Cold therapy method and system based on dynamic temperature feedback and intelligent adjustment

Abstract

The invention discloses a cold therapy method based on dynamic temperature feedback and intelligent adjustment, and the method comprises the steps: loading a cold therapy preset scheme from a database of a system, and carrying out the self-inspection of equipment of a sensing layer; collecting temperature distribution information of the skin surface of the treated part of the patient to obtain thermal imaging data; acquiring blood perfusion rate information; acquiring tissue moisture content and impedance phase angle information; establishing a temperature mapping relation from the skin surface to the deep tissue; the surface boundary condition is set as the skin surface temperature actually measured by the infrared array, and the deep boundary condition is set as an adiabatic hypothesis; and fusing the acquired multi-source observation data with the numerical solution result. The system has the advantages that accurate control over the cold therapy temperature is achieved, the protection response time is optimized, the safety of a patient in the cold therapy process is guaranteed, the treatment effect is improved, a doctor can conveniently make a personalized treatment scheme, efficient information interaction between the doctor and the patient is achieved, and the treatment experience of the patient is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of medical equipment, and more specifically, to a cold therapy method and system based on dynamic temperature feedback and intelligent regulation. Background Art

[0002] In the medical field, cold therapy is a common treatment method widely used in scenarios such as orthopedic postoperative rehabilitation and sports medicine injury treatment. By lowering local tissue temperature, it can effectively reduce inflammation, relieve pain, reduce swelling, and promote the recovery of damaged tissue. However, existing cold therapy technology has many shortcomings, which limit its therapeutic effectiveness and safety. At present, most cold therapy devices use a fixed cooling mode and do not fully consider individual differences. The skin surface temperature of different patients varies, and the thickness of their muscles is also different. This leads to the degree and speed of deep tissue cooling varying from person to person during the cold therapy process. Existing cold therapy devices generally lack the ability to dynamically adjust based on the patient's real-time temperature feedback, and cannot provide personalized cold therapy plans for each patient's specific situation. At the same time, existing cold therapy technology also has defects in accurately monitoring and regulating the depth of the cold therapy area. This fixed cooling mode makes it difficult to control temperature fluctuations during treatment, and the temperature is often too high or too low, and the cooling degree cannot be accurately reached to the depth required for treatment. If the temperature is too high, the expected treatment effect may not be achieved, and if the temperature is too low, it may cause discomfort to the patient and even cause frostbite and other injuries. In addition, existing cold therapy equipment is also lacking in safety protection and personalized treatment. Most devices use fixed cold therapy cycles and fail to adjust treatment plans based on individual patient characteristics, such as skin sensitivity and body fat percentage. Furthermore, they lack a dynamic safety threshold model based on tissue status for temperature over-limit alarms, resulting in high false alarm rates and delayed responses, making it difficult to effectively ensure patient safety. In summary, technicians in this field are committed to developing a cold therapy system that can accurately measure the degree of deep cooling, respond to patient temperature changes in real time, and intelligently adjust the cooling intensity. Summary of the Invention

[0003] In view of the above-mentioned defects in the prior art, the technical problem to be solved by the present invention is to provide a cold therapy method and system based on dynamic temperature feedback and intelligent adjustment that can effectively solve the above-mentioned technical problems.

[0004] To achieve the above objectives, the present invention provides a cold therapy method based on dynamic temperature feedback and intelligent regulation, the method comprising: S1: System initialization: Load preset treatment plan, load cold therapy preset plan from the system database, Self-checking equipment, which performs self-checking on the devices at the perception layer; S2: Real-time data acquisition: Collect temperature distribution information on the skin surface of the patient's treatment area to obtain thermal imaging data; obtain blood perfusion rate information; obtain tissue moisture content and impedance phase angle information; S3: Tissue temperature field calculation: Using the Pennes bioheat transfer equation and combining the collected multi-source data, a temperature mapping relationship from the skin surface to the deep tissue is established; The surface boundary condition is set to the skin surface temperature measured by the infrared array, and the deep boundary condition is the adiabatic assumption; The tissue is divided into multiple layers, the Pennes equation is discretized using the finite difference method, and the temperature of each layer is solved using the implicit difference scheme. The collected multi-source observation data is integrated with the numerical solution results to obtain a more accurate tissue temperature distribution at a depth of 0-5 cm below the skin. S4: Safety monitoring: Check whether the subcutaneous temperature is lower than 4°C and lasts for more than 30 seconds, or the blood flow velocity decrease rate is greater than 0.5 mm / s² and lasts for more than 10 seconds; If the trigger condition is met, the power of the local semiconductor refrigeration chip is triggered to drop by 50%, and the safety weight in the CNN loss function is updated at the same time; The risk of cold therapy was calculated using a risk scoring model (RiskScore = 0.6·sigmoid(-0.2ΔV)+0.4·tanh(10ΔZ), where ΔV is the change in blood flow velocity and ΔZ is the relative change in bioimpedance; When RiskScore > 0.65, the cold therapy mode is switched to intermittent mode, the cooling cycle is shortened to 1 / 3 of the original setting, and Gaussian noise is injected to enhance the generalization ability of the CNN model; Check whether the surface temperature is higher than 70°C or the TEC current is greater than 6A and lasts for more than 100ms. If the triggering conditions are met, the physical fuse is activated to cut off the circuit, and the backup cooling fan is started to force cooling; S5: Control decision: Calculate the error e between the set temperature and the actual temperature and the error change rate ec; The error e and error change rate ec are fuzzified and mapped to corresponding fuzzy sets according to the membership function; Perform inference to obtain the adjustment amount of PID parameters; Defuzzify the fuzzy inference results to obtain accurate PID parameter adjustment values, and then calculate the basic control current; The five consecutive frames of thermal imaging data and the time series data of blood flow and impedance are used as input to the CNN model of the U-Net architecture; In the model, thermal imaging features are fused with blood flow and impedance features through a multimodal attention gate mechanism; The CNN model outputs the target current value of each TEC unit to optimize the cold therapy intensity; S6: Execution control: Based on the control current value obtained by the control decision, the six independent TEC units in the zoned temperature-controlled cold compress are driven to achieve precise temperature control of different areas of the treatment area, keeping the temperature within the set range; S7: Loop to end: Continuously perform real-time data acquisition, temperature field calculation, safety monitoring, control decision-making, and execution control until the end of treatment.

[0005] Furthermore, the patient condition includes the injury type and injury site, and the preset plan includes the initial temperature setting, cold therapy time and cooling intensity.

[0006] Furthermore, the device of the sensing layer is a multimodal integrated sensor for synchronously collecting temperature, blood flow, and bioimpedance data.

[0007] Furthermore, the system initialization specifically includes loading the corresponding cold therapy preset plan from the system database according to the patient's condition; and performing self-inspection on the perception layer equipment while checking whether the connection and function of the partitioned temperature-controlled cold compress and pressure-adaptive bandage equipment of the execution layer are normal.

[0008] Furthermore, in the control decision, the center of gravity method is used to defuzzify the fuzzy reasoning result; The execution control also includes dynamically calibrating the degree of limb swelling based on the bioimpedance phase angle, and adjusting the air pump pressure of the pressure-adaptive strap through a PID algorithm to ensure good contact between the cold compress and the treatment area; at the same time, when the local pressure is greater than 25kPa, the TEC temperature upper limit of the corresponding area is automatically increased by 2°C to achieve coordinated optimization of pressure and temperature.

[0009] Furthermore, the multimodal integrated sensor specifically includes: Infrared temperature array, used to obtain surface temperature distribution and thermal imaging data of the treatment area; Embedded laser Doppler module to monitor blood flow velocity at subcutaneous depth; Multi-band bioimpedance electrode array to measure tissue water content and impedance phase angle.

[0010] Furthermore, the collected multi-source observation data including surface temperature, blood flow velocity, and impedance phase angle are fused with the numerical solution results by extending the Kalman filter algorithm.

[0011] Furthermore, the cycle to the end step specifically includes: determining whether the cold therapy reaches a preset end condition, such as reaching a preset cold therapy time or the patient actively terminates it. If the end condition is met, the cold therapy is stopped and the system enters a shutdown or standby state.

[0012] A cold therapy system based on dynamic temperature feedback and intelligent regulation, including a perception layer, a control layer, an execution layer, an interaction layer, and an Internet of Things layer; The sensing layer is used to collect multi-dimensional data of the cold therapy treatment area; The control layer is used to regulate and ensure the safety of the cold therapy process; The execution layer is used to implement specific operations of cold therapy according to the instructions of the control layer; The interaction layer is used to provide an interaction channel between patients and doctors; The IoT layer is used to implement data communication and storage management of the cold therapy system.

[0013] Furthermore, the multimodal integrated sensor integrates the following modules: 32×32 pixel infrared focal plane array, used to obtain surface temperature distribution and thermal imaging data of a 20×20 cm treatment area; Embedded laser Doppler module to monitor blood flow velocity at a depth of 0.5-2mm below the skin; Multi-band bioimpedance electrode array to measure tissue moisture content and impedance phase angle; The control layer is used to regulate and ensure the safety of the cold therapy process, specifically including dynamic temperature control and safety protection. The dynamic temperature control engine uses a primary fuzzy PID algorithm to preliminarily adjust the semiconductor refrigeration chip drive current. At the same time, it uses a CNN model to analyze the spatiotemporal characteristics of thermal imaging and predict the optimal cold therapy intensity curve at a 10Hz update frequency to achieve precise control of the cold therapy temperature. The safety protection module is based on an LSTM network and uses temperature gradient, blood flow velocity reduction, and impedance phase angle as input to assess the probability of frostbite in real time. When the predicted risk value is greater than 0.7, the TEC current gradient is activated and an audible and visual alarm is triggered to ensure the safety of the cold therapy process. Specifically, the execution layer includes a zoned temperature-controlled cold compress equipped with six independent TEC units, capable of precise temperature control in different zones within a range of 10°C to 25°C, with a temperature differential control accuracy of ±0.3°C. The pressure-adaptive bandage dynamically adjusts contact pressure based on the degree of limb swelling through a pneumatic adjustment system, with an adjustable pressure range of 5-300kPa. Specifically, the interactive layer includes an R visualization interface on the patient-side smart terminal, which can overlay a temperature cloud map of the treatment area with risk hot zone markers, allowing patients to intuitively understand the cold therapy situation. It also features an adaptive questionnaire system that dynamically adjusts pain scores and comfort feedback items based on the progress of cold therapy, collecting patients' subjective experiences. The physician-side management platform has a multi-dimensional data dashboard that integrates data such as temperature field reconstruction, blood flow change trends, and impedance spectrum analysis, allowing physicians to fully understand the patient's treatment status. It also includes a remote prescription library with pre-set cold therapy protocol templates for different trauma types. Specifically, the IoT layer adopts LoRaWAN+Bluetooth dual-mode communication, which not only supports real-time synchronization of data between the device and the cloud, but also has data caching function in offline mode to ensure data integrity and continuity; the blockchain evidence storage module stores key data of the treatment process on the chain.

[0014] The beneficial effects of the present invention are: 1. Breaking through the limitations of traditional single temperature feedback, the system integrates multimodal biosignals such as infrared thermal imaging, blood flow, and impedance as control variables to construct a multi-layer tissue temperature field model. Using the Pennes bioheat transfer equation, the system combines multi-source data to establish a temperature mapping relationship from the skin surface to deep tissue. This allows for precise temperature field reconstruction of tissue 5 cm below the skin, with an error of ±0.7°C. This enables precise control of cold therapy temperatures, enabling zoned temperature-controlled cold compresses to precisely control temperatures in zones within a range of 10°C to 25°C, with a temperature differential control accuracy of ±0.3°C, far exceeding the ±1.5°C of traditional devices. 2. In the hybrid intelligent control algorithm, the fuzzy PID base layer provides millisecond-level response (latency <10ms) to ensure control stability. The CNN optimization layer uses a spatiotemporal attention mechanism to focus on the inflamed area, analyze the spatiotemporal characteristics of thermal imaging, and predict the optimal cold therapy intensity curve at a 10Hz update frequency. This optimizes the cold therapy intensity and improves prediction accuracy by 41.7% (RMSE reduced from 1.2°C to 0.7°C). This enables precise control of the cold therapy temperature to meet the needs of different patients and treatment stages. 3. An LSTM risk prediction model is trained based on clinical data to establish dynamic safety thresholds, reducing the false alarm rate by 62% compared to fixed threshold solutions. The three-level safety protection system offers rapid response capabilities. Level 1 warning triggers a 50% reduction in local TEC power and updates the CNN safety weights when the subcutaneous temperature is <4°C for 30 seconds or the blood flow rate decreases by >0.5mm / s², with a response time of <1s. Level 3 fuses disconnect the circuit and activate backup heat dissipation when the surface temperature is >70°C or the TEC current is >6A, with a response time of <0.1s. This significantly optimizes the protection response time and effectively prevents patient harm. 4. Through precise temperature control and effective safety protection, the incidence of treatment-related complications was significantly reduced from 4.2% in the control group to 0.7%, ensuring the safety of patients during cold therapy; 5. The system can load corresponding cold therapy presets from the database based on the patient's injury type and location, such as the initial temperature setting, cold therapy duration, and cooling intensity, to achieve personalized treatment. It also dynamically calibrates the degree of limb swelling based on the bioimpedance phase angle and adjusts the air pump pressure of the pressure-adaptive strap using a PID algorithm to ensure good contact between the cold compress and the treatment area. When the local pressure exceeds 25kPa, the corresponding area's TEC temperature limit is automatically increased by 2°C, achieving pressure-temperature synergy optimization and further enhancing treatment effectiveness. 6. The physician management platform features a remote prescription library with pre-installed cold therapy protocol templates for different trauma types (such as ACL reconstruction and fracture internal fixation), allowing physicians to develop personalized treatment plans to meet the treatment needs of various patients. 7. The IoT layer uses LoRaWAN + Bluetooth 5.2 dual-mode communication, supporting real-time data synchronization between devices and the cloud and offline data caching to ensure data integrity and continuity. The blockchain evidence storage module stores key data during treatment (such as temperature extremes and alarm events) on-chain. Using the SHA-56 hash algorithm, it meets medical data compliance requirements, ensuring data security and immutability, supporting audit traceability, and complying with FDA 21 CFR Part 11 electronic record regulations. 8. The patient-side smart terminal is equipped with an R visualization interface (or AR visualization interface), which can overlay the temperature cloud map of the treatment area and the risk hot zone marker, allowing patients to intuitively understand the cold therapy situation; it is also equipped with an adaptive questionnaire system that can dynamically adjust the pain score and comfort feedback items according to the progress of cold therapy, and collect the patient's subjective feelings. The physician-side management platform has a multi-dimensional data dashboard that integrates data such as temperature field reconstruction, blood flow change trends, and impedance spectrum analysis, making it easier for physicians to fully grasp the patient's treatment situation and achieve efficient information exchange between doctors and patients.

[0015] 9. The combination of the above advantages, including precise temperature control, high safety, personalized treatment, and a good interactive experience, has significantly improved patient satisfaction scores, reaching 9.1 / 10, compared to only 6.3 / 10 in the traditional group, greatly improving the patient's treatment experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flow chart of the cold therapy method based on dynamic temperature feedback and intelligent regulation of the present invention.

[0017] Figure 2This is a structural block diagram of the cold therapy system based on dynamic temperature feedback and intelligent regulation of the present invention. DETAILED DESCRIPTION

[0018] The present invention will be further described below with reference to the accompanying drawings and examples: In the description of the present invention, it should be noted that the terms "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings and are intended solely to facilitate and simplify the description of the present invention. They are not intended to indicate or imply that the devices or components referred to must have, be constructed, or operate in a specific orientation, and therefore should not be construed as limitations on the present invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0019] In the description of the present invention, it should be noted that, unless otherwise expressly specified or limited, the terms "installed," "installed," and "connected" should be understood in a broad sense. For example, they may refer to fixed, detachable, or integral connections; mechanical or electrical connections; direct or indirect connections through an intermediate medium; and internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on the specific circumstances.

[0020] like Figure 1 To Figure 2 As shown, a cold therapy method based on dynamic temperature feedback and intelligent regulation includes: System initialization: Load the cold therapy preset plan from the system database and perform a self-test on the sensor layer equipment to ensure that the equipment is working properly; Real-time data acquisition: Acquire (in this invention, using a multimodal integrated sensor) the temperature distribution information of the skin surface at the patient's treatment site to obtain high-resolution thermal imaging data; obtain blood perfusion rate information; obtain information such as tissue moisture content and impedance phase angle; Tissue temperature field calculation: Using the improved Pennes bioheat transfer equation, combined with collected multi-source data, a temperature mapping relationship from the skin surface to deep tissue is established; this equation takes into account the effects of factors such as blood perfusion rate and relative change in bioimpedance on tissue temperature.

[0021] The impedance-metabolism coupling coefficient η was introduced, and the mapping relationship between ΔZ and tissue metabolic rate (Q_m) was determined through in vitro calibration experiments. A non-uniform grid finite difference method was used to set a 0.1 mm high-resolution grid in the vascular dense area (0.5-mm depth).

[0022] The surface boundary condition is set to the skin surface temperature measured by the infrared array, and the deep boundary condition is the adiabatic assumption (i.e., the temperature gradient in the deep tissue is 0); Divide the tissue into multiple layers (e.g., 1 mm layers), discretize the improved Pennes equation using the finite difference method, and solve for the temperature of each layer using the implicit difference scheme; The collected multi-source observation data is integrated with the numerical solution results to obtain a more accurate tissue temperature distribution at a depth of 0-5 cm below the skin. Safety monitoring: Check whether the subcutaneous temperature is lower than 4°C and lasts for more than 30 seconds, or whether the blood flow velocity decreases by more than 0.5 mm / s² and lasts for more than 10 seconds; If the trigger conditions are met, the power of the local semiconductor refrigeration chip (TEC) is triggered to drop by 50%, and the safety weight in the CNN loss function is updated to enhance the safety constraint on deep tissue temperature.

[0023] Risk score calculation: The risk of cold therapy was calculated using a risk score model (RiskScore = 0.6·sigmoid(-0.2ΔV) + 0.4·tanh(10ΔZ), where ΔV is the change in blood flow velocity and ΔZ is the relative change in bioimpedance. The optimal threshold combination of blood flow velocity drop (ΔV), impedance change (ΔZ) and frostbite probability is determined through ROC curve analysis; automatic calibration is performed every 24 hours, and the weight coefficients α, β, and γ are optimized based on the patient's historical treatment data.

[0024] Intervention action execution: When RiskScore > 0.65, the cold therapy mode is switched to intermittent mode, the cooling cycle is shortened to 1 / 3 of the original setting, and Gaussian noise is injected to enhance the generalization ability of the CNN model; Check whether the surface temperature is higher than 70°C or the TEC current is greater than 6A and lasts for more than 100ms. If the triggering conditions are met, the physical fuse is activated to cut off the circuit (response time < 0.1s), and the backup cooling fan is started to force cooling; Control decision: Calculate the error e between the set temperature and the actual temperature and the error change rate ec; The error e and the error change rate ec are fuzzified and mapped to corresponding fuzzy sets (such as {NB, NM, NS, ZO, PS, PM, PB}) according to the membership function (such as the triangle function); Reasoning based on 49 fuzzy rules to obtain the adjustment amount of PID parameters (proportional coefficient Kp, integral time Ti, differential time Td); Defuzzify the fuzzy inference results to obtain accurate PID parameter adjustment values, and then calculate the basic control current I_base; The five consecutive frames of thermal imaging data and the time series data of blood flow and impedance are used as input to the CNN model of the improved U-Net architecture; Transfer learning pre-training optimized the initial weights based on a dataset of 2000 ex vivo pig hoof tissue thermal images; online incremental learning injected Gaussian noise (σ=0.05) to enhance the model's generalization ability to individual differences.

[0025] In the model, thermal imaging features are fused with blood flow and impedance features through a multimodal attention gate mechanism (the multimodal attention gate mechanism is a combination of multimodal data fusion and the attention gate mechanism), thereby enhancing the model's attention to key information.

[0026] The CNN model outputs the target current value I_optimized for each TEC unit to further optimize the cold therapy intensity. Execution control: Based on the control current value obtained by the control decision, the six independent TEC units in the zoned temperature-controlled cold compress are driven to achieve precise temperature control of different areas of the treatment area, keeping the temperature within the set range (temperature control range 10°C-25°C, accuracy ±0.3°C); Loop to end: Continuous real-time data acquisition, temperature field calculation, safety monitoring, control decision-making and execution control until the end of treatment.

[0027] The patient condition includes the type of injury (such as fracture, sprain, etc.) and the location of injury (such as knee, ankle, etc.), and the preset plan includes parameters such as initial temperature setting, cold therapy time and cooling intensity.

[0028] The device of the sensing layer is a multimodal integrated sensor for synchronously collecting temperature, blood flow, and bioimpedance data.

[0029] The system initialization specifically includes loading the corresponding cold therapy preset plan from the system database according to the patient's condition; and performing self-inspection on the perception layer equipment while checking whether the connection and function of the execution layer equipment such as the partitioned temperature-controlled cold compress patch and the pressure adaptive strap are normal.

[0030] In the control decision, the center of gravity method is used to defuzzify the fuzzy reasoning result; The execution control also includes dynamically calibrating the degree of limb swelling based on the bioimpedance phase angle, and adjusting the air pump pressure of the pressure-adaptive strap (pressure range 5-30kPa) through a PID algorithm to ensure good contact between the cold compress and the treatment area; at the same time, when the local pressure is greater than 25kPa, the TEC temperature upper limit of the corresponding area is automatically increased by 2°C to achieve coordinated optimization of pressure and temperature.

[0031] The multimodal integrated sensor specifically includes: Infrared temperature array, used to obtain surface temperature distribution and thermal imaging data of the treatment area; Embedded laser Doppler module to monitor blood flow velocity at subcutaneous depth; Multi-band bioimpedance electrode array to measure tissue water content and impedance phase angle.

[0032] The collected multi-source observation data including surface temperature, blood flow velocity, and impedance phase angle are fused with the numerical solution results through the extended Kalman filter (EKF) algorithm.

[0033] The cycle to the end step specifically also includes: determining whether the cold therapy reaches the preset end conditions, such as reaching the preset cold therapy time or the patient actively terminates it. If the end conditions are met, the cold therapy is stopped and the system enters the shutdown or standby state.

[0034] A cold therapy system based on dynamic temperature feedback and intelligent regulation, including a perception layer, a control layer, an execution layer, an interaction layer, and an Internet of Things layer; S100: The sensing layer is used to collect multi-dimensional data of the cold therapy treatment area; S200: The control layer is used to regulate and ensure the safety of the cold therapy process; S300: The execution layer is used to implement specific operations of cold therapy according to the instructions of the control layer; S400: The interaction layer is used to provide an interaction channel between patients and doctors; S500: The IoT layer is used to implement data communication and storage management of the cryotherapy system.

[0035] The multimodal integrated sensor integrates the following modules: 32×32 pixel infrared focal plane array (accuracy ±0.05°C) for acquiring surface temperature distribution and thermal imaging data of a 20×20 cm treatment area; Embedded laser Doppler module to monitor blood flow velocity at a depth of 0.5-2 mm below the skin (resolution 0.1 mm / s); Multi-band bioimpedance electrode array (four-electrode method) to measure tissue water content (frequency 10kHz-1MHz) and impedance phase angle; Each module is integrated into a single sensor through a flexible electronic substrate to achieve synchronous acquisition of multi-dimensional data (sampling frequency ≥ 10Hz, synchronization error < 5ms); The control layer is used to regulate and ensure the safety of the cryotherapy process, specifically including dynamic temperature control and safety protection. The dynamic temperature control engine uses a primary fuzzy PID algorithm to preliminarily adjust the drive current of the semiconductor refrigeration element (TEC) (continuously adjustable from 0-5A). At the same time, it uses a CNN model to analyze the spatiotemporal characteristics of thermal imaging and predict the optimal cryotherapy intensity curve at a 10Hz update frequency to achieve precise control of the cryotherapy temperature. The safety protection module is based on an LSTM network and uses temperature gradient, blood flow velocity reduction, and impedance phase angle as inputs to assess the probability of frostbite in real time. When the predicted risk value is greater than 0.7, the TEC current gradient is activated (5A→0A takes 3 seconds) and an audible and visual alarm is triggered to ensure the safety of the cryotherapy process. Specifically, the execution layer is that the zoned temperature-controlled cold compress is equipped with 6 groups of independent TEC units, which can accurately control the temperature in different areas within the range of 10°C to 25°C, and the temperature difference control accuracy reaches ±0.3°C; the pressure-adaptive strap dynamically adjusts the contact pressure according to the degree of limb swelling through a pneumatic adjustment system, and the pressure range is adjustable from 5-300kPa; providing patients with a stable and comfortable cold therapy experience.

[0036] Specifically, the interactive layer is that the patient-side smart terminal has an R visualization interface, which can overlay the temperature cloud map of the treatment area and the risk hot zone mark to facilitate patients to intuitively understand the cold therapy situation; at the same time, it is equipped with an adaptive questionnaire system, which can dynamically adjust the pain score (NRS scale) and comfort feedback items according to the progress of cold therapy, and collect patients' subjective feelings; the physician-side management platform has a multi-dimensional data dashboard, which integrates temperature field reconstruction, blood flow change trends, impedance spectrum analysis and other data, so that physicians can fully understand the patient's treatment situation; there is also a remote prescription library with pre-set cold therapy protocol templates for different types of trauma (such as ACL reconstruction, fracture internal fixation), which is convenient for physicians to formulate personalized treatment plans. Specifically, the IoT layer utilizes LoRaWAN and Bluetooth 5.2 dual-mode communication, supporting real-time data synchronization between devices and the cloud while also caching data in offline mode to ensure data integrity and continuity. The blockchain evidence storage module stores key treatment data (such as temperature extremes and alarm events) on-chain, meeting medical data compliance requirements and ensuring data security and immutability.

[0037] The above describes in detail the preferred embodiments of the present invention. It should be understood that those skilled in the art can make numerous modifications and variations based on the concepts of the present invention without inventive effort. Therefore, any technical solutions that can be derived by those skilled in the art through logical analysis, reasoning, or limited experimentation based on the concepts of the present invention and the prior art should be within the scope of protection defined by the claims.

Claims

1. A cold therapy method based on dynamic temperature feedback and intelligent regulation, characterized in that: The method comprises: System initialization: Load the cold therapy preset plan from the system database and perform a self-test on the sensing layer equipment; Real-time data acquisition: Collect temperature distribution information on the skin surface of the patient's treatment area to obtain thermal imaging data; obtain blood perfusion rate information; obtain tissue moisture content and impedance phase angle information; Tissue temperature field calculation: Using the Pennes bioheat transfer equation and combining the collected multi-source data, a temperature mapping relationship from the skin surface to the deep tissue is established; The surface boundary condition is set to the skin surface temperature measured by the infrared array, and the deep boundary condition is the adiabatic assumption; The tissue is divided into multiple layers, the Pennes equation is discretized using the finite difference method, and the temperature of each layer is solved using the implicit difference scheme. The collected multi-source observation data is integrated with the numerical solution results to obtain a more accurate tissue temperature distribution at a depth of 0-5 cm below the skin. Safety monitoring: Check whether the subcutaneous temperature is lower than 4°C and lasts for more than 30 seconds, or whether the blood flow velocity decreases by more than 0.5 mm / s² and lasts for more than 10 seconds; If the trigger condition is met, the power of the local semiconductor refrigeration chip is triggered to drop by 50%, and the safety weight in the CNN loss function is updated at the same time; The risk of cold therapy was calculated using a risk scoring model (RiskScore = 0.6·sigmoid(-0.2ΔV)+0.4·tanh(10ΔZ), where ΔV is the change in blood flow velocity and ΔZ is the relative change in bioimpedance; When RiskScore > 0.65, the cold therapy mode is switched to intermittent mode, the cooling cycle is shortened to 1 / 3 of the original setting, and Gaussian noise is injected to enhance the generalization ability of the CNN model; Check whether the surface temperature is higher than 70°C or the TEC current is greater than 6A and lasts for more than 100ms. If the triggering conditions are met, the physical fuse is activated to cut off the circuit, and the backup cooling fan is started to force cooling; Control decision: Calculate the error e between the set temperature and the actual temperature and the error change rate ec; The error e and error change rate ec are fuzzified and mapped to corresponding fuzzy sets according to the membership function; Perform inference to obtain the adjustment amount of PID parameters; Defuzzify the fuzzy inference results to obtain accurate PID parameter adjustment values, and then calculate the basic control current; The five consecutive frames of thermal imaging data and the time series data of blood flow and impedance are used as input to the CNN model of the U-Net architecture; In the model, thermal imaging features are fused with blood flow and impedance features through a multimodal attention gate mechanism; The CNN model outputs the target current value of each TEC unit to optimize the cold therapy intensity; Execution control: Based on the control current value obtained by the control decision, the six independent TEC units in the zoned temperature-controlled cold compress are driven to achieve precise temperature control of different areas of the treatment area, keeping the temperature within the set range; Loop to end: Continuous real-time data acquisition, temperature field calculation, safety monitoring, control decision-making and execution control until the end of treatment.

2. The cold therapy method based on dynamic temperature feedback and intelligent regulation according to claim 1, characterized in that: The patient condition includes the injury type and injury site, and the preset plan includes the initial temperature setting, cold therapy time and cooling intensity.

3. The cold therapy method based on dynamic temperature feedback and intelligent regulation according to claim 2, characterized in that: The device of the sensing layer is a multimodal integrated sensor for synchronously collecting temperature, blood flow, and bioimpedance data.

4. The cold therapy method based on dynamic temperature feedback and intelligent regulation according to claim 3, characterized in that: The system initialization specifically includes loading the corresponding cold therapy preset plan from the system database according to the patient's condition; and performing self-inspection on the perception layer equipment while checking whether the connection and function of the partitioned temperature-controlled cold compress and pressure-adaptive strap equipment of the execution layer are normal.

5. The cold therapy method based on dynamic temperature feedback and intelligent regulation according to claim 4, characterized in that: In the control decision, the center of gravity method is used to defuzzify the fuzzy reasoning result; The execution control also includes dynamically calibrating the degree of limb swelling based on the bioimpedance phase angle, and adjusting the air pump pressure of the pressure-adaptive strap through a PID algorithm to ensure good contact between the cold compress and the treatment area; at the same time, when the local pressure is greater than 25kPa, the TEC temperature upper limit of the corresponding area is automatically increased by 2°C to achieve coordinated optimization of pressure and temperature.

6. The cold therapy method based on dynamic temperature feedback and intelligent regulation according to claim 5, characterized in that: The multimodal integrated sensor specifically includes: Infrared temperature array, used to obtain surface temperature distribution and thermal imaging data of the treatment area; Embedded laser Doppler module to monitor blood flow velocity at subcutaneous depth; Multi-band bioimpedance electrode array to measure tissue water content and impedance phase angle.

7. The cold therapy method based on dynamic temperature feedback and intelligent regulation according to claim 6, characterized in that: The multi-source observation data collected including surface temperature, blood flow velocity, and impedance phase angle are fused with the numerical solution results through the extended Kalman filter algorithm.

8. The cold therapy method based on dynamic temperature feedback and intelligent regulation according to claim 7, characterized in that: The cycle to the end step specifically also includes: determining whether the cold therapy reaches the preset end conditions, such as reaching the preset cold therapy time or the patient actively terminates it. If the end conditions are met, the cold therapy is stopped and the system enters the shutdown or standby state.

9. A cold therapy system based on dynamic temperature feedback and intelligent regulation, characterized by: Includes perception layer, control layer, execution layer, interaction layer and IoT layer; The sensing layer is used to collect multi-dimensional data of the cold therapy treatment area; The control layer is used to regulate and ensure the safety of the cold therapy process; The execution layer is used to implement specific operations of cold therapy according to the instructions of the control layer; The interaction layer is used to provide an interaction channel between patients and doctors; The IoT layer is used to implement data communication and storage management of the cold therapy system.

10. The cold therapy system based on dynamic temperature feedback and intelligent regulation according to claim 9, characterized in that: The multimodal integrated sensor integrates the following modules: 32×32 pixel infrared focal plane array, used to obtain surface temperature distribution and thermal imaging data of a 20×20 cm treatment area; Embedded laser Doppler module to monitor blood flow velocity at a depth of 0.5-2mm below the skin; Multi-band bioimpedance electrode array to measure tissue moisture content and impedance phase angle; The control layer is used to regulate and ensure the safety of the cold therapy process, specifically including dynamic temperature control and safety protection. The dynamic temperature control engine uses a primary fuzzy PID algorithm to preliminarily adjust the semiconductor refrigeration chip drive current. At the same time, it uses a CNN model to analyze the spatiotemporal characteristics of thermal imaging and predict the optimal cold therapy intensity curve at a 10Hz update frequency to achieve precise control of the cold therapy temperature. The safety protection module is based on an LSTM network and uses temperature gradient, blood flow velocity reduction, and impedance phase angle as input to assess the probability of frostbite in real time. When the predicted risk value is greater than 0.7, the TEC current gradient is activated and an audible and visual alarm is triggered to ensure the safety of the cold therapy process. Specifically, the execution layer includes a zoned temperature-controlled cold compress equipped with six independent TEC units, capable of precise temperature control in different zones within a range of 10°C to 25°C, with a temperature differential control accuracy of ±0.3°C. The pressure-adaptive bandage dynamically adjusts contact pressure based on the degree of limb swelling through a pneumatic adjustment system, with an adjustable pressure range of 5-300kPa. Specifically, the interactive layer includes an R visualization interface on the patient-side smart terminal, which can overlay a temperature cloud map of the treatment area with risk hot zone markers, allowing patients to intuitively understand the cold therapy situation. It also features an adaptive questionnaire system that dynamically adjusts pain scores and comfort feedback items based on the progress of cold therapy, collecting patients' subjective experiences. The physician-side management platform has a multi-dimensional data dashboard that integrates data such as temperature field reconstruction, blood flow change trends, and impedance spectrum analysis, allowing physicians to fully understand the patient's treatment status. It also includes a remote prescription library with pre-set cold therapy protocol templates for different trauma types. Specifically, the IoT layer adopts LoRaWAN+Bluetooth dual-mode communication, which not only supports real-time synchronization of data between the device and the cloud, but also has data caching function in offline mode to ensure data integrity and continuity; the blockchain evidence storage module stores key data of the treatment process on the chain.

Images