A MEDICAL SYSTEM

Abstract

This invention relates to a system (1) that enables the patient and / or user with painful areas in their body to wear at least one medical fabric which sends electrical stimuli to the painful area, thereby relieving the patient's painful area.

Description

1 TARIFF A MEDICAL SYSTEM Technical Area This invention is for patients and / or users who have painful areas in their bodies. 5 the subject is at least one medical fabric that sends electrical impulses to the painful area. a system that allows the patient to wear something that relieves pain in the affected area It is related to. Previous Technique Chronic pain management constitutes one of the most significant challenges in modern medicine. Millions of people worldwide suffer from musculoskeletal disorders. Managing neuropathic pain, fibromyalgia, and similar chronic pain conditions 15 This problem is being addressed with transcutaneous electrical nerve stimulation. (TENS) method is a non-invasive and drug-free treatment option. It is widely used. However, current TENS (Transcutaneous Electrical) Nerve stimulation (transcutaneous electrical nerve stimulation) devices are important. It has disadvantages. Traditional TENS devices require 20 separate steps for the user's body. It is attached as an electronic unit; the electrodes are attached to the skin with adhesive gel tapes. is secured and the device requires regular battery replacement or charging. This requires... This situation imposes serious limitations in terms of user comfort. This leads to restrictions on freedom of movement in daily activities and long-term consequences. It causes skin irritation at the electrode site during use. 25 The most critical disadvantage of current TENS devices is their energy dependence. Users, They have to regularly charge their devices or replace the batteries, which 2 This disrupts the continuity of treatment, especially for patients with chronic pain. Given the need for ongoing therapy, after using it for a few hours a day The need for charging significantly reduces treatment effectiveness. Furthermore, traditional methods... The devices' bulky and aesthetically striking design allows users to socialize... This causes them to hesitate to use the device in certain environments. Adhesive 5 The marks left on the skin by the electrodes, the need for frequent replacement, and the discomfort caused by using the gel Discomfort is also among the significant user complaints. In recent years, significant progress has been made in the field of wearable health technologies and smart textiles. Developments are taking place. Piezoelectric energy harvesting converts mechanical energy into electricity. 10 It is a technology that converts energy into usable energy, especially for low-power devices. It offers a suitable energy source. However, this technology is problematic for medical devices. especially its integration with TENS systems is not yet fully complete. This has not been realized. Existing piezoelectric textile applications are generally sensor applications. 15 applications or areas requiring very low power, such as LED lighting It remains. The voltage levels required for TENS application are 20-50 V. Generating sufficient energy to reach and provide sustained stimulation is a technical requirement. This presents us with a challenge. In the current state of the art, 20 long-life TENS electrodes and transmission lines are included. the provision of a user-friendly device, the TENS device in question having a traditional battery and It is powered by a rechargeable battery, and the patient inputs a pain intensity scale. It can operate a TENS device, input data, time and storing, analyzing, and using density data for pain treatment. spinal column 25, which allows for presentation to the clinician to evaluate its effectiveness. molded plastic for transcutaneous electrical nerve stimulation of the region enabling the use of a device containing electrodes integrated into a component Various solutions are being encountered. 3 However, in the known state of the technique, energy harvesting and TENS therapy are synergistic. the integration involves converting the user's kinetic energy into electricity. by transforming it, it can both produce its own energy and use this energy for therapeutic purposes. four-layer textile architecture that enables its use, antibacterial inner layer, piezoelectric energy generation layer, flexible electronic layer and protective outer layer 5 Providing an integrated solution with the layer, AC / DC converter, DC-DC boost Low-voltage piezoelectricity thanks to converter and supercapacitor integration. Artificial power supply that increases the output to the 20-50V level required for TENS. Personalized treatment protocols using intelligent algorithms, addressing the user's pain By learning their patterns and activity levels, optimal stimulation is achieved. 10 a solution that allows the parameters to be adjusted automatically It has not been encountered. In the Chinese patent document numbered CN113730806, which is included in the known state of the art... An invention related to a wearable transcutaneous electrical nerve stimulation sock 15 It is explained that the sock in question consists of an elastic sock, and inside this sock... A power generation device is being installed, and the power generation device uses nanometer friction. Power generation is achieved using fabrics and non-conductive materials. Electricity a capacitor that generates energy, a boost element, and a pulse oscillation element It is placed inside the elastic sock. Booster element and shock absorption 20 The element is electrically connected to the power generation device; also the elastic sock It also contains numerous electrical stimulation massage particles, and Electrical stimulation massage particles are connected to a power generation device. Transcutaneous electrical nerve stimulation stocking, simple design structure, high It has advantages such as functionality, ease of use and portability, nanometer 25 By applying the principle of friction energy generation holistically, the traditional approach... The problem of not having a constant and easily portable energy source in treatments. being resolved, electrical nerve stimulation through the skin to the feet of the human body By applying this method, an analgesic effect is achieved, and patients' daily pains are relieved. A new treatment method is offered to alleviate the symptoms. Clinically, many patients 30 4 They suffer from Achilles tendinitis and degenerative diseases of the foot bones, and Their daily walks inevitably cause them pain and psychological stress. This invention offers a new treatment method to alleviate this pain. The device, mainly from nano-triboelectric materials and non-conductive fabrics a wearable transcutaneous electrical nerve device containing an elastic stocking made of 5 It is a stimulation sock. The sock in question is made of nano-triboelectric fabric and conductive material. It houses a power generation device that contains non-fabric. Inside the sock there is also a It includes a capacitor, a voltage booster, and a pulse oscillator. The capacitor is located between nano-triboelectric fabric and non-conductive fabric. It collects electrical energy generated by friction and feeds it into a power generation device. It is connected. The voltage booster element converts the alternating voltage into energy for the capacitor. It converts it in a way that will provide sufficient power for storage. Pulse oscillation This element operates two electrical components. However, there are 15 differences between the aforementioned document numbered CN113730806 and our invention, as follows: These are some of the differences. • The energy harvesting mechanism is fundamentally different: CN113730806 operates on the triboelectric principle — between two different fabrics (nano- from the friction (contact 20) between triboelectric fabric + non-conductive fabric (electrification + electrostatic induction) produces electricity. Our invention, however... It operates on the piezoelectric principle — in the crystal structure of the PVDF / PZT material. Electricity is generated when dipoles reorient themselves under mechanical strain. These are physically different energy conversion mechanisms (friction, etc.). deformation), requires different material classes and presents different engineering challenges 25 She gives birth. • Differences in power management architecture: CN113730806 defines a single "boost element", No specific voltage value, topology, or efficiency is given. In our invention, however, 30 Two-stage cascade boost to achieve the 20–50V range required by TENS. converter (LTC3108 + flyback, 78–85% overall efficiency), piezoelectric source specific impedance matching-based MPPT algorithm (Z_opt = 1 / (2πf·Cp)) and The mathematical model of supercapacitor charging / discharging (t_charge = C·ΔV / I) is concretely presented. It is defined. This is much more than a general statement like "an upgrade element". 5 It is a specific and technical solution. • Differences in terms of sensors, feedback, and personalization: CN113730806 has no sensors (accelerometer, gyroscope, EMG, skin impedance, etc.) or no artificial intelligence is present. Stimulation parameters are fixed ("2Hz, 100Hz, 10 "2 / 100Hz waves") and do not adapt to the user or activity. Our invention uses a 15-dimensional feature vector, Random Forest classification + Reinforcement learning-based closed-loop optimization (with reward function) (pain / energy / comfort / habituation balance) CN113730806 has no equivalent. It is a layer. 15 • The application area and purpose are different: CN113730806 is available only in foot / sock form, targeting acupuncture points (Yongquan, It provides stimulation with "massage particles" (for taichong, etc.) — acupuncture It is based on the theory. Our discovery is based on the general TENS theory (gate control + endorphin 20 It offers a modular architecture based on (swinging) motion, with multi-zone (knee, waist, shoulder, elbow) support. Brief Description of the Invention The aim of this invention is to combine piezoelectric energy harvesting technology with TENS therapy. By combining these elements, a fully self-charging, textile-integrated medical device is created. mechanical processes that enable the presentation of information generated during the user's daily activities energy (walking, joint movements, body movements) piezoelectric materials converting it into electrical energy and using this energy for TENS therapy 30 6 enabling its use without the need for external charging or battery replacement. enabling continuous therapy and daily life through textile integration. It provides a comfortable and aesthetically pleasing solution that can be worn under clothing. It is implementing a medical system. Detailed Description of the Invention This invention was created in the form of a "Medical System" to achieve its purpose. shown in this figure; 10 Figure 1: Detailed schematic of the textile material in a medical system that is the subject of the invention. It is the appearance. Figure 1 shows a schematic view of a medical system that is the subject of the invention. The parts shown in the figure are individually numbered, and the corresponding numbers correspond to these numbers. It is given below. 1. System 2. Textile materials 20 21. Inner layer 22. piezoelectric layer 221. energy harvesting module 23rd electronic layer 231. Power management module 25 232. Energy Budget Analysis Module 233. TENS generator circuit Sensor 234. 235. Microcontroller 236. Wireless communication module 30 7 24th outer layer 3. Electronic devices 4. Application 5. External server the patient and / or user who has painful areas in their body, the painful area in question wear at least a medical-grade fabric that sends electrical signals to the area, This invention is a medical device that provides relief to the patient's painful area. system (1); 10 - inner layer (21), piezoelectric layer (22), electronic layer (23) and outer layer (24) formed by bringing together four main layers, the inner layer (21) the inner layer which makes contact with the user's and / or patient's skin (21) antibacterial and moisture-absorbing properties from the base textile layer that comes into contact with the skin. silver coated conductive threads in the inner layer (21) that make the product a finished product 15 Because it is made of nylon (AgNy) filaments, the TENS current is distributed homogeneously. PVDF nanofibers that enable distribution on the patient's skin produced by the “electrospinning” method and laminated to the inner layer (21) Energy related to TENS current is generated thanks to the piezoelectric layer (22). providing, the piezoelectric layer (22) has a low piezoelectric coefficient (d33 ≈ -33 20 Despite its superior flexibility (Young's modulus ≈ 2-4 GPa), biocompatibility, and (pC / N) Wearable medical devices in terms of washability and suitability for textile integration. It has the feature of being made of PVDF material, which is preferred in its applications. The electronic layer (23) is on a 50 μm thick flexible polyimide PCB. piezoelectric layer 25 to accommodate the mounted electronic components (22) to operate at least one energy harvesting module (221) on it to provide at least one power management module (231) of the electronic layer (23), energy budget analysis module (232) and intermittent stimulation protocol, TENS generator circuit (233), sensor (234), microcontroller (235) wireless communication to ensure that the module (236) is included, the outer layer (24) and the inner layer (21), 30 8 Protecting the piezoelectric layer (22) from external factors, electronic layer (23) to provide at least an energy harvest thanks to the piezoelectric layer inside (22) operation of the mechanism, piezoelectric with energy harvesting mechanism mechanical dipole moments in the PVDF material in the layer (22) its reorientation under deformation and opposite electric charge on the surfaces 5 to cause it to occur when the patient bends or straightens their knee. each incoming bending cycle produces an AC voltage, and this voltage is then passed through a Schottky diode. rectifying it with its bridge and converting it to DC, thanks to its multi-layered structure. Thanks to the stack configuration, a voltage typically between 4 V and 12 V is provided. To ensure proper operation of the TENS application, 20 to 50 V 10 two-stage cascade boost converter topology is used when reaching these values to enable its use, Maximum Power on the piezoelectric layer (22) Compared to constant impedance loading, thanks to the Point Tracking (MPPT) algorithm. Ensuring that 30-40% more energy is harvested, the energy TENS 15 is stored in a supercapacitor and released when the threshold value (e.g., 30 V) is reached. to enable the module to be activated automatically, electronic layer (23) with its artificial intelligence module, it analyzes the user's activity pattern. learning and automatically optimizing the most effective TENS protocol for the user. to ensure that the artificial intelligence algorithm in question makes two-stage decision-making the mechanism should include, in the first stage, the Random Forest classifier (150 trees, 20 Using the maximum depth 8, minimum leaf node sample 5) related estimation of user / patient activity classification and pain level the second stage involves maximizing therapeutic efficacy under energy constraints. performing multi-objective optimization to provide limited reinforcement learning within the optimization in question 25 To enable the implementation of the (constrained reinforcement learning) approach. All real-time with the microcontroller (235) on the electronic layer (23) sensor (234) data and artificial intelligence model inferences are made entirely electronically. to enable therapeutic decisions to be processed on the cloud layer (23). To enable local delivery without requiring a connection, raw sensor data, 30 9 Information in the form of personal health data and model parameters shall under no circumstances be used to prevent data from being transferred outside the device, thanks to its layered data architecture, to optimize the privacy-functionality balance, at the electronic layer (23) Bluetooth Low Energy 5.0 (BLE) chip in the wireless communication module (236) This allows authorized data obtained from the user to be transferred to external environments, 5 Two-phase (biphasic) with TENS Generator Circuit (233) in the electronic layer (23) generating pulses (vibrations), low-frequency TENS (Transcutaneous Electrical Nerve Imaging) By stimulation (2-10 Hz), endorphin and enkephalin release is increased in the user, resulting in a longer-lasting effect. to achieve a sustained analgesic effect, high-frequency TENS (50-150 Hz) through gate control theory, the acute pain signals present in the relevant user are 10 at least one textile material structured to ensure that it is blocked (2), - the user and / or patient using the textile material (2) with other people at least one electronic device configured to enable communication (3), - wireless 15 located on textile material (2) and operated on electronic device (3). To connect with the communication module (236), the user's interface on it Customize TENS parameters, the energy status of textile material (2) user monitoring, user pain diary, and treatment history. to enable the display of all real-time transmitted from textile material (2) 20 stimulation control data, sensor (234) data, artificial intelligence processing to enable processing within the electronic device (3), after obtaining permission from the user Enabling the optional cloud backup feature, textile material (2) Securely transmitting the user's treatment data to at least one external server. (5) at least one application (4) configured to enable sending and - only 25 that connect with application (4) and are transmitted from application (4) anonymized treatment summary data (VAS scores, stimulation durations, energy) (statistics) to be stored in HIPAA / GDPR compliant and encrypted format. It includes at least one external server (5) configured. The textile material (2) in the system subject to the invention (1) has a four-layer structure (21) silver in the inner layer which is in contact with the user’s skin This allows the use of bamboo or merino wool coated with nanoparticles. Inside layer (21) also contains conductive textile lines (conductive thread) and acts as an electrode. The inner layer (21) has silver-coated nylon 5 TENS using (AgNy) filaments while maintaining its surface resistance at a level of 1-5 Ω / cm. The current is distributed homogeneously across the user's skin. The pieozoelectric layer (22) in the system subject to the invention (1) is PVDF (polyvinylidene (fluoride) nanofibers with a diameter of 200-500 nm are produced by electrospinning technique and 10 by laminating lead zirconate titanate (PZT) nanoparticles onto textiles Composite panel in flexible silicone matrix (30-40% volume ratio) It was created by bringing in the piezoelectric layer (22) low piezoelectric Despite its coefficient (d33 ≈ -33 pC / N), it has high elasticity and low density (1.78 (g / cm³), preferably 15 due to its wearable comfort compatibility and non-toxic nature. It is made of PVDF material. Piezoelectric layer (22) to compensate for the disadvantage of flexibility multilayer Enabling the implementation of a piezoelectric stack configuration: 4-6 panels per panel. It is created by laminating PVDF layers on top of each other in series, this 20 Thus, compared to a single layer, the voltage output is directly proportional to the number of layers. It ensures its increase. Piezoelectric layer (22) panels are strategically placed in areas with high mechanical stress. areas (around the knee joint, ankle, hip joint, mid-back, shoulder region and 25 (elbow circumference) 5×10 cm in size, placed on the user, preferably A total of 6 panels are used, and each panel has different activity levels. In the achieved sitting position (minimal movement), 0.05-0.2 mW per panel, 0.8-1.5 mW per panel during normal walking (4-5 km / h), during fast walking (6-7 km / h) 1.5-2.5 mW per panel, 2.0-3.5 mW per panel for stair climbing activity 30 11 to enable power generation, 6 panels under normal walking conditions for the user To ensure that the total power generation averages 9 mW (6 × 1.5 mW). It is structured accordingly. Piezoelectric layer (22) multilayer piezoelectric stack 5 in each panel by applying the configuration (series bonding of 3-5 layers of PZT-silicon composite). It enables a 2-3 fold increase in output voltage per single panel, when connected in series. In a stack structure, the voltage generated by each layer is summed (V_total = n × V_layer, (n: number of layers) provides an increase in the input voltage to the boost converter. The energy harvesting module (221) is located in the piezoelectric layer (22), PVDF the β-phase in the crystal structure of the material under mechanical deformation It enables the material to exhibit a piezoelectric effect, thereby transferring it to its own material. the redirection of dipole moments of the applied stress (strain) and This causes an accumulation of electrical charge on surfaces. 15 The energy harvesting module (221) is placed on the textile material (2) by the user. When the array is bent, the panel has compressive stress (εc = -0.5% to -2%). (between), the knee being subjected to tensile stretching (εt = between +0.5% and +2%) when opened. It provides AC voltage for each bending cycle (single panel: typical 2-8 V, stack 20 In this configuration, it enables the production of 6-24 V; current: 0.1-0.5 mA. Schottky diode bridge (low forward) on energy harvesting module (221) It enables the rectification of AC to DC with a voltage drop of <0.3V. It enables the calculation of the generated power using the formula P = V × I × η (η: conversion 25 Efficiency ~60-75%), 1.5 mW / panel at the user's average walking activity. This enables the achievement of a certain value, with a total of 9 mW using six panels. It ensures continuous power generation. 12 Energy harvesting module (221) to the required 20-50 V level for TENS application. It utilizes a two-stage cascade boost converter topology during its implementation: The first stage piezoelectric output (4-12 V) is increased to an intermediate voltage level (15-20 V) 5 the first stage is to raise the intermediate voltage to the target TENS voltage (20-50 V, (adjustable) extracts the cascade structure in question, single-stage high-rate operating with higher efficiency (85-92%) compared to the previous model It provides. In one arrangement of the invention, the energy harvesting module (221) flyback Using the converter topology, the transformer winding ratio (N2 / N1) and voltage 10 The boost rate can be controlled directly. Energy harvesting module (221) internal impedance of piezoelectric sources capacitive because of its character, it differs from traditional photovoltaic MPPT algorithms (Perturb & Unlike Observe, impedance matching-based MPPT (Maximum Power Point 15) Enabling the use of the Tracking approach, in the piezoelectric layer (22) the equivalent capacitance (Cp) of the panel and the mechanical excitation frequency (f_mech) real-time measurement, optimal load impedance Z_opt = 1 / (2π × with the formula f_mech × Cp) (f: mechanical vibration frequency, C_p: piezoelectric Calculation of capacitance on the microcontroller (235) and DC-DC 20 the duty cycle of the converter must satisfy this impedance matching condition It is configured to allow for adaptive adjustment. Energy harvesting module (221) in MPPT algorithm, perturb-and-observe (P&O) 25 to provide maximum power by adjusting the load impedance every 100 ms. to enable the transfer of fixed load with this approach It is possible to obtain 15-25% more energy harvesting compared to its impedance. It is structured to ensure that it is carried out. 13 Electronic layer (23) AC / DC converter on flexible PCB, cascade boost converter topology, supercapacitor (100-500 mF capacity, preferably 470 mF), microcontroller (235) (ultra low power ARM Cortex-M0+ or ​​RISC-V based, (Power consumption <1 mW in active mode, <5 μW in sleep mode) and TENS generator It contains the circuit (233). There is at least one 5 on the electronic layer (23). power management module (231), energy budget analysis module (232), sensor (234) and also contains the wireless communication module (236). With the energy budget analysis module (232) located on the electronic layer (23) Under normal walking conditions, the total power generation P_generation = 9 mW. 10 Conventional continuous TENS therapy requires 0.4-4 W of power, which affects the user. Information regarding the application of an intermittent stimulation protocol. After acceptance, the energy consumption of a single TENS pulse is E_pulse = V Calculation using the formula × I × t_pulse, where V = 40 V (typical operating voltage), I = 40 mA (average therapeutic current), t_pulse = 200 μs (pulse width) when 15 The E_pulse should be calculated as 40 × 0.04 × 200 × 10⁻⁶ = 0.32 mJ at 80 Hz. The stimulation frequency is 4800 pulses per minute, and the instantaneous average power is... The consumption is calculated as P_TENS = 0.32×10⁻³ × 80 = 25.6 mW. It is structured to provide this. With the energy budget analysis module (232) located on the electronic layer (23) The calculated P_TENS value must be above 9 mW of continuous production capacity. In this case, ensuring the duty cycle principle is applied, Duty cycle 20-25% when set as (e.g., 30 seconds stimulation / 90 seconds charging cycle or) (60 seconds stimulation / 180 second charging cycle), average power consumption is 25 P_average = P_TENS × duty_cycle = 32 mW × 0.25 = 8 mW should be reduced to this. Ensuring this value remains below 9 mW of production capacity, using artificial intelligence. The algorithm determines the current energy status and user activity level. by evaluating and ensuring the dynamic optimization of the duty cycle. It is being structured. 30 14 With the power management module (231) located on the electronic layer (23) The voltage from the piezoelectric layer (22) must reach the required level for TENS application. upgraded with a two-stage cascade boost converter topology To enable its implementation, in the first stage, the LTC3108 ultra-low power DC-DC 5 boost converter piezoelectric output (6-24 V from stack configuration or raising the voltage from a single panel (2-8 V) to an intermediate voltage level (12-15 V) It is structured to provide this. In the second stage, with the power management module (231) on the electronic layer (23), 10 thanks to the flyback converter topology (e.g. LT3573 or custom-designed flyback) (circuit) intermediate voltage to the high voltage required by TENS (20-50 V, adjustable) This is achieved by ensuring the flyback converter provides galvanic isolation, To provide an additional layer of protection in terms of user security. It is structured as follows. The power management module on the electronic layer (23) 15 (231) with cascade topology total conversion efficiency in the range of 78-85% the energy produced is in a supercapacitor with a capacity of 100-500 mF (ESR <50 mΩ, to enable the storage of Maxwell Technologies BCAP series or equivalent. It is being structured. power management module (231) on the electronic layer (23) and supercapacitor The charging time should be calculated using the formula “t_charge = C × ΔV / I_charge”, where C = 200 mF, ΔV = 20 V (charge from 10V to 30V) and I_charge = 9 mW / 20 V = 0.45 When mA is taken, the full charge time is approximately t_charge = 0.2 × 20 / 0.00045 ≈ 8889 Calculated in seconds (approximately 148 minutes, or 2.5 hours of walking activity) 25 It is structured to provide this. The voltage threshold value with the microcontroller (235) on the electronic layer (23) (e.g., 30V) Continuous monitoring, and activation of the TENS generator circuit when the threshold is exceeded. (233) activation, power consumption optimized in deep sleep mode <5 μA, 30 to ensure current draw of <2 mA in active mode It is being structured. In the energy budget analysis module (232) on the electronic layer (23) Thanks to the intermittent stimulation protocol implemented, 5 of the intermittent TENS applications (30-60 seconds of active stimulation / 2-3 minutes of rest) compared to continuous stimulation. It prevents neural habituation and preserves therapeutic efficacy, and even some To increase the therapeutic efficacy in cases of chronic pain. It is structured accordingly. In the energy budget analysis module (232) on the electronic layer (23) thanks to the intermittent stimulation protocol that was run, using textile material (2) the user's statement during periods of intense physical activity (climbing stairs, running) The subject is textile material (2) enabling the power generation to increase to 18-24 mW. allow for longer stimulation times or higher duty cycle values ​​15 It is structured to ensure recognition. TENS generator circuit (233) on electronic layer (23) and H-bridge Generating biphasic (two-phase) pulses using MOSFET configuration. to provide 20 range of width (50-250 μs), frequency (2-150 Hz), intensity (10-80 mA). The pulse parameters are controlled by PWM (Pulse Width Modulation). It is structured to enable this. On the electronic layer (23) Current limiting circuit (IEC 60601-2-10) for TENS generator circuit (233) (<80 mA according to standard), short circuit protection and overvoltage protection It is structured to enable integration. 25 3-axis MEMS accelerometer with sensors (234) on the electronic layer (23) thanks to (for example, ADXL345) the user's step count, movement intensity, and joint activity can be monitored. gyroscope allows monitoring of physical activity in the form of bending frequency. thanks to the detailed 30 movement patterns of the user using textile material (2). 16 It is structured to enable analysis. Electronic layer (23) electrode-skin contact with capacitive pressure sensors (234) on it quality monitoring with a temperature sensor (e.g. TMP116, ±0.2°C) (2) thermal safety of textile material by monitoring skin surface temperature) sensitivity) to ensure that the data in question is checked using 100 Hz sampling 5 to ensure that it is transmitted to the microcontroller (235) with its frequency It is being structured. The microcontroller (235) works on the electronic layer (23) of the textile material (2) is structured to function as the brain. Microcontroller (235) 10 to the machine learning algorithm on it, TENS parameter optimization specially designed multi-objective optimization It is structured to implement this approach. 15 in the machine learning algorithm on the microcontroller (235) [accelerometer_x, accelerometer_y, accelerometer_z, gyro_roll, gyro_pitch, gyro_yaw, step_frequency, motion_intensity_RMS, joint_flexion_angle, skin_impedance, supercapacitor_voltage, last_VAS_score, time_of_day, a 15-dimensional form [last_stimulation_time, cumulative_energy_state] It is configured to run the feature vector. Microcontroller (235) 20 The lightweight version of the Random Forest algorithm consists of 12 decision trees. to ensure that each tree has a maximum depth of 6 levels its limitation, with this restriction, within 4KB RAM and 32KB flash memory. to ensure the model size is optimized in a way that allows it to work It is being structured. 25 Reward on the Reinforcement learning module on the microcontroller (235) function R = -VAS_score × α + energy_efficiency × β + comfort_index × γ To ensure it is defined in this form; where α = 0.6 (pain reduction weight), β = 0.25 (energy efficiency weighting), γ = 0.15 (comfort index weighting), VAS_score 30 17 Pain level reported by the user (0-10), energy_efficiency = (therapeutic_effect / (energy expended) normalized value and comfort_index electrode-skin contact To ensure the use of a combination of quality and stimulation comfort level. It is structured accordingly. The reward function run on the microcontroller (235) of the energy constraint, E_consumed(t) ≤ E_available(t) × safety_margin (safety_margin = 0.7) It should be defined as follows: the reward function is never a supercapacitor. not to consume more than 70% of its energy in a single stimulation session to ensure that the action space of the reward function in question is a = 10 as [frequency, pulse_width, intensity, duty_cycle, electrode_pair_selection] to ensure that each parameter is defined and selected from a discrete set of values. It is structured in such a way. The machine is run on the microcontroller (235). the fixed model of traditional TinyML solutions in the learning approach Unlike inference-only learning, on-device incremental learning (incremental 15 (learning) implementation, user-specific model parameters in each session It is configured to ensure that it is eventually updated. Microcontroller (235) model size is limited to 48 KB RAM and 256 KB flash memory The ability to run in real-time on an ARM Cortex-M0+ processor, inference 20 to ensure that the duration of a single decision cycle is less than 10 ms. It is being structured. Multi-objective in machine learning on microcontroller (235) the optimization approach that maximizes therapeutic efficacy under energy constraints energy-aware decision-making mechanism—the algorithm, currently 25 using the supercapacitor energy level directly as a decision variable by simultaneously optimizing stimulation parameters and duty cycle from existing TinyML solutions (TensorFlow Lite Micro, Edge Impulse) It is structured primarily to enable separation. 18 Multi-objective in machine learning on microcontroller (235) optimization approach adapted to piezoelectric energy harvesting dynamics adaptive sampling—both energy production and efficiency during periods of high activity While increasing the frequency of data collection, ultra of textile material (2) at low activity Switching to low power mode enables temporal pattern learning specific to TENS—algorithm 5 and the temporal correlations of the user's pain patterns (morning stiffness, learning about (such as increased pain after activity) and initiating proactive stimulation with points like these from existing TinyML solutions (TensorFlow Lite Micro, Edge Impulse is primarily designed to enable separation. Multi-objective in machine learning on microcontroller (235) the optimization approach uses 8-bit integers of the machine learning model in question. Because it is compressed using quantization, and the inference time is less than 5 ms. with points like these from existing TinyML solutions (TensorFlow Lite Micro, Edge Impulse is primarily designed to enable separation. 15 inside the wireless communication module (236) on the electronic layer (23) The use of a Bluetooth Low Energy 5.0 (BLE) chip in the form of nRF52832. by providing textile material (2) on the user’s electronic device (3) It is configured to enable communication with the application (4). 20 Protocol in the wireless communication module (236) on the electronic layer (23) The use of GATT (Generic Attribute Profile) and the user's TENS parameters such as energy status, activity data, and pain scores to ensure that it is transmitted to the application (4) securely (AES-128 encryption) It is being structured. 25 Electronic layer (23) automatically cuts off when current greater than 80 mA arrives. Overcurrent protection, short circuit protection, overvoltage protection (discharge at >60V), thermal protection (system shutdown at >45°C) and electrode-skin contact monitoring (contact 30 safety layers (stimulation stoppage when interrupted) 19 It is configured to enable operation. Electronic layer (23) supercapacitor overcharge protection (charge cut-off at >55V) and balancing circuit It is configured to run on (cell balancing). The external server (5) in the system (1) which is the subject of the invention must operate in compliance with HIPAA / GDPR. with the explicit consent of the user via the application (4) and only BLE anonymous statistical data (daily) transferred to the application (4) via the link (total stimulation time, mean pain score trend, energy production / consumption summary), optionally stored internally according to user preference. It is being structured. 10 Industrial application of the invention In the system that is the subject of the invention (1) pains in certain parts of the user's body The textile material (2) included in the invention applies to the user's painful area 15 (2) electronic and piezoelectronic components in textile material are being placed thanks to this, the textile material gains its own energy through the movement of the user. It produces (2) electronic components in textile materials, thanks to modules. TENS pulses are transmitted to the user's painful area, relieving the user's pain. There is weakening. User data produced by textile material (2) is secure 20 In this way, it is sent to the application (4) on the user's electronic device (3). Study of textile material (2) in TENS treatment with user application (4) It can change its parameters. If the user approves, the application (4) User information is stored in encrypted form on an external server (5). The advantages of the system in question (1) over existing technologies are as follows: (a) Energy independence - For users who take an average of 3000 steps a day, intermittent Stimulation protocol with 36-72 minutes of effective TENS over a 12-hour period. being able to receive therapy, this period is clinically significant for chronic pain management. Providing a therapeutic dose, 30 in more active users (8000+ steps per day) Supercapacitor energy excess and stimulation duty cycle automatically (b) Increased comfort and fit - so comfortable you'll forget about it thanks to textile integration. comfortable, machine washable (after removing the electronic module) (c) Cost effectiveness - No battery / charging costs after initial investment, 3-5 years. (d) Environmentally friendly - Does not generate single-use battery waste, 5 (e) Adaptive therapy - AI-assisted personalized treatment protocol, fixed Compared to conventional parameterized TENS devices, it offers 25-40% better results in the long term. Providing high analgesic efficacy, and in terms of safety, TENS current complies with IEC 60601. Limited to 80 mA according to standard 2-10, for short circuit and overcurrent protection. Protection circuits are integrated, and the supercapacitor voltage has an upper limit of 50 V. a hardware voltage clamping circuit that cannot be exceeded The addition of piezoelectric materials isolated from the skin via biocompatible encapsulation. to be done. The invention system (1) and piezoelectric energy harvesting technology for TENS therapy 15 Combined with other components, it creates a fully self-charging, textile-integrated medical device. the presentation of mechanical energy generated during the user's daily activities (walking, joint movements, body movements) piezoelectric materials and this energy is converted into electrical energy and used for TENS therapy. its use allows for continuous therapy without the need for external charging or battery replacement. 20 The realization of this possibility, thanks to textile integration, allows for the use of textiles under everyday clothing. This ensures a wearable, comfortable, and aesthetically pleasing solution. The system in question (1) is within the scope of the Personal Data Protection Law (KVKK). It operates. 25 The subject of the invention is the development of a wide variety of applications of a medical system (1). It is possible, and the invention is not limited to the examples described here, but is essentially As stated in the requests.

Claims

21 REQUESTS 1. The patient and / or user who has painful areas in their body... wear at least one medical-grade fabric that sends electrical impulses to the painful area, 5. This helps to relieve the patient's painful area. - inner layer (21), piezoelectric layer (22), electronic layer (23) and outer layer (24) formed by bringing together four main layers, the inner layer (21) the inner layer which makes contact with the user's and / or patient's skin (21) antibacterial and moisture-absorbing properties from the base textile layer that comes into contact with the skin. silver coated conductive threads in the inner layer (21) that enable the product to be made Because it is made of nylon (AgNy) filaments, the TENS current is distributed homogeneously. PVDF nanofibers that enable distribution on the patient's skin produced by the “electrospinning” method and laminated to the inner layer (21) Energy related to TENS current is generated thanks to the piezoelectric layer (22). providing, the piezoelectric layer (22) has a low piezoelectric coefficient (d33 ≈ -33 15 Despite its superior flexibility (Young's modulus ≈ 2-4 GPa), biocompatibility, and (pC / N) Wearable medical devices in terms of washability and suitability for textile integration. It has the feature of being made of PVDF material, which is preferred in its applications. The electronic layer (23) is on a 50 μm thick flexible polyimide PCB. piezoelectric layer 20 to accommodate the mounted electronic components (22) to operate at least one energy harvesting module (221) on it. to provide at least one power management module (231) of the electronic layer (23), energy budget analysis module (232) and intermittent stimulation protocol, TENS generator circuit (233), sensor (234), microcontroller (235) wireless communication to ensure that the module (236) is located between the outer layer (24) and the inner layer (21), 25 Protecting the piezoelectric layer (22) from external factors, electronic layer (23) to provide at least an energy harvest thanks to the piezoelectric layer inside (22) operation of the mechanism, piezoelectric with energy harvesting mechanism mechanical dipole moments in the PVDF material in the layer (22) its reorientation under deformation and opposite electric charge on the surfaces 30 22 to cause it to occur when the patient bends or straightens their knee. each incoming bending cycle produces an AC voltage, and this voltage is then passed through a Schottky diode. rectifying it with its bridge and converting it to DC, thanks to its multi-layered structure. Thanks to the stack configuration, a voltage typically between 4 V and 12 V is provided. To ensure proper operation of the TENS application, a voltage of 20 to 50 V 5 is required. two-stage cascade boost converter topology is used when reaching these values to enable its use, Maximum Power on the piezoelectric layer (22) Compared to constant impedance loading, thanks to the Point Tracking (MPPT) algorithm. Ensuring that 30-40% more energy is harvested, the energy TENS 10 is activated when the supercapacitor is activated and the threshold value (e.g., 30 V) is reached. to enable the module to be activated automatically, electronic layer (23) with its artificial intelligence module, it analyzes the user's activity pattern. learning and automatically optimizing the most effective TENS protocol for the user. to ensure that the artificial intelligence algorithm in question makes two-stage decision-making the mechanism should include, in the first stage, the Random Forest classifier (150 trees, 15 Using the maximum depth 8, minimum leaf node sample 5) related estimation of user / patient activity classification and pain level the second stage involves maximizing therapeutic efficacy under energy constraints. performing multi-objective optimization to provide limited reinforcement learning within the optimization in question 20 To enable the implementation of the (constrained reinforcement learning) approach. All real-time with the microcontroller (235) on the electronic layer (23) sensor (234) data and artificial intelligence model inferences are made entirely electronically. to enable therapeutic decisions to be processed on the cloud layer (23). To enable local delivery without requiring a connection, raw sensor data, 25 Information in the form of personal health data and model parameters shall under no circumstances be used to prevent data from being transferred outside the device, thanks to its layered data architecture, to optimize the privacy-functionality balance, at the electronic layer (23) Bluetooth Low Energy 5.0 (BLE) chip in the wireless communication module (236) This allows authorized data received from the user to be transferred to external environments, 30 23 Two-phase (biphasic) with TENS Generator Circuit (233) in the electronic layer (23) generating pulses (vibrations), low-frequency TENS (Transcutaneous Electrical Nerve Imaging) By stimulation (2-10 Hz), endorphin and enkephalin release is increased in the user, resulting in a longer-lasting effect. to achieve a sustained analgesic effect, high-frequency TENS (50-150 Hz) through gate control theory, the acute pain signals present in the relevant user are 5 at least one textile material structured to ensure that it is blocked (2), - the user and / or patient using the textile material (2) with other people at least one electronic device configured to enable communication (3), - wireless 10 located on textile material (2) and operated on electronic device (3). To connect with the communication module (236), the user's interface on it Customize TENS parameters, the energy status of textile material (2) user monitoring, user pain diary, and treatment history. to enable the display of all real-time transmitted from textile material (2) stimulation control data, sensor (234) data, artificial intelligence processing 15 to enable processing within the electronic device (3), after obtaining permission from the user Enabling the optional cloud backup feature, textile material (2) Securely transmitting the user's treatment data to at least one external server. (5) at least one application (4) configured to enable sending and - only 20 that connect with application (4) and are transmitted from application (4) anonymized treatment summary data (VAS scores, stimulation durations, energy) (statistics) to be stored in HIPAA / GDPR compliant and encrypted format. a medical system characterized by having at least one external server (5) configured (1).

2. It has a four-layered structure, with the part that comes into contact with the user's skin having... bamboo or merino wool coated with silver nanoparticles in the inner layer (21) Claim 1, characterized by textile material (2) which enables its use. a medical system like (1). 24 3. It also contains conductive textile lines (conductive threads) and electrodes. medical as in Claim 1 or 2, characterized by the inner layer (21) which serves the function. a system (1).

4. With its silver-coated nylon (AgNy) filaments, it increases its surface resistance by 1-5. By maintaining a level of Ω / cm, the TENS current is distributed homogeneously across the user's skin. any of the above claims characterized by the inner layer (21) which provides a medical system like one (1).

5. Electrospinning of PVDF (polyvinylidene fluoride) nanofibers with a thickness of 200-500 10 It is produced with a diameter of nm and laminated into textiles or by lead zirconate titanate (PZT). composite of nanoparticles (30-40% by volume) in a flexible silicone matrix characterized by the pieozoelectric layer (22) created by turning it into a panel. a medical system as in any of the above requests (1).

6. High flexibility despite low piezoelectric coefficient (d33 ≈ -33 pC / N), low density (1.78 g / cm³), wearable comfort and non-toxic. pieozoelectric layer, preferably made of PVDF material due to its structure (22) A medical claim like any of the above claims characterized by (22). system (1). 20 7. Multilayer piezoelectric stack to compensate for the flexibility disadvantage. The configuration is implemented by: each panel having 4-6 layers of PVDF. It is created by laminating layers on top of each other via serial connection, thus creating a single layer. 25 a medical system as in Claim 6 characterized by a piezoelectric layer (22) (1).

8. The panels should be strategically placed in areas with high mechanical stress (knee joint). circumference, ankle, hip joint, mid-back, shoulder area and elbow circumference) 5×10 30 Preferably a total of 6 panels, each measuring in cm, to be placed for the user. It is in use and seating is obtained from each panel at different activity levels. In the position (minimal movement) 0.05-0.2 mW per panel, in normal walking (4-5 0.8-1.5 mW per panel at high speed (km / h), 1.5-2.5 mW per panel at high speed (6-7 km / h). mW, generating 2.0-3.5 mW of power per panel during stair climbing activity 5 to provide the total power output of 6 panels under normal walking conditions for the user. configured to ensure an average output level of 9 mW (6 × 1.5 mW) Any of Claims 5 to 7 characterized by a pieozoelectric layer (22) a medical system like one (1).

9. Multilayer piezoelectric stack configuration in each panel (3-5 layers of PZT- By applying series connection of silicon composite, the output voltage per single panel is 2- A series-connected stack structure that enables a 3x increase, with each layer producing... The voltage is summed (V_total = n × V_layer, n: number of layers) and sent to the boost converter. 15 characterized by a piezoelectric layer (22) that enables the input voltage to be increased A medical system as in any of Claims 5 to 8 (1).

10. It is located in the piezoelectric layer (22), the crystal of PVDF material. the β-phase in its structure exhibits a piezoelectric effect under mechanical deformation providing, in this way, the strain applied to its own material, 20 reorientation of dipole moments and accumulation of electric charge on surfaces the above characterized by the energy harvesting module (221) which causes a medical system like any of the requests (1).

11. The user who places the textile material (2) on it has 25 when his knee is bent. the panel is subjected to compressive stress (εc = -0.5% to -2%) when the knee is opened each bend allows it to be subjected to tensile stress (εt = between +0.5% and +2%). AC voltage of the loop (single panel: typical 2-8 V, stack configuration: 6-24 V; It is characterized by its energy harvesting module (221) which enables it to produce current: 0.1-0.5 mA). a medical system as in any of the above requests (1). 30 26 12. With its Schottky diode bridge (low forward voltage drop <0.3V) The process of rectifying AC to convert it to DC, and the resulting power output is calculated using the formula P = V × I × η. enabling calculation using the formula (η: conversion efficiency ~60-75%), Achieving a value of 1.5 mW / panel during the user's average walking activity is recommended. providing a total of 9 mW of continuous power generation with the use of six panels. characterized by the energy harvesting module (221) which enables its implementation a medical system like any of the above requests (1).

13. To reach the 20-50 V level required for TENS application, two 10 Using a staged cascade boost converter topology: first stage The second one increases the piezoelectric output (4-12 V) to an intermediate voltage level (15-20 V). The step is the one that raises the intermediate voltage to the target TENS voltage (20-50 V, adjustable). The subject is the cascade structure, which has a higher rate of conversion compared to a single-stage high-rate conversion. with energy harvesting module (221) which enables operation with efficiency (85-92%) 15 a medical system as in any of the claims characterized above (1).

14. Using the flyback converter topology, the transformer winding ratio (N2 / N1) Energy harvesting module (221) 20 with voltage boost rate that can be directly controlled a medical condition as in any of the above claims characterized by system (1).

15. Since the internal impedance of piezoelectric sources is capacitive, Unlike traditional photovoltaic MPPT algorithms (Perturb & Observe), 25 Impedance matching-based MPPT (Maximum Power Point Tracking) approach to enable its use, the equivalent of the panel in the piezoelectric layer (22) capacitance (Cp) and mechanical excitation frequency (f_mech) in real time measurement of the optimal load impedance using the formula Z_opt = 1 / (2π × f_mech × Cp). (f: mechanical vibration frequency, C_p: piezoelectric capacitance) microcontroller (235) 30 27 the calculation of this impedance and the duty cycle of the DC-DC converter. to ensure that it adjusts adaptively in a way that satisfies the matching condition the above characterized by the structured energy harvesting module (221) a medical system like any of the requests (1).

16. In the MPPT algorithm, the perturb-and-observe (P&O) method uses piezoelectric technology. enabling the use of a modified version for the source, load Maximum power transfer by adjusting the impedance every 100 ms to enable this, with the approach in question, to constant load impedance making it possible to obtain 15-25% more energy harvest compared to 10 characterized by the energy harvesting module (221) structured to provide a medical system like any of the above requests (1).

17. AC / DC converter on flexible PCB, cascade boost converter topology, supercapacitor (100-500 mF capacity, preferably 470 mF), microcontroller (235) 15 (ultra-low power ARM Cortex-M0+ or ​​RISC-V based, <1 mW in active mode, power consumption <5 μW in sleep mode) and containing the TENS generator circuit (233) any of the above claims characterized by the electronic layer (23) a medical system like one (1).

18. It has at least one power management module (231), energy budget analysis including the module (232), sensor (234) and wireless communication module (236) any of the above claims characterized by the electronic layer (23) a medical system like one (1).

19. Normal walk with the energy budget analysis module (232) on it. Under these conditions, the total power generation P_generation = 9 mW, using conventional continuous TENS therapy requires 0.4-4 W of power, allowing the user to experience intermittent (intermittent) therapy. After accepting the information regarding the implementation of the stimulation protocol, The energy consumption of a single TENS pulse is calculated using the formula E_pulse = V × I × t_pulse, which is 30. 28 calculation; where V = 40 V (typical operating voltage), I = 40 mA (average (therapeutic current), when t_pulse = 200 μs (pulse width) E_pulse = 40 × 0.04 Calculated as × 200×10⁻⁶ = 0.32 mJ at a stimulation frequency of 80 Hz. The number of pulses per minute is 4800, and the instantaneous average power consumption is P_TENS = 5 is configured to allow calculation in the form of 0.32×10⁻³ × 80 = 25.6 mW. any of the above claims characterized by the electronic layer (23) a medical system like one (1).

20. P_TENS calculated with the energy budget analysis module (232) on it. If the value exceeds 9 mW of continuous production capacity, duty cycle 10 To ensure the cycle principle is applied, the duty cycle is 20-25%. when adjusted (e.g., 30 seconds stimulation / 90 seconds charging cycle or 60 (second stimulation / 180 second charging cycle), average power consumption P_average = P_TENS × duty_cycle = 32 mW × 0.25 = 8 mW should be reduced to this. To ensure this value remains below 9 mW of production capacity, artificial intelligence 15 The algorithm determines the current energy status and user activity level. by evaluating and ensuring the dynamic optimization of the duty cycle. as in Claim 19, characterized by the structured electronic layer (23). a medical system (1).

21. With the power management module (231) and the piezoelectric layer (22) Raising the incoming voltage to the level required for TENS application is a two-stage process. To enable its implementation with a cascade boost converter topology, the first At this stage, the piezoelectric output of the LTC3108 ultra-low power DC-DC boost converter Intermediate voltage 25 (6-24 V from stack configuration or 2-8 V from single panel) electronically configured to allow the voltage to rise to the level (12-15 V) as in any of the above requests characterized by layer (23) a medical system (1). 29 22. In the second stage, with the power management module (231) on it, flyback converter thanks to its topology (e.g., LT3573 or custom-designed flyback circuit), the intermediate the voltage to the high voltage required by TENS (20-50 V, adjustable) This is achieved by ensuring the flyback converter provides galvanic isolation, 5. To ensure the creation of an additional layer of protection in terms of user security. The above is characterized by the electronic layer (23) which is structured to be a medical system like any of the requests (1).

23. Total conversion of cascade topology with power management module (231) on it its efficiency should be in the range of 78-85%, and the energy produced should be 100-500 mF 10 in high-capacity supercapacitor (ESR <50 mΩ, Maxwell Technologies BCAP series) electronic layer (23) configured to enable storage of (or equivalent) a medical condition as in any of the above claims characterized by system (1).

24. Charging time of the supercapacitor with the power management module (231) on it The charge should be calculated using the formula “t_charge = C × ΔV / I_charge”, where C = 200 mF. When ΔV = 20 V (charging from 10V to 30V) and I_charge = 9 mW / 20 V = 0.45 mA, The approximate full charge time is t_charge = 0.2 × 20 / 0.00045 ≈ 8889 seconds (approximately To ensure it is calculated as 148 minutes, or 2.5 hours of walking activity, use 20. the above is characterized by the structured electronic layer (23) a medical system like any of the requests (1).

25. With the microcontroller (235) on it, the voltage threshold value (e.g. 30V) is kept constant. monitoring, activation of the TENS generator circuit (233) when the threshold is exceeded, power 25 Power consumption is optimized to <5 μA in deep sleep mode and <2 mA in active mode. an electronic layer configured to allow current to be drawn at certain values (23) A medical claim like any of the above claims characterized by (23). system (1). 30 26. Intermittent energy budget analysis module (232) is run on it. Thanks to the stimulation protocol, intermittent TENS application (30-60 seconds active) (stimulation / 2-3 minutes rest) neural habituation compared to continuous stimulation. (Habituation) prevention and preservation of therapeutic efficacy, even in some chronic pain in these cases, to ensure that the therapeutic efficacy is increased 5 the above is characterized by the structured electronic layer (23) a medical system like any of the requests (1).

27. Intermittent energy budget analysis module (232) is run on it. thanks to the stimulation protocol, the user using textile material (2) has an intense 10 textiles during periods of physical activity (climbing stairs, running) by increasing the power generation of the material (2) to 18-24 mW, it will last longer allowing for longer stimulation durations or higher duty cycle values characterized by the electronic layer (23) structured to provide a medical system as in any of the above requests (1). 15 28. H-bridge MOSFET with TENS generator circuit (233) on it. To enable the generation of biphasic (two-phase) pulses using this configuration. having a bandwidth (50-250 μs), frequency (2-150 Hz), and intensity (10-80 mA) range. Control of pulse parameters using PWM (Pulse Width Modulation) 20 characterized by the electronic layer (23) structured to provide a medical system like any of the above requests (1).

29. Current limiting circuit (IEC 60601-) on the TENS generator circuit (233). Compliant with 2-10 standard (<80 mA), short circuit protection and overvoltage protection 25 an electronic layer configured to enable the integration of protection (23) A medical claim like any of the above claims characterized by (23). system (1). 31 30. Thanks to the sensors (234) on it and the 3-axis MEMS accelerometer (for example ADXL345) user's step count, movement intensity, joint flexion frequency enabling monitoring of physical activity in textiles thanks to the gyroscope detailed analysis of the movement patterns of the user using the material (2) 5 characterized by the electronic layer (23) structured to provide a medical system like any of the above requests (1).

31. Electrode-skin with capacitive pressure sensors (234) on it monitoring of contact quality with a temperature sensor (e.g. TMP116, ±0.2°C) (2) Thermal safety of textile material by monitoring skin surface temperature) sensitivity 10 to ensure that the data in question is checked using 100 Hz sampling configured to transmit to the microcontroller (235) with its frequency any of the above claims characterized by the electronic layer (23) a medical system like one (1).

32. Working on the electronic layer (23) as the brain of the textile material (2) characterized by a microcontroller (235) configured to perform the task a medical system like any of the above requests (1).

33. The machine learning algorithm on it, TENS parameter 20 Multi-objective optimization designed specifically for optimization structured to implement the objective optimization approach any of the above requests characterized by the microcontroller (235) a medical system like one (1).

34. In the machine learning algorithm on it [accelerometer_x, accelerometer_y, accelerometer_z, gyro_roll, gyro_pitch, gyro_yaw, step_frequency, motion_intensity_RMS, joint_flexion_angle, skin_impedance, supercapacitor_voltage, last_VAS_score, time_of_day, last_stimulation_duration, Running a 15-dimensional feature vector in the form of [cumulative_energy_state] 30 32 The above is characterized by the microcontroller (235) structured to be built on. a medical system like any of the requests (1).

35. The lightweight version of the Random Forest algorithm has 12 decision trees. to ensure its formation, the maximum depth of each tree is between 6 and 5 levels. its limitation, with this restriction, within 4KB RAM and 32KB flash memory. to ensure the model size is optimized so that it can work from the above requests characterized by the structured microcontroller (235) a medical system like any other (1).

36. In the Reinforcement learning module, the reward function R = -VAS_score × α + energy_efficiency × β + comfort_index × γ to ensure its definition; where α = 0.6 (pain reduction weight), β = 0.25 (energy VAS_score is the user's weighting (efficiency weighting), γ = 0.15 (comfort index weighting). reported pain level (0-10), energy_efficiency = (therapeutic_effect / 15 (energy expended) normalized value and comfort_index electrode-skin contact To ensure the use of a combination of quality and stimulation comfort level. The above is characterized by the microcontroller (235) structured to be built on. a medical system like any of the requests (1).

37. The energy constraint of the reward function being operated on, E_consumed(t) ≤ E_available(t) × safety_margin (safety_margin = 0.7) its definition, the reward function never being the energy of the supercapacitor To ensure that no more than 70% is consumed in a single stimulation session, the promise is... The subject is the action space of the reward function a = [frequency, 25 pulse_width, intensity, duty_cycle, electrode_pair_selection] to ensure that each parameter is defined and selected from a discrete set of values. The above is characterized by the microcontroller (235) structured to be built on. a medical system like any of the requests (1). 33 38. The machine learning approach being run on it is a traditional TinyML approach. Unlike fixed model inference (inference-only) solutions, on-device implementing incremental learning, user-specific to ensure model parameters are updated at the end of each session 5 of the above requests characterized by the structured microcontroller (235) a medical system like any other (1).

39. Model size is limited to 48 KB RAM and 256 KB flash memory. The ability to run in real-time on an ARM Cortex-M0+ processor, inference to ensure that the duration of a single decision cycle is less than 10 ms from the above requests characterized by the structured microcontroller (235) a medical system like any other (1).

40. The multi-objective optimization approach in machine learning, Energy-awareness (energy-15) that maximizes therapeutic efficacy under energy constraints. (aware) decision-making mechanism—the algorithm's ability to determine the current supercapacitor energy level using stimulation parameters and duty cycle as direct decision variables The point is that existing TinyML optimizes the cycle simultaneously. its fundamental difference from solutions (TensorFlow Lite Micro, Edge Impulse) 20 characterized by a microcontroller (235) configured to provide a medical system like any of the above requests (1).

41. The multi-objective optimization approach in machine learning, Adaptive sampling tailored to the dynamics of piezoelectric energy harvesting—high During activity periods, both energy production and data collection frequency are 25. while increasing, the textile material (2) in low activity ultra low power mode This involves transitioning to temporal pattern learning—an algorithm specific to TENS—that allows the user to understand pain. temporal correlations of patterns (morning stiffness, post-activity pain increase) (such as) learning about and initiating proactive stimulation are among the points present. Basically, 30 TinyML solutions (TensorFlow Lite Micro, Edge Impulse) 34 characterized by a microcontroller (235) configured to enable its separation. a medical system as in any of the above requests (1).

42. The multi-objective optimization approach in machine learning, the machine learning model in question is compressed with 8-bit integer quantization. The points that exist include the fact that the inference period is less than 5 ms. Based on TinyML solutions (TensorFlow Lite Micro, Edge Impulse) characterized by a microcontroller (235) configured to enable its separation. a medical system as in any of the above requests (1).

43. In the wireless communication module (236) on it, in the form of nRF52832 By enabling the use of Bluetooth Low Energy 5.0 (BLE) chips, textiles (2) the material communicates with (3) the application (4) on the user’s electronic device. characterized by the electronic layer (23) which is structured to enable it to establish a medical system as in any of the above requests (1). 15 44. The wireless communication module (236) on it uses GATT as the protocol. (Generic Attribute Profile) usage, user-specific TENS parameters, data such as energy status, activity data and pain scores can be securely stored. 20 configured to enable transmission to the application (4) (AES-128 encryption) any of the above claims characterized by the electronic layer (23) a medical system like one (1).

45. Overcurrent protection that automatically cuts off the power when the current exceeds 80 mA. Short circuit protection, overvoltage protection (discharge at >60V), thermal protection 25 (System shutdown at >45°C) and electrode-skin contact monitoring (when contact is interrupted) to enable the activation of safety layers (such as stimulation stopping) The above is characterized by the electronic layer (23) which is structured to be a medical system like any of the requests (1). 35 46. ​​Supercapacitor overcharge protection (charge cut-off at >55V) and balancing. electronic circuit configured to operate on cell balancing as in any of the above requests characterized by layer (23) a medical system (1).

47. To operate in compliance with HIPAA / GDPR, the user must open the application (4) with the approval of the application and only via BLE connection (4) anonymous statistical data transferred (total daily stimulation time, average) (pain score trend, energy production / consumption summary), according to user preference. 10 with external server (5) configured to store on its own optionally. a medical system as in any of the claims characterized above (1).

Images