-loop control device is formed as a ring structure and is worn on a finger to sense first non-invasive biosignal data regarding PPG (Photoplethysmography), blood oxygen saturation (SpO2), bioimpedance, skin temperature, tremor, and electromyography (EMG), and transmits the sensed first non-invasive biosignal data to a wrist-attached smartwatch; an externally worn wearable patch part (200) attached to the surface of the chest skin to sense second non-invasive biosignal data regarding ECG (Electrocardiography) signals, heart rate, and heart rate variability flowing from the surface of the chest skin, and transmits the sensed second non-invasive biosignal data to a wrist-attached smartwatch; and 3-axis acceleration data regarding the posture in which the user moves, which is worn on the user's wrist. A wrist-mounted smartwatch (300) that, after measurement, collects the first non-invasive biosignal data transmitted from the smart ring-type wearable sensor unit and the second non-invasive biosignal data transmitted from the external wearable patch unit, transmits them to the field edge-type AI module via a wireless communication network, receives a photo-stimulation response signal from the field edge-type AI module, and photostimulates the photo-stimulation response type body-implantable PTH-secreting object implanted inside the PTH minor neck muscle, and is located in a hospital or public institution within a short distance of 1m to 5000m relative to the wrist-mounted smartwatch and is connected to the wrist-mounted smartwatch via a wireless communication network, and uses the first non-invasive biosignal data, second non-invasive biosignal data, and 3-axis acceleration data transmitted from the wrist-mounted smartwatch to determine the QT (Q wave to T wave) interval, HRV (Heart Rate Variability), Pulse waveform, and neural excitability index. After calculating the indicator, the blood calcium status is estimated, andA non-invasive multi-biosignal-based field system characterized by comprising: a field edge type AI module (400) that transmits an estimated blood calcium status to a calcium balance smart app module and outputs a photo-stimulation response signal to secrete body-implantable PTH to a photo-stimulation response type body-implantable PTH secreting object located in the wrist muscle using the estimated blood calcium status and performs closed-loop control; a calcium balance smart app module (500) that is activated on a patient smart device app and a guardian smart device app, is connected to the field edge type AI module via a wireless network, displays the estimated blood calcium status, risk level, PTH injection status, and biosignal status on the smart device screen, and performs an emergency notification role in case of low calcium risk; and a photo-stimulation response type body-implantable PTH secreting object (600) that is implanted in the user's wrist muscle and secretes body-implantable PTH (Parathyroid Hormone) according to the photo-stimulation signal of a wrist-attached smartwatch. Edge-type AI blood ionized calcium estimation formula body-implantable PTH secretion closed-loop controller. Claim 3 In paragraph 2, the smart ring-type wearable sensor unit (100) is formed in a ring shape for inserting a finger and comprises: a ring-type main body unit (110) that protects and supports each device from external pressure; a PPG photoplethysmography sensor unit (120) located on one side of the inner surface of the ring-type main body unit, which irradiates LED light onto the skin and detects reflected light to sense changes in peripheral blood flow and pulse waves of the finger; an SpO₂ oxygen saturation measurement unit (130) located on one side of the PPG photoplethysmography sensor unit, which calculates blood oxygen saturation using the difference in absorbance between red light and infrared light; a bioimpedance measurement unit (140) located on one side of the SpO₂ oxygen saturation measurement unit, which applies a micro alternating current to finger tissue and measures voltage changes to calculate indicators related to tissue impedance, body fluid changes, and electrolyte changes; and a skin temperature sensor located on the inner skin contact surface of the ring-type main body unit, which senses the surface temperature of the finger skin. A sensor unit (150), a tremor detection sensor unit (160) located on a ring-shaped internal PCB for sensing fine tremor, vibration, and abnormal movement of the finger, a finger EMG electrode unit (170) located on the other side of the PPG photoplethysmography sensor unit for measuring fine electromyography signals according to finger muscle or neuromuscular excitation, a first signal processing unit (180) located on a ring-shaped internal PCB for amplifying, filtering, digitally converting, and feature extracting raw signals collected from the PPG photoplethysmography sensor unit, SpO₂ oxygen saturation measurement unit, bioimpedance measurement unit, skin temperature sensor unit, tremor detection sensor unit, and finger EMG electrode unit to convert them into first non-invasive biosignal data, a ring-shaped wireless communication unit (190) located on one side of the first signal processing unit for wirelessly transmitting the first non-invasive biosignal data generated by the first signal processing unit to a wrist-worn smartwatch, and a ring-shaped internal PCB formed in an arc-shaped structure on one side of the ring-shaped internal PCB for supplying power to each device. A non-invasive multi-biometric signal-based field edge AI body-implantable PTH secretion closed-loop control device characterized by being composed of a battery unit (190a). Claim 4 In paragraph 2, the external wearable patch portion (200) comprises: a patch-type adhesive substrate portion (210) positioned on the lower surface of the entire patch that is in direct contact with the surface of the chest skin, which supports the ECG electrode portion and electronic circuit to adhere to the skin and prevents the patch from deviating; an ECG electrode portion (220) positioned on the skin contact surface of the patch-type adhesive substrate portion to detect minute electrical signals generated from the heart and transmitted to the surface of the chest skin; an ECG amplifier portion (230) positioned on one side of the ECG electrode portion to amplify the minute electrical signals measured by the ECG electrode portion to an analyzable size; an ECG filter portion (240) positioned on one side of the rear end of the ECG amplifier portion to filter power noise, electromyography noise, motion noise, and baseline fluctuations from the ECG signal amplified by the ECG amplifier portion; a heart rate calculation portion (250) positioned on one side of the rear end of the ECG filter portion to detect the R-wave of the ECG signal and calculate the heart rate; and a heart rate calculation portion A non-invasive multi-biometric signal-based field edge AI body-implantable PTH secretion closed-loop control device characterized by comprising: an HRV calculation unit (260) located on one side that analyzes changes in RR intervals to calculate heart rate variability data; a raw ECG storage unit (270) located on one side of a memory unit inside a patch-type adhesive substrate unit that stores raw ECG or high-quality filtered ECG data of a certain time interval for calculating QT interval and QTc; a patch wireless communication unit (280) located on a flexible PCB of the patch-type adhesive substrate unit that wirelessly transmits second non-invasive biosignal data regarding ECG (Electrocardiography) signals, heart rate, and heart rate variability to a wrist-attached smartwatch; and a patch battery unit (290) located on one side of the upper part of the patch-type adhesive substrate unit and formed in a curved structure to supply power to each device. Claim 5 In paragraph 2, the wrist-mounted smartwatch (300) is formed in the shape of a wristwatch worn on a user's wrist and comprises a watch-type main body (310) that protects and supports each device from external pressure, a 3-axis acceleration sensor unit (320) located on the internal PCB of the watch-type main body and measuring linear acceleration changes in the X-axis, Y-axis, and Z-axis directions according to the user's wrist movement, a gyro sensor unit (330) located on one side of the 3-axis acceleration sensor unit and performing the role of measuring angular velocity data according to the user's wrist rotation, twisting, and change of direction, a watch-type data receiver (340) located on one side of the gyro sensor unit and receiving first non-invasive biosignal data transmitted from a smart ring-type wearable sensor unit and second non-invasive biosignal data transmitted from an external wearable patch unit, and a smart ring-type wearable sensor unit, an external wearable patch unit, and a wrist-mounted smartwatch located on one side of the watch-type data receiver unit. A time synchronization unit (350) that synchronizes first non-invasive biosignal data, second non-invasive biosignal data, and 3-axis acceleration data of different time axes generated internally to the same reference time; a data packet generation unit (360) located on one side of the time synchronization unit that generates the synchronized first non-invasive biosignal data, second non-invasive biosignal data, and 3-axis acceleration data into data packets that can be analyzed by a field edge-type AI module; a watch-type wireless communication unit (370) located on one side of the data packet generation unit that transmits the integrated data packet to a field edge-type AI module via a wireless communication network; a watch-type memory unit (380) located on one side of the watch-type wireless communication unit that stores first non-invasive biosignal data, second non-invasive biosignal data, 3-axis acceleration data, communication status, notification history, and transmission failure data; and a sensor connection status, data transmission status, battery low, wearing failure, and low calcium located on one side of the front display window of the watch-type main body. A user notification unit (390) that performs the role of notifying the user of danger, and located on one side of the lower part of the internal space,A non-invasive multi-biosignal-based field-edge AI blood ionized calcium estimation type body-implantable PTH secretion closed-loop control device, characterized by comprising a watch-type battery unit (390a) that supplies power to each device of a wrist-attached smartwatch, and a photo-stimulation unit (390b) located on a skin contact surface on one side of the bottom surface of a watch-type main body unit that contacts the wrist muscle, which photostimulates a photo-stimulation response type body-implantable PTH secretion object implanted inside the wrist muscle according to a photo-stimulation response signal received from a field-edge type AI module. Claim 6 In paragraph 2, the field edge-type AI module (400) comprises a data receiving unit (410) that receives an integrated data packet from a wrist-mounted smartwatch, a preprocessing unit (420) that corrects noise, abnormal values, and missing values ​​of the biosignal received through the data receiving unit, a motion artifact removal unit (430) that removes PPG, ECG, and EMG signal distortions caused by movement using 3-axis acceleration data received through the data receiving unit, a feature value calculation control unit (440) that controls the integrated calculation of feature values ​​necessary for estimating blood calcium status from the first non-invasive biosignal data, the second non-invasive biosignal data, and the 3-axis acceleration data, a QT interval calculation control unit (450) that controls the calculation of the QT interval and QTc from the Q-wave start point to the T-wave end point in the ECG signal, an HRV calculation control unit (460) that controls the calculation of a heart rate variability index by analyzing changes in the RR interval, and PPG-based A pulse waveform analysis control unit (470) that analyzes a pulse waveform to calculate peripheral blood flow, vascular tone, and pulse wave changes; a neural excitability analysis control unit (480) that analyzes neuromuscular excitability changes using EMG, Tremor, bioimpedance, and skin temperature data; an index calculation unit (490) that quantifies and indexes QT, HRV, pulse waveform, and neural excitability results; a blood calcium state estimation control unit (490a) that controls the estimation of the blood ionized calcium state based on AI using the data calculated from the index calculation unit; a low calcium risk prediction control unit (490b) that controls the prediction of the risk in the current calcium state and the future low calcium risk; and a photostimulation response signal that controls the generation of a photostimulation response signal to be sent toward a photostimulation response body-implantable PTH secreting object located in the wrist muscle according to the blood calcium state and the low calcium risk. A generation control unit (490c) and a wrist-mounted smartwatch,A non-invasive multi-biosignal-based field edge-type AI body-implantable PTH secretion closed-loop control device for estimating blood ionized calcium, characterized by being composed of an edge-type wireless communication unit (490d) that transmits and receives data by being connected to a calcium balance smart application module via a wireless communication network, and an edge AI processor unit (490e) that controls AI inference, feature value calculation, risk prediction algorithm, and closed-loop control algorithm to be executed on-site. Claim 7 In claim 6, the feature value calculation control unit (440) comprises an ECG feature extraction unit (441) that extracts cardiac electrical conduction-related feature values ​​from an ECG signal transmitted from an external wearable patch unit, a PPG feature extraction unit (442) that extracts peripheral blood flow and pulse wave-related feature values ​​from a PPG signal transmitted from a smart ring-type wearable sensor unit, an EMG feature extraction unit (443) that extracts neuromuscular excitability-related feature values ​​from an electromyography signal measured by a finger EMG electrode unit, a Tremor feature extraction unit (444) that extracts fine tremor feature values ​​of a finger or wrist from 3-axis acceleration data from a Tremor detection sensor unit, a Bio-Z feature extraction unit (445) that extracts tissue conductivity, body fluid change, and electrolyte change-related feature values ​​from a Bio-Z signal measured by a bioimpedance measurement unit, and a skin temperature feature that extracts peripheral circulation and external environment correction feature values ​​from finger skin temperature data measured by a skin temperature sensor unit. A non-invasive multi-biometric signal-based field edge AI body-implantable PTH secretion closed-loop control device characterized by being composed of an extraction unit (446) and a feature value integration unit (447) that forms a single integrated feature vector from feature values ​​calculated in each feature extraction unit. Claim 8 In claim 6, the QT interval calculation control unit (450) is characterized by being composed of: a QRS detection unit (451) that detects a QRS complex in an ECG signal and sets the R-wave position as a reference point; a Q-wave start point detection unit (452) that detects the point where the Q-wave starts within the QRS complex and determines the start point of the QT interval; a T-wave end point detection unit (453) that detects the point where the T-wave ends in an ECG signal and determines the end point of the QT interval; a QT interval calculation unit (454) that calculates the QT interval by calculating the time interval between the Q-wave start point and the T-wave end point; a QTc correction unit (455) that calculates QTc by correcting the QT interval that varies according to the heart rate; and a QT extension judgment unit (456) that determines whether the QT is extended and the risk grade by comparing the calculated QT interval and QTc with a reference value. Secretion closed-loop control device. Claim 9 In claim 6, the HRV calculation control unit (460) is characterized by being composed of an R-wave detection unit (461) that detects an R-wave from an ECG signal and generates a reference point for calculating an RR interval, an RR interval calculation unit (462) that calculates a time interval (RR interval) between two consecutive R-waves, a time domain HRV calculation unit (463) that calculates a time domain HRV index using an RR interval array, a frequency domain HRV calculation unit (464) that converts RR interval fluctuations into a frequency spectrum to analyze the autonomic nervous system balance state, a non-linear HRV calculation unit (465) that analyzes the complexity and unpredictability of heart rate variability, and an autonomic nervous system balance judgment unit (466) that performs the role of determining the autonomic nervous system state by integrating the results of the time domain HRV, frequency domain HRV, and non-linear HRV. Claim 10 In claim 6, the pulse waveform analysis control unit (470) comprises a PPG waveform detection unit (471) that detects a valid pulse waveform from a raw PPG signal collected from a PPG photoplethysmography sensor unit, a peak detection unit (472) that detects a systolic peak, a diastolic peak, or an auxiliary peak of the pulse in the PPG waveform, a pulse amplitude calculation unit (473) that calculates the pulse amplitude by calculating the peak height relative to a baseline in the PPG waveform, a pulse width calculation unit (474) that calculates the temporal width of the PPG pulse wave to analyze changes in vascular elasticity and pulse wave shape, a rise time calculation unit (475) that calculates the time from the starting point of the PPG pulse wave to the systolic peak, a peripheral blood flow index calculation unit (476) that calculates a peripheral blood flow index by synthesizing the pulse amplitude, pulse width, rise time, waveform stability, etc., and the collected PPG waveform can be used to estimate blood calcium status. A non-invasive multi-biometric signal-based field edge AI body-implantable PTH secretion closed-loop control device characterized by being composed of a pulse wave quality judgment unit (477) that performs the role of judging whether it is reliable to a certain degree. Claim 11 In claim 6, the neural excitability analysis control unit (480) is characterized by being composed of an EMG burst analysis unit (481) that detects and analyzes abnormal muscle activation signals in the form of a burst from an electromyography (EMG) signal, a tremor frequency analysis unit (482) that converts finger tremor data into the frequency domain and analyzes abnormal tremor components, a micro-spasm detection unit (483) that detects micro-spasms invisible to the naked eye using EMG burst and tremor data, a Bio-Z change analysis unit (484) that evaluates changes in tissue conductivity and extracellular fluid by analyzing changes in bioimpedance, a skin temperature correction unit (485) that corrects the EMG, tremor, and Bio-Z analysis results using changes in skin temperature, and a neuromuscular response judgment unit (486) that determines the neuromuscular excitability state by integrating the EMG burst, tremor, micro-spasm, Bio-Z, and skin temperature correction results. Blood ionized calcium estimation formula body-implantable PTH secretion closed-loop control device. Claim 12 In claim 6, the blood calcium status estimation control unit (490a) is characterized by being composed of an AI estimation model algorithm engine unit (490a-1) that receives a multi-biometric signal-based index and calculates the blood ionized calcium status and estimated value, a personal reference value storage unit (490a-2) that stores normal biometric signal patterns and personal reference calcium status for each user, a calcium status classification unit (490a-3) that classifies the AI ​​estimation results into clinically understandable status grades, an estimation value correction unit (490a-4) that corrects the AI ​​estimation value by reflecting movement, skin temperature, sensor quality, personal reference value, and recent test value, a trend analysis control unit (490a-5) that controls the analysis of the direction and speed of calcium change over time as well as the current calcium status, and a reliability calculation control unit (490a-6) that controls the calculation of the reliability of the blood calcium estimation result, thereby forming a non-invasive multi-biometric signal-based field edge-type AI blood ionized calcium estimation formula body-implantable PTH secretion closed loop. Control unit. Claim 13 In paragraph 2, the calcium balance smart app module (500) is characterized by being composed of: a calcium balance type activity control unit (510) that controls the activation and management of the main screen, calcium status screen, PTH injection status screen, notification screen, and history inquiry screen visible when a user runs the app; a calcium balance type service control unit (520) that continuously communicates with the field edge type AI module and monitors the calcium status and risk level even when the app is in a background state; a calcium balance type broadcast receiver control unit (530) that detects system events and external events of the smart device and automatically controls the operation of the app; a calcium balance type content provider control unit (540) that performs the role of safely storing and providing internal app data and external linked data; and a calcium balance type UI smart control unit (550) that controls the intuitive display of the estimated blood calcium status, risk level, PTH injection status, and biosignal status on the smart device screen, and controls color, vibration, voice, and popup notifications according to the risk level, thereby forming a non-invasive multi-biosignal based Field edge-type AI blood ionized calcium estimation formula body-implantable PTH secretion closed-loop control device. Claim 14 In claim 13, the calcium balance type UI smart control unit (550) comprises: a calcium status display control unit (551) that controls the display of the blood ionized calcium (Ca2+) estimation result in a form that is easy for the user and guardian to understand; a risk color display control unit (552) that controls the display of the low calcium risk level intuitively in color; a PTH injection status display control unit (553) that controls the display of the operation status of the photo-stimulation response type body-implantable PTH secreting object; a biosignal status display control unit (554) that controls the display of the biosignal status used for blood calcium estimation; a real-time graph display control unit (555) that controls the display of the calcium status and biosignal changes over time as a graph; a guardian sharing screen control unit (556) that controls the display of the patient's status on the guardian app screen; an emergency notification popup control unit (557) that controls the display of a popup when an emergency occurs; and a vibration, sound, and voice control unit that controls the output of vibration, warning sound, and voice guidance according to the risk level. A non-invasive multi-biometric signal-based field edge AI body-implantable PTH secretion closed-loop control device characterized by being composed of a guide control unit (558). Claim 15 A step (S10) of sensing first non-invasive biosignal data regarding PPG (Photoplethysmography), blood oxygen saturation (SpO2), bioimpedance, skin temperature, tremor, and electromyography (EMG) by wearing a smart ring-type wearable sensor unit on a finger, and transmitting the sensed first non-invasive biosignal data to a wrist-mounted smartwatch; a step (S20) of sensing second non-invasive biosignal data regarding ECG (Electrocardiography) signals, heart rate, and heart rate variability flowing from the surface of the chest skin by attaching an external wearable patch unit to the surface of the chest skin, and transmitting the sensed second non-invasive biosignal data to a wrist-mounted smartwatch; and a step (S20) of measuring 3-axis acceleration data regarding the posture of the user moving by wearing a wrist-mounted smartwatch on the user's wrist, and then the data transmitted from the smart ring-type wearable sensor unit A step (S30) of collecting the first non-invasive biosignal data and the second non-invasive biosignal data transmitted from an externally worn wearable patch and transmitting them to a field edge AI module via a wireless communication network; a step (S40) of estimating the blood calcium status after calculating the QT (Q wave to T wave) interval, HRV (Heart Rate Variability), Pulse waveform, and neural excitability index using the first non-invasive biosignal data, second non-invasive biosignal data, and 3-axis acceleration data transmitted from a wrist-mounted smartwatch via the field edge AI module; a step (S50) of transmitting the estimated blood calcium status to a calcium balance smart application module via the field edge AI module; and a step of displaying the estimated blood calcium status, risk level, PTH injection status, and biosignal status on a smart device screen via the calcium balance smart application module, and in case of a low calcium risk, A step (S60) for performing an emergency notification role, and through a field edge-type AI module,A closed-loop control method for the secretion of implantable PTH based on blood ionized calcium estimation formula of a non-invasive multi-biosignal-based field edge AI, characterized by comprising the steps of: outputting a photostimulation response signal to secrete implantable PTH toward a wrist-attached smartwatch located on a wrist muscle using an estimated blood calcium status (S70); and secreting implantable PTH (Parathyroid Hormone) according to the photostimulation signal of the wrist-attached smartwatch through a photostimulation response type implantable PTH secretion object (S80). Claim 16 In claim 15, the above step (S20) comprises: a step (S21) of supporting the ECG electrode and electronic circuit so as not to detach from the patch position by placing them on the lower surface of the entire patch that is in direct contact with the chest skin surface through a patch-type adhesive substrate, thereby adhering the ECG electrode and electronic circuit to the skin; a step (S22) of detecting a microelectric signal generated from the heart and transmitted to the chest skin surface through an ECG electrode; a step (S23) of amplifying the microelectric signal measured by the ECG electrode to an analyzable size through an ECG amplifier; a step (S24) of filtering power noise, electromyography noise, motion noise, and baseline fluctuations from the ECG signal amplified by the ECG amplifier through an ECG filter; a step (S25) of calculating the heart rate by detecting the R wave of the ECG signal through a heart rate calculation unit; a step (S26) of calculating heart rate variability data by analyzing changes in the RR interval through an HRV calculation unit; and a step of, for a certain period of time, through a raw ECG storage unit for QT·QTc extraction A non-invasive multi-biometric signal-based field edge AI body-implantable PTH secretion closed-loop control method characterized by comprising the step (S27) of storing raw ECG or high-quality filtered ECG data of a segment, and the step (S28) of wirelessly transmitting second non-invasive biosignal data regarding ECG (Electrocardiography) signals, heart rate, and heart rate variability to a wrist-worn smartwatch through a wireless communication unit for a patch. Claim 17 In claim 15, the above step (S40) comprises: a step of receiving an integrated data packet from a wrist-mounted smartwatch through a data receiving unit (S41); a step of correcting noise, abnormal values, and missing values ​​of the biosignal received through the data receiving unit in a preprocessing unit (S42); a step of removing PPG, ECG, and EMG signal distortions caused by movement using 3-axis acceleration data received through the data receiving unit in a motion artifact removal unit (S43); a step of controlling the calculation of integrated feature values ​​necessary for estimating blood calcium status from the first non-invasive biosignal data, the second non-invasive biosignal data, and the 3-axis acceleration data through a feature value calculation control unit (S44); a step of controlling the calculation to calculate the QT interval from the Q-wave start point to the T-wave end point and QTc in the ECG signal through a QT interval calculation control unit (S45); and a step of analyzing the RR interval change through an HRV calculation control unit to determine the heart rate variability A non-invasive multi-biosignal-based field edge-type AI blood ionized calcium estimation formula body-implantable PTH secretion closed-loop control method characterized by comprising: a step (S46) of controlling to calculate an index; a step (S47) of calculating peripheral blood flow, vascular tone, and pulse wave changes by analyzing a PPG-based pulse waveform through a pulse waveform analysis control unit; a step (S48) of analyzing neuromuscular excitability changes using EMG, Tremor, bioimpedance, and skin temperature data through a neural excitability analysis control unit; a step (S49) of quantifying and indexing QT, HRV, pulse waveform, and neural excitability results through an index calculation unit; and a step (S49a) of controlling to estimate the blood ionized calcium state based on AI using the data calculated from the index calculation unit through a blood calcium state estimation control unit.

Images