A system and method for monitoring and managing exercise-induced fatigue

Abstract

The present application relates to a kind of motion fatigue monitoring and management system and motion fatigue monitoring management method, first provide a kind of motion fatigue monitoring and management system, including electronic fabric patch, multi-modal sensing module and data transmission and interaction module;Electronic fabric patch can be attached to the skin surface of subject, the electrode structure in it forms multi-modal sensing module, can be detected to the sweat flowing or directly to subject body;Electronic fabric patch is realized by porous fabric and gradient microcolumn structure dynamic adjustment of microscale sweat enrichment and excess sweat drainage, synchronously collect electromyogram, electrolyte, metabolite and hormone and other multidimensional physiological and biochemical indexes;In addition, AI evaluation module is deployed based on the multi-algorithm fusion model trained in the observation data in rest-exercise-recovery process, identifies 6 kinds of fatigue-related states and generates intervention scheme, in some embodiments, can reduce fatigue degree 51.82%, accelerate chronic fatigue recovery, applicable to athlete training monitoring and rehabilitation patient exercise management.

Description

Technical Field

[0001] This invention belongs to the field of wearable sensing technology, and in particular relates to a sports fatigue monitoring and management system and a sports fatigue monitoring and management method. Background Technology

[0002] Exercise-induced fatigue refers to a temporary decrease in the body's working capacity due to exercise, which can be recovered after rest. It is not a disease, but rather a "protective signal" from the body when the exercise load exceeds its current capacity, prompting you to pause or adjust the training intensity. This fatigue typically manifests as muscle soreness, decreased strength, slowed reaction time, and a slower heart rate recovery, and can be divided into two main categories: physical fatigue and psychological fatigue. While exercise-induced fatigue is a normal protective signal from the body, if it accumulates over a long period without recovery, it may develop into over-fatigue, causing various harms to both physical and mental health. With increasing awareness of exercise health, the risks of muscle damage and immune disorders caused by exercise-induced fatigue are becoming increasingly prominent. Marathons are a typical example of exercise-induced fatigue; historical data indicates that the incidence of sudden death among marathon participants due to dehydration and electrolyte imbalance is as high as 1 in 50,000.

[0003] Existing fatigue monitoring methods have significant limitations; traditional Borg scales and blood lactate tests cannot achieve real-time dynamic monitoring; conventional wearable devices can only collect single physiological signals and cannot simultaneously acquire biochemical indicators; although flexible electronic skin sensors can take into account multimodal detection, sweat accumulation during exercise interferes with electrophysiological signals, and insufficient sweat during rest limits biochemical monitoring, failing to cover the entire exercise-rest scenario; at the same time, there is a lack of intelligent assessment systems that can integrate multidimensional data, making it difficult to provide accurate and personalized intervention plans.

[0004] The core problems with existing sports fatigue monitoring devices can be summarized as follows: 1. Signal instability caused by dynamic changes in sweat; excessive sweat accumulation during exercise and insufficient sweat during rest both affect monitoring accuracy; 2. Disconnection between physiological and biochemical signal acquisition, failing to simultaneously acquire multidimensional indicators such as electromyography, electrolytes, metabolites, and hormones. Therefore, developing a wearable system that can adapt to changes in sweat, achieve stable acquisition of multidimensional signals, and provide intelligent fatigue management has become an urgent need in the field of sports and health. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a sports fatigue monitoring and management system and a sports fatigue monitoring and management method.

[0006] The technical solution adopted in this invention is: a sports fatigue monitoring and management system, comprising,

[0007] The electronic fabric patch is formed by bonding an ion-electroosmotic hydrogel layer, an electronic textile fabric layer, a PDMS micropillar layer, and an absorbent pad layer in sequence. The electronic textile fabric layer is printed with electromyographic sensing electrodes and biochemical sensing electrodes. The PDMS micropillar layer has multiple gradient micropillars and its surface is modified with a PVA / SiO2 composite coating.

[0008] The multimodal sensing module includes an electromyography sensing electrode and an iontophoresis electrode disposed on one side of the electronic textile fabric layer, and a biochemical sensing electrode disposed on the other side of the electronic textile fabric layer.

[0009] The data transmission and interaction module includes a Bluetooth Low Energy communication unit, a cloud storage unit, a mobile terminal APP, and an integrated circuit module. The integrated circuit module connects to the multimodal sensing module and interacts with the multimodal sensing module.

[0010] Preferably, the electronic textile fabric layer has a superhydrophobic substrate and also includes a liquid transport channel formed after plasma treatment, which is capable of unidirectional transport.

[0011] Preferably, the liquid transmission channel includes a sensing (SP) hole and a perspiration (DP) hole, the sensing (SP) hole corresponding to the position of the biochemical sensing electrode, and the diameter of the perspiration (DP) hole being smaller than that of the sensing (SP) hole.

[0012] Preferably, the biochemical sensing electrode includes an ISE, an enzyme biosensing electrode, and a MIP electrode;

[0013] Preferably, ISE includes the detection of Na + K + and NH4 + One or more of the following electrodes; enzyme biosensing electrodes include electrodes for detecting one or more of glucose, lactate, and urea; MIP electrodes include electrodes for detecting cortisol and / or testosterone.

[0014] Preferably, the micropillars in the PDMS micropillar layer are arranged radially outward from the center, with the height of the outer micropillars being higher than that of the inner micropillars, and the circumferential spacing of the outer micropillars being greater than that of the inner micropillars.

[0015] Preferably, the absorbent pad layer comprises absorbent paper and PET fabric, the PDMS micropillar layer is the annular PET fabric, and the absorbent paper is disposed on the side of the PDMS micropillar layer without micropillars and is bonded to the annular PET fabric.

[0016] Preferably, the mobile terminal APP is deployed with an AI evaluation module based on a cloud platform, including a data preprocessing unit, a multi-algorithm fusion model, and an intervention plan generation unit.

[0017] The method for monitoring and managing exercise fatigue using an exercise fatigue monitoring and management system includes the following steps;

[0018] S1: The electronic fabric patch is attached to the user's skin. During exercise, sweat is collected directly. During rest, the iontophoresis module is activated to stimulate sweat secretion and collect sweat.

[0019] S2: The data information collected by the multimodal sensing module is transmitted to the mobile terminal APP and uploaded to the cloud platform through the Bluetooth Low Energy communication unit. The preprocessing includes linear interpolation of missing values, Fourier transform of electromyography signals and feature standardization.

[0020] S3: The AI ​​assessment module of the cloud platform calls the optimized model to classify and assess six fatigue-related states: subjective fatigue (SF), blood glucose supply (BG), lactate accumulation (BL), hydration status (Hydration), muscle fatigue (MF), and protein level (Protein), and calculate the comprehensive fatigue index.

[0021] Preferably, the detection response time of the multimodal sensing module is: Na + / K + / NH4 + The data for glucose / lactic acid / urea ratio was 6 seconds, electromyography (MEF / MDF) ratio was 1 minute, and cortisol / testosterone ratio was 10 minutes. Linear interpolation was used for data alignment.

[0022] Preferably, the comprehensive fatigue index is calculated in the following manner;

[0023] Overall Fatigue Index = Score BG ×1 / 6 + Score BL ×1 / 6 +Score Protein ×1 / 6 +Score Hydration ×1 / 6 +Score SF ×1 / 6+Score MF ×1 / 6

[0024] Among them, Score BG Score for blood glucose supply BL Score for lactic acid accumulation Protein Score for protein levels Hydration Score for hydration status SF Score is used to rate subjective fatigue. MF Scoring muscle fatigue.

[0025] The advantages and positive effects of this invention are as follows: Through sweat adaptive regulation technology, it achieves rapid drainage of excess sweat during exercise (at a rate 2533 times that of the human physiological limit) and efficient enrichment of trace amounts of sweat during rest, resulting in a decrease in the electromyographic signal-to-noise ratio of only 5.74%, which is 4.53 times higher than that of commercial electrodes; the patch has good breathability, mechanical stability (stable performance after 100 bends / rubbings) and biocompatibility, can be adapted to different parts of the limb, and supports 24-hour continuous monitoring;

[0026] This system can simultaneously collect multidimensional physiological and biochemical indicators, covering four dimensions: electromyography, electrolytes, metabolites, and hormones. It can stably monitor for up to 60 minutes even under extremely low sweat flow conditions. Furthermore, it uses an AI model to accurately identify six types of fatigue states. This system solves the problems of unstable signals, lack of multidimensional monitoring, and disconnect between assessment and intervention in existing devices across the entire exercise-rest scenario. The clinical consistency rate reaches 86.51%, and personalized intervention plans can reduce the degree of fatigue by 51.82% and accelerate the recovery of chronic fatigue by 24 hours. It is suitable for monitoring athletes' training and managing the exercise of rehabilitation patients. Attached Figure Description

[0027] Figure 1 Schematic diagram of physiological and biochemical electrode structures fabricated by printing on the surface of hydrophobic cotton fabric; 11. Biochemical sensing electrode, 12. Electromyographic sensing electrode;

[0028] Figure 2 Schematic diagram of selective plasma treatment of electronic textile patch; 13, electronic textile fabric layer; 14, PET tape with specific pattern; 15, DP die hole; 16, SP die hole; 17, PET tape without pattern.

[0029] Figure 3 Schematic diagram of PDMS gradient micropillar design;

[0030] Figure 4 Schematic diagram of the layered structure of the e-ASRHT patch; 2. PDMS micropillar layer, 3. PET fabric, 4. Absorbent paper, 51. Anode hydrogel, 52. Cathode hydrogel.

[0031] Figure 5 Sensitivity characterization of biochemical sensors in e-ASRHT patches; (a) Na + (b) K + (c) NH4 + (d) Urea, (e) Glucose, (f) Lactic acid, (g) Cortisol, (h) Testosterone;

[0032] Figure 6 Schematic diagram of an integrated circuit module;

[0033] Figure 7AI model training and optimization flowchart;

[0034] Figure 8 Images and wearing instructions for the e-ASRHT patch. Detailed Implementation

[0035] The embodiments of the present invention will now be described with reference to the accompanying drawings.

[0036] This invention relates to a sports fatigue monitoring and management system and a sports fatigue monitoring and management method. Firstly, it provides a sports fatigue monitoring and management system based on an adaptive sweat-regulating electronic fabric. This system includes an electronic fabric patch, a multimodal sensing module, and a data transmission and interaction module. The electronic fabric patch can be attached to the skin surface of a subject. The electrode structure within it forms the multimodal sensing module, which can analyze the composition of flowing sweat and collect electromyographic signals. The multimodal sensing module is electrically connected to and interacts with the data transmission and interaction module. Data collected by the multimodal sensing module can be uploaded to a cloud platform via the data transmission and interaction module for data analysis and processing. Similarly, the multimodal sensing module can also receive instructions from the data transmission and interaction module. The sports fatigue monitoring and management system can adaptively regulate sweat state, simultaneously monitor multidimensional physiological and biochemical signals, and, combined with artificial intelligence, achieve accurate assessment and personalized intervention for sports fatigue.

[0037] The adaptive sweat-regulating multimodal electronic fabric patch (e-ASRHT) is formed by bonding an ion-electroosmotic hydrogel layer, an electronic textile fabric layer, a ring-shaped PET absorbent pad layer, a PDMS micropillar layer, and a circular absorbent pad layer. The electronic textile fabric layer is a Janus porous fabric layer, where cotton fabric is modified with superhydrophobic materials to form a hydrophobic material. Specifically, perfluorosilane-functionalized TiO2 nanoparticles can be used to modify hydrophilic cotton fabric, forming a superhydrophobic substrate with a contact angle of 156°. Then, a patterned mask is used to selectively plasma-treat the side of the hydrophobic electronic fabric away from the skin, forming liquid transport channels on the cotton fabric surface. To flexibly address different sweat detection scenarios, the liquid transport channels include sensing (SP) pores and perspiration (DP) pores. The SP pores correspond to the positions of the biochemical sensing electrodes, and the DP pores have a smaller diameter than the SP pores. In some embodiments of the present invention, SP holes with a diameter of 1.6 mm and DP holes with a diameter of 0.8 mm are formed. Utilizing the dynamic difference in breakthrough pressure between the two, SP preferentially enriches sweat at low sweat flow rates, while DP and SP work together to drain sweat at high flow rates. Testing shows that the sweat transfer rate reaches 7.22 mL / cm². -2 min -1 This is three orders of magnitude higher than the physiological limits of the human body.

[0038] Conductive electrode points are fabricated on the electronic textile fabric layer. Specifically, this can be achieved by 3D printing self-made silver nanosheets (AgFKs) / gallium indium liquid metal hybrid conductive ink (ECI), which is then mechanically stretched to form horizontally low-diffusion, vertically permeable conductive electrode points with a diameter of approximately 1.0 mm. A serpentine electrode and an arc-shaped dual electrode are printed on the side of the electronic textile fabric layer closest to the skin to form electromyography (EMG) sensing electrodes and ion sweat induction electrodes; a biochemical sensing electrode is printed on the other side to prepare biochemical index detection sensing electrodes. In some embodiments of this method, the electrode points can be printed on the electronic textile fabric layer first, followed by regional plasma treatment; alternatively, regional plasma treatment can be performed first, followed by electrode point printing. The electrode point printing positions are matched with the patterned mask shape so that the SP holes correspond to the electrode point positions.

[0039] The multimodal sensing module includes electromyography (EMG) sensing electrodes and biochemical sensing electrodes. The biochemical sensing electrodes utilize ion-selective permeable membranes, enzyme-containing mixtures, or molecularly imprinted polymers to cover the conductive electrode points away from the skin, thereby forming biochemical sensing electrodes capable of detecting different targets. Specifically, the biochemical sensing electrodes include ion-selective electrodes (ISE), enzyme biosensing electrodes, and molecularly imprinted polymer (MIP) sensing electrodes; the ISE electrode is used to detect Na+ in sweat. + K + and NH4 + Enzyme biosensing electrodes are used to detect glucose, lactic acid, and urea in sweat; molecularly imprinted polymer (MIP) sensing electrodes are used to detect cortisol and testosterone. Pre-concentration calibration is performed before use, and a multiple regression model is established using temperature and sodium ion sensor data to dynamically calibrate the detection results for glucose, lactic acid, cortisol, and testosterone. The serpentine mesh electrodes printed on the skin side of the electronic fabric patch are EMG electrodes used to detect electromyographic signals for dynamic monitoring of muscle fatigue. Iontoosmotic electrodes connected to a carbacholine-loaded conductive hydrogel form an iontoosmotic module, which promotes the release of sweat-inducing agents from the hydrogel, allowing the user to secrete sufficient sweat even at rest. The iontoosmotic electrodes use a low-current stimulation of 50 μA, combined with the carbacholine-loaded conductive hydrogel, to achieve highly efficient sweat induction. The excellent self-adhesive properties of the hydrogel ensure stable skin adhesion. When the user is exercising, the system directly collects naturally excreted sweat. When the user is at rest, the iontophoresis module is activated, controlling the iontophoresis electrodes to output a 50 μA pulsed current at a frequency of 1 Hz for 60 seconds to promote the release of carbacholine loaded in the conductive hydrogel, thereby inducing sweat secretion. During iontophoresis stimulation, the system monitors sweat flow rate and biochemical sensor output signals in real time; this module ensures that the user secretes sufficient sweat for accurate monitoring even at rest. Furthermore, the excellent self-adhesive properties of the hydrogel ensure stable skin adhesion.

[0040] This exercise-induced fatigue monitoring and management system can be used to monitor the wearer throughout the entire exercise cycle, including detecting fatigue during exercise, collecting data on fatigue levels before exercise, and collecting data on fatigue levels during the recovery process after exercise. This resting-state monitoring requires the iontophoresis module. Pre-exercise monitoring can determine the subject's baseline level, while post-exercise monitoring can determine fatigue recovery. For example, when monitoring is needed 30 minutes before exercise, 1 hour after exercise, 6 hours after exercise, or longer, the iontophoresis module needs to be activated to promote sweating. During exercise, since sweat secretion is sufficient, the iontophoresis module does not need to be activated.

[0041] The electronic fabric patch also includes a highly efficient sweat-wicking structure, specifically a PDMS micropillar layer and a double-layer absorbent pad. The PDMS micropillar layer features multiple rows of gradient micropillars, arranged radially outwards from the center. The outer micropillars are taller than the inner ones, and the circumferential spacing between the outer micropillars is smaller than that between the inner ones. This gradient arrangement further enhances sweat wicking efficiency. The double-layer absorbent pad consists of circular absorbent paper and a ring-shaped PET fabric. The ring-shaped PET fabric surrounds the PDMS micropillar layer. The circular absorbent paper is positioned on the side of the PDMS micropillar layer without micropillars and is bonded to the ring-shaped PET fabric. The circular absorbent paper is detachably bonded to both the PDMS micropillar layer and the ring-shaped PET fabric. The PDMS gradient micropillar layer features a gradient distribution of micropillar height and spacing. Its surface is modified with a PVA / SiO2 composite coating to form a long-lasting hydrophilic layer. The PDMS micropillar layer, combined with a ring-shaped PET fabric and circular absorbent paper, constitutes a directional sweat transport channel. The spatial arrangement of the micropillars allows sweat to be transported from the center outwards. The PET fabric further provides capillary force, allowing sweat to enter the ring-shaped PET fabric from within the micropillar area. The ring-shaped PET fabric has a larger pore size than the circular absorbent paper; based on capillary action, liquid spontaneously travels from the ring-shaped PET fabric to the ring-shaped absorbent paper. Through the rational combination of the PDMS micropillar layer, the ring-shaped PET fabric, and the circular absorbent paper, rapid directional sweat transport is achieved. In particular, the gradient-set micropillars significantly improve the sweat transport rate; tests showed a 2.2-fold increase in liquid transport rate compared to PDMS columns of uniform height. This high liquid transport efficiency enables directional sweat transport and rapid replenishment, avoiding sample mixing. Furthermore, the replaceable circular absorbent pad extends the product's lifespan.

[0042] An electronic fabric patch (e-ASRHT) was formed by assembling an iontophoresis hydrogel layer, an electronic textile fabric layer, a ring-shaped PET fabric layer, a PDMS micropillar layer, and a circular absorbent paper. Testing showed that the e-ASRHT patch achieved a sweat transfer rate of up to 7.22 mL / cm². -2 min -1This surpasses the physiological limits of the human body by three orders of magnitude, with a signal-to-noise ratio (SNR) decrease of only 5.74% after 30 minutes of exercise, significantly better than commercially available electromyography electrodes. This is also true when sweat flow is < 4.0 × 10⁻⁶. -4 mL cm -2 min -1 It can still be stably monitored for 60 minutes under the conditions.

[0043] The data transmission and interaction module includes a Bluetooth Low Energy communication unit, a cloud storage unit, a mobile terminal APP, and an integrated circuit module. The integrated circuit module connects to and interacts with the multimodal sensing module, enabling real-time data transmission, visualization, and intervention command push. The mobile terminal APP, deployed on a cloud platform, includes an AI assessment module, comprising a data preprocessing unit and a multi-algorithm fusion model. The multi-algorithm fusion model, built based on algorithms such as random forest and extreme tree, is used to identify and assess six types of fatigue states, including subjective fatigue, blood glucose levels, and lactic acid accumulation. Furthermore, it can rate the output data based on different types of fatigue states and generate intervention plans based on the ratings. An intervention plan generation unit is built within the AI ​​assessment module to guide the wearer in improving their fatigue state through behavioral compensation.

[0044] The AI ​​fatigue assessment module model was pre-trained. Training data included all observations from 10 subjects during a 30-minute baseline, 120-minute exercise, and a 72-hour recovery period. Model input parameters included eight types of sweat biomarkers (Na). + K + NH4 + The study used data including concentrations of glucose, lactate, urea, cortisol, and testosterone, electromyographic parameters (mean frequency MEF / median frequency MDF), and subject weight and gender. Based on clinical data from urine specific gravity, blood glucose, blood lactate, salivary cortisol / testosterone ratio, and salivary urea nitrogen, the study categorized each fatigue label and established threshold values ​​for grades 0, 1, and 2.

[0045] The raw multidimensional acquisition data first underwent standardized preprocessing: linear interpolation was used to fill in missing data items, and the raw electromyography signals were filtered and then subjected to Fourier transform to extract MEF and MDF feature parameters. After normalization, the preprocessed data was divided into training and test sets in an 8:2 ratio. This scheme uses a multi-machine learning algorithm fusion strategy to build an evaluation model. Based on the above-mentioned real-world test dataset of the entire exercise-recovery cycle of 10 subjects, algorithm optimization was performed. The extreme tree algorithm was selected for subjective fatigue assessment, and the random forest algorithm was selected for identification of 6 states, including blood glucose supply / lactate accumulation. The overall accuracy of the model reached 86.51%.

[0046] The wearable device of this invention relies on a low-power Bluetooth BLE unit to realize the real-time transmission of physiological and biochemical test data to the mobile terminal. The collected data is stored and processed online by the AI ​​model via the cloud platform, and outputs fatigue label classification results in various dimensions to realize real-time monitoring of exercise fatigue. The system combines the classification results to automatically generate the timing and dosage of carbohydrate, electrolyte and protein supplementation, as well as exercise intensity regulation instructions to form an individualized exercise intervention plan.

[0047] In some preferred embodiments: when the fatigue index is ≥1, the APP pops up a suggestion for intervention; when the hydration level is ≥1, it is recommended to drink electrolyte water (200ml each time); when the blood glucose level is ≥1, it is recommended to consume Gu energy gel (1 packet / time, containing 23g carbohydrates); when the protein level is ≥1, it is recommended to drink MYPROTEIN whey protein powder (1 bottle / time, containing 20g protein); when the muscle fatigue level is ≥1, it is recommended to reduce the exercise intensity (reduce cycling resistance by 20%) or rest for 5-10 minutes.

[0048] After pushing out intervention plans, the system continuously collects physiological and biochemical data and updates the comprehensive fatigue index. If the comprehensive fatigue index decreases by more than 20%, the current intervention level is maintained; otherwise, the intervention level is upgraded. This constructs a closed-loop fatigue management system encompassing data collection, status assessment, precise intervention, and effect feedback. The e-ASRHT patch connects to the mobile terminal in real time via a Bluetooth Low Energy (BLE) communication unit. The cloud platform handles data storage and AI model computation, while the accompanying mobile app provides visualization of fatigue status, intervention plan push notifications, and historical data query functions.

[0049] The method for monitoring and managing exercise fatigue using a system includes the following steps;

[0050] S1: The electronic fabric patch is attached to the user's skin. During movement, sweat is collected directly. At rest, the iontophoresis module is activated to stimulate sweat secretion and collect the sweat. The multimodal sensing module is activated to collect real-time data. The detection response time of the multimodal sensing module is: Na + / K + / NH4 + The data were aligned using linear interpolation for the following parameters: glucose / lactic acid / urea (6 s), electromyographic signal characteristics (MEF, MDF) (1 min), and cortisol / testosterone (10 min).

[0051] S2: The data information collected by the multimodal sensing module is transmitted to the mobile terminal APP and uploaded to the cloud platform through the Bluetooth Low Energy communication unit. The preprocessing includes linear interpolation of missing values, Fourier transform of electromyography signals and feature standardization.

[0052] S3: The AI ​​assessment module of the cloud platform calls the optimized model to classify and assess six fatigue-related states: subjective fatigue (SF), blood glucose supply (BG), lactate accumulation (BL), hydration status (Hydration), muscle fatigue (MF), and protein level (Protein), and calculate the comprehensive fatigue index; the comprehensive fatigue index is calculated in the following manner.

[0053] Overall Fatigue Index = Score BG ×1 / 6 + Score BL ×1 / 6 +Score Protein ×1 / 6 +Score Hydration ×1 / 6 +Score SF ×1 / 6+Score MF ×1 / 6

[0054] Among them, Score BG Score for blood glucose supply BL Score for lactic acid accumulation Protein Score for protein levels Hydration Score for hydration status SF Score is used to rate subjective fatigue. MF Scoring muscle fatigue.

[0055] By monitoring exercise fatigue through the above process, personalized intervention plans can be generated based on the assessment results. These plans include the timing and dosage of carbohydrate, electrolyte, and protein supplementation, as well as suggestions for adjusting exercise intensity, and are pushed to users via a mobile app. The personalized intervention plan uses SHAP analysis to identify key influencing factors and generates suggestions on the timing and dosage of carbohydrate, electrolyte, and protein supplementation, as well as adjustments to exercise intensity, based on fatigue level scores. Specifically, when the fatigue index is ≥1, the app pops up with intervention suggestions; when the hydration level is ≥1, it is recommended to drink electrolyte water (200ml each time); when the blood glucose level is ≥1, it is recommended to consume Gu energy gel (1 packet / time, containing 23g carbohydrates); when the protein level is ≥1, it is recommended to drink MYPROTEIN whey protein powder (1 bottle / time, containing 20g protein); when the muscle fatigue level is ≥1, it is recommended to reduce exercise intensity (reduce cycling resistance by 20%) or rest for 5-10 minutes. After implementing the intervention plan, the system continuously collects physiological and biochemical data and recalculates the comprehensive fatigue index. If the comprehensive fatigue index decreases by more than 20%, the current intervention level is maintained; otherwise, the intervention measures are upgraded by one level, thus forming a closed-loop management of "monitoring-evaluation-intervention-feedback".

[0056] Continuous monitoring of changes in physiological and biochemical indicators after intervention, dynamic updating of assessment results and intervention plans, forming a closed-loop management system.

[0057] The exercise fatigue monitoring and management system based on adaptive sweat-regulated multimodal electronic fabric can solve the problems of unstable signals, lack of multidimensional monitoring, and disconnect between assessment and intervention in the entire exercise-rest scenario of existing equipment. It has a clinical consistency of 86.51%, can reduce the degree of fatigue by 51.82%, and accelerate the recovery of chronic fatigue. It is suitable for monitoring athletes' training and managing the exercise of rehabilitation patients.

[0058] The present invention will now be described with reference to the accompanying drawings. Experimental methods not specifically described in terms of operation steps are performed in accordance with the corresponding product manuals. Unless otherwise specified, the instruments, reagents, and consumables used in the embodiments can be purchased from commercial companies.

[0059] Example 1: Fabrication of an Adaptive Sweat-Regulating Multimodal Electronic Fabric Patch (e-ASRHT)

[0060] 1.1 Preparation of electronic textile fabric layers

[0061] First, a superhydrophobic fabric is prepared. The fabric used in this embodiment is a woven cotton fabric, which is pretreated by traditional desizing, scouring, and bleaching processes before use. The pretreated cotton fabric is then immersed in a coating suspension to form a superhydrophobic coating. The coating suspension is prepared by adding 2g of PFOTES (C 14 H 19 F 13 TiO2 nanoparticles (97%) were dissolved in 198 g of anhydrous ethanol and stirred vigorously for 2 hours. Then, 10 g of Degussa P25 TiO2 nanoparticles were added to the mixed solution and stirred for 30 min to form a coating suspension. Pretreated cotton fabric cut to an appropriate size was immersed in the coating suspension for 5 min and dried at room temperature for 10 min to obtain double-sided superhydrophobic cotton fabric.

[0062] A conductive paste was prepared and then 3D printed onto the surface of a superhydrophobic fabric to fabricate electromyography (EMG) and biochemical sensing electrodes. The conductive paste consisted of liquid metal microspheres (LMMS), silver nanosheets (AgFKs, 10 nm), and a two-component elastomer, styrene-isoprene-styrene / vinyl acetate (SIS-EVA). The preparation of the conductive paste included the following steps: First, liquid metal microspheres (LMMS) were prepared; 4 g of eutectic gallium-indium alloy (EGaIn, 75% Ga, 25% In) was mixed with 20 mL of anhydrous ethanol and sonicated in an ice bath for 5 min using an ultrasonic cell disruption system (Biosafer 650-92, Nanjing Biosafer Co., Ltd.); after standing overnight, the supernatant solvent was decanted to obtain LMMS. A 35 wt% SIS-EVA (1:1) toluene solution was added to a sample vial containing LMMS, and the mixture was vortexed (Scientific Industries Vortex-Genie 2) for 10 min. AgFKs were then added and vortexed for another 20 min to obtain a mixed liquid metal conductive paste. Figure 1 As shown, a biochemical sensing electrode 11 was printed on one side of a superhydrophobic cotton fabric using a 3D printer, and an electromyography sensing electrode 12 and an ion sweat induction electrode were printed on the other side. The fabric was then cured at 100°C for 30 min.

[0063] The electronic textile fabric layer 13 is subjected to localized plasma treatment. For example... Figure 2 As shown, a PET tape 14 with a specific pattern was prepared. The size of the PET tape matched the shape and size of the pretreated cotton fabric of the prototype. The pattern included DP mold holes 15 and SP mold holes 16. The SP mold holes were located in the center, and their number and position corresponded to the sensing electrode points of the biochemical electrode. The hole diameter was 1.6 mm. The DP mold holes were distributed throughout the area of ​​the PET tape excluding the SP mold holes, and the hole diameter was 0.8 mm. The spacing between adjacent holes on the PFT tape was 3.0 mm. A PET tape 17 without a pattern was also prepared, with its shape and size matching the pretreated cotton fabric. The PET tape with the specific pattern was attached to the side of the circular cotton fabric printed with the biochemical sensing electrode, with the biochemical sensing electrode points precisely aligned with the center of the SP mold holes in the PET tape. The PET tape without a pattern was attached to the other side of the circular cotton fabric. The electronic fabric covered with the tape was placed in an oxygen plasma etching machine (PE100RIE, Plasma Etch Inc.) and etched at 50 cm⁻¹ for a set time. 3 The process involves an O2 flow rate of / min and a power of 300 W. Due to the patterned tape covering the area, only the perforated regions are exposed to the plasma environment, giving them hydrophilic properties and forming channels that can unidirectionally transport sweat. The 1.6 mm and 0.8 mm perforations on the tape form SP and DP pores on the hydrophobic cotton fabric, respectively.

[0064] The above steps are used to prepare a hydrophobic cotton fabric with printed physiological and biochemical electrodes and formed sweat transmission pores DP and SP, which is the electronic textile fabric layer.

[0065] 1.2 Preparation of PDMS micropillar layers

[0066] The micropillars in the circular PDMS micropillar layer are arranged in a gradient pattern. The PDMS template was designed using AutoCAD and fabricated using photolithography. Figure 3 As shown, the PDMS micropillar layer has an array of n concentric rings of micropillars arranged radially from the central center. Each micropillar has a diameter of 200 μm, and the spacing between adjacent rings is 100 μm. The outermost ring has a spacing of 200 μm between each micropillar, and the spacing between the nth ring from the outside in is 200 + 20(n-1) μm. The micropillars at the center are the shortest, with the innermost ring having a height of 50 μm, and the height of the nth ring from the inside out is 50 + 20(n-1) μm. The PDMS template undergoes OTS vapor silanization treatment to facilitate the release of the PDMS micropillars. PDMS (matrix to curing agent ratio of 10:1, Sylgard 184, Dow Corning) is poured into a mold, degassed under vacuum for 10 min, and then baked at 80°C for 30 min before demolding. During use, the side with the micropillars faces the electronic textile fabric layer.

[0067] A hydrophilic coating was prepared by dispersing 0.1 g SiO2 and 0.05 g sodium dodecyl sulfate in 10 mL of a 1.0 wt% polyvinyl alcohol solution and sonicating for 10 min. The hydrophilic coating solution was then sprayed onto plasma-treated PDMS micropillars and heated in an oven at 110°C for 10 min to form a durable superhydrophilic layer on the surface of the PDMS micropillars. The PDMS micropillar layer was thus constructed.

[0068] 1.3 Preparation of the hydrogel layer

[0069] 4.0 mL of acrylic acid was added to 7.0 mL of 10 wt% glucose aqueous solution, followed by the addition of 0.3 g of chitosan and stirring to form a homogeneous dispersion (solution A). 0.1 g of boric acid was dissolved in 3.0 mL of deionized water to form solution B. 0.005 g of N,N'-methylenebisacrylamide and solution B were added to solution A and sonicated for 10 min. Subsequently, 0.015 g of ammonium persulfate was added and thoroughly mixed with rapid stirring; then 3.0 wt% carbachol was added, the mixture was poured into a mold, and heated at 55 °C for 3.0 h to form a carbachol anodic hydrogel.

[0070] The preparation process for the cathode hydrogel is the same, except that carbacholine is replaced with NaCl.

[0071] 1.4 Assembly of e-ASRHT patch

[0072] Take PET fabric and absorbent paper. The hydrophilic (plasma-treated) PET fabric is laser-cut into a ring shape, with the inner side capable of embedding the PDMS micropillar layer 2. The absorbent paper 4 is cut into a circle to match the shape of the electronic textile fabric layer 13, with the outer edge of the ring-shaped PET fabric 3 aligned with the edge of the absorbent paper 4. For example... Figure 4 As shown, absorbent paper 4, PDMS micropillar layer 2, annular PET fabric 3, electronic textile fabric layer 13, and hydrogel layer are sequentially bonded together and glued together using Sil-Poxy silicone adhesive. The anode hydrogel 51 and cathode hydrogel 52 can adhere to and connect to the ion sweat induction electrode; finally, the e-ASRHT patch is assembled.

[0073] Example 2: Fabrication of a multimodal sensing module on an e-ASRHT patch

[0074] Multimodal sensing modules, including EMG sensing electrodes, ISE, and MIP sensing electrodes, were fabricated on biochemical electrodes in electronic textile fabric layers using 3D printing.

[0075] 2.1 Fabrication of electromyography sensing electrodes

[0076] Conductive paste was used to print electromyography (EMG) sensing electrodes 12 on the surface of superhydrophobic fabric using 3D printing technology. The EMG sensing electrodes 12 were designed with a serpentine mesh structure to improve structural stability and durability under bending deformation. They are used to collect EMG signals and obtain MDF and MEF parameters for muscle fatigue assessment.

[0077] 2.1 Preparation of the reference electrode

[0078] A shared reference electrode was prepared by drop-casting 2.5 μL of 0.1M FeCl3 onto the surface of a 3D-printed SP electrode for 30 s, followed by rinsing with deionized water to obtain an Ag / AgCl reference electrode. Then, 1.0 μL of a polyvinyl butyral (PVB) reference mixture (prepared by dissolving 79.1 mg PVB and 50 mg NaCl in 1 mL methanol) was drop-coated onto the Ag / AgCl surface and dried overnight to obtain the reference electrode.

[0079] 2.3 Preparation of ion-selective electrodes

[0080] Sodium selective permeation membrane (Na) +The preparation method of the -ISM) solution is as follows: 1 mg of sodium ion carrier X, 0.55 mg of sodium tetrakis(3,5-bis(trifluoromethyl)phenyl)borate (NaTFPB), 33 mg of polyvinyl chloride (PVC) and 65.45 mg of diisooctyl sebacate (DOS) are dissolved in 660 μL of tetrahydrofuran (THF), and then continuously stirred to ensure thorough mixing.

[0081] K + The -ISM ​​solution was prepared as follows: 2 mg valproic acid, 0.5 mg NaTFPB, 32.7 mg PVC and 64.7 mg DOS were dissolved in 660 μL THF.

[0082] NH4 + The ISM solution was prepared as follows: 1.0 mg of non-actin, 0.14 mg of potassium tetra(4-chlorophenyl)borate (KTpClPB), 30 mg of PVC and 65 mg of DOS were dissolved in 660 μL of THF.

[0083] 1.0 μL Na + -ISM solution, 1.0 μL K + -ISM solution and 1.0 μL NH4 + -ISM was drop-cast onto electrode sites and dried overnight to obtain samples capable of detecting Na. + / K + / NH4 + ISE.

[0084] 2.4 Preparation of enzyme biosensing electrodes

[0085] The enzyme biosensor was prepared using an electrochemical workstation (CHI 760E, CH Instruments). It is specifically used for the detection of glucose, lactic acid, and / or urea.

[0086] AuNPs were deposited in 1.0 mM HAuCl4 (0.1 M KCl) by cyclic voltammetry within a voltage range of 0.2–1.0 V at a scan rate of 50 mV / s for 10 cycles to increase surface area and improve sensitivity. Prussian blue (PB) transducer layers were deposited by cyclic voltammetry in freshly prepared solutions containing 2.5 mM MFeCl3, 2.5 mM K3Fe4(CN)6, 100 mM KCl, and 100 mM HCl. Electrode deposition for glucose detection was performed for 10 cycles, and electrode deposition for lactate detection was performed for 20 cycles. The deposition was achieved by adding 10 μL of 50% glutaraldehyde and 10 μL of enzyme solution (1000 U / mL). -1To prepare the enzyme mixture solution, add 1.0 mL of 1% BSA solution (w / v). Then, drop 0.5 μL of the enzyme mixture onto the surface of the enzyme sensing electrode and dry at 4 °C overnight.

[0087] To improve the linear response range and sensitivity of the lactic acid sensor, 1.0 μL of polyurethane solution (20 mg / mL) was dropped onto the surface of the lactic acid electrode. -1 Dissolved in THF) as a diffusion-limiting membrane. To prepare a urea potential sensor, 0.5 μL of a solution containing urease (100 U / mL) was used. -1 The enzyme mixture solution was added dropwise to the NH4 prepared above. + The surface of the potential sensor was prepared and placed overnight at 4°C. Finally, Nafion solution (2.0% v / v) was dropped onto its surface to improve detection stability.

[0088] 2.5 Fabrication of Molecularly Imprinted Polymer (MIP) Sensing Electrodes

[0089] The MIP electrode for detecting cortisol or testosterone was prepared by electropolymerization in a PBS (pH = 7.4) deposition solution containing 2.5 mM FeCl3, 5 mM K3Fe4(CN)6, 5 mM KCl, 100 mM HCl, and 6 mM cortisol template via cyclic voltammetry at a scan rate of 50 mV / s (10 cycles). Following electropolymerization, the cortisol template was removed from the MIP matrix by peroxidation in PBS solution for 15 cycles at a scan rate of 50 mV / s within the range of -0.2 to 0.8 V, yielding a polymer membrane for cortisol detection. For the preparation of the testosterone molecularly imprinted polymer membrane, a similar scheme was used to synthesize the testosterone MIP by replacing the cortisol template with a testosterone template.

[0090] As a control, non-indented polymers (NIPs) were prepared and deposited using the same procedure, but without template molecules in the deposition solution.

[0091] 2.6 Characterization and Verification of Multimodal Sensing Module

[0092] All in vitro biosensor characterizations were performed using open-circuit potential and amperometric methods on a multichannel electrochemical workstation (CHI 1430, CH Instruments).

[0093] For ISE detection and characterization; preparations targeting Na+ were prepared in deionized water. + / K + / NH4 +The analyte solutions used for electrode detection included NaCl solutions with concentrations ranging from 0.1 to 500 mM, KCl solutions with concentrations ranging from 0.5 to 20 mM, and NH₄Cl solutions with concentrations ranging from 0.05 to 10 mM. The ISE sensor was characterized using the open-circuit potential method.

[0094] For the detection and characterization of the enzyme biosensor electrode, analyte solutions for glucose, lactate, and urea were prepared in PBS, with concentrations ranging from 5–250 μM for glucose, 1–50 mM for lactate, and 0.05–50 mM for urea. The enzyme biosensor was evaluated using a chronoamperometric method at an applied potential of 0 V.

[0095] For the detection and characterization of the MIP sensing electrode, analyte solutions for cortisol and testosterone were prepared, with cortisol concentrations ranging from 0 to 1000 μg / L. -1 The concentration range of the testosterone solution is 0-500 ng / L. -1 The MIP sensor was evaluated using the chronoamperometry method at an applied potential of 0 V.

[0096] like Figure 5 The figure shows the detection results of each sensing electrode in the e-ASRHT patch. As can be seen from the data in the figure, each sensing electrode has good sensitivity.

[0097] Example 3: AI Evaluation Module Training and Deployment

[0098] An e-ASRHT patch was assembled using the method described in Example 1. Multimodal sensing modules were then fabricated on the biochemical electrode surface of the e-ASRHT patch according to the steps in Example 2. The multimodal sensing modules were connected to a data transmission and interaction module. The data transmission and interaction module includes a Bluetooth Low Energy (BLE) communication unit, a cloud storage unit, and a mobile terminal APP, as well as necessary integrated circuit modules. The structure of the integrated circuit module is as follows: Figure 6 As shown.

[0099] The integrated circuit module includes a front-end analog circuit, a conversion circuit, a microcontroller circuit, and a power management circuit. The front-end analog circuit, conversion circuit, and microcontroller circuit are connected sequentially, and the power management circuit is electrically connected to these circuits. A multimodal sensing module is connected to the front-end analog circuit. An iontophoresis electrode is connected to the microcontroller circuit, and the power management circuit is connected to the iontophoresis circuit to supply power. The iontophoresis electrode is connected to the hydrogel layer. The microcontroller circuit connects to a Bluetooth Low Energy (BLE) communication unit and a cloud storage unit. The BLE communication unit enables data exchange with a mobile terminal app, allowing users to view data in real time and perform in-depth data analysis on their mobile devices.

[0100] The mobile app has an AI evaluation module deployed on it, and this module is used for training. For example... Figure 7 As shown, it includes the following steps:

[0101] 3.1 Data Acquisition

[0102] Subject Recruitment and Grouping: Ten healthy subjects (6 males and 4 females, aged 23-30 years, weighing 55-75 kg, with no history of metabolic diseases) were recruited and informed consent was obtained.

[0103] Experimental Procedure Design: First, subjects wore e-ASRHT patches, and the iontophoresis module (50 μA current) was activated to induce sweating. Baseline physiological and biochemical data were collected after 30 minutes of sitting still. Subsequently, subjects performed moderate-intensity cycling using a Keep K0102C fitness machine in a constant temperature and humidity environment (21-32℃, RH 29-35%), maintaining a heart rate of 70-80% of their maximum heart rate for 120 minutes. Subjective fatigue levels (Borg scale) were recorded every 10 minutes. e-ASRHT patch data acquisition parameters: Na + / K + / NH4 + The sampling frequency of glucose, lactate, and urea sensors was 6 s / time, EMG was 1000 Hz, and cortisol / testosterone was 10 min / time. After exercise, physiological and biochemical data were collected at 0.5 h, 24 h, 48 h, and 72 h.

[0104] During the sitting and exercise processes, physiological and biochemical data were collected simultaneously via the e-ASRHT patch, along with blood (finger-prick blood), saliva, and urine samples for label calibration. Blood glucose was measured using a Roche Accu-Chek blood glucose meter, lactate was measured using an EKF Lactate Scout 4 analyzer, salivary cortisol / testosterone was measured using an enzyme-linked immunosorbent assay kit (Nanjing Jiancheng), salivary urea nitrogen was measured using a urea nitrogen test kit / spectrophotometry (Nanjing Jiancheng), and urine specific gravity (USG) was measured using a refractometer.

[0105] 3.2 Data Preprocessing Steps

[0106] Raw data cleaning: Remove obvious outliers (exceeding the mean ± 3 standard deviations), and fill in missing data using linear interpolation (interpolation error ≤ 5%).

[0107] EMG signal processing: Baseline drift was eliminated using a polynomial fitting method; noise was removed using a 5-500Hz bandpass filter; after Fourier transform, MEF and MDF were calculated;

[0108] Dataset partitioning: The training set and the test set are randomly partitioned in an 8:2 ratio, and stratified sampling is used to ensure that the sample proportions for each fatigue level are consistent.

[0109] 3.3 Model Training and Optimization

[0110] Model selection: Twelve machine learning algorithms were selected, including Extra Trees, Random Forest, LightGBM, Decision Tree, K-nearest Neighbor, Multi-layer Perceptron, Support Vector Machine, Logistic Regression, Gaussian Naive Bayes, SGD Classifier, XGBoost, and AdaBoost. Accuracy, precision, recall, and F1 score were used as evaluation metrics to select the model with the best overall performance. The fatigue index was comprehensively evaluated using subjective fatigue, blood glucose supply, lactate accumulation, hydration status, muscle fatigue, and protein levels as key assessment factors.

[0111] Based on the body fluid test results, each judgment element is classified. Data collected by the e-ASRHT patch is input into each model, and the output results are correlated with the classification of the judgment elements to select the appropriate model for each element. The optimal models are subjective fatigue (XGboost), blood glucose supply (Extra Tree), lactate accumulation (K-nearest neighbor), hydration status (Extra Tree), muscle fatigue (Random forest), and protein level (Extra Tree).

[0112] SHAP analysis: The contribution of each feature to fatigue assessment was quantified by SHAP values, and glucose, cortisol, MDF and other key influencing factors were identified.

[0113] The AI ​​assessment module of the cloud platform calls the optimized model and uses the best machine learning model to classify and assess six fatigue-related states, namely subjective fatigue (SF), blood glucose supply (BG), lactate accumulation (BL), hydration status, muscle fatigue (MF), and protein level, based on real-time biochemical and physiological signal data obtained by wearable sensors. The comprehensive fatigue index is calculated, and then a personalized intervention plan is generated based on the assessment results.

[0114] The comprehensive fatigue index is calculated as follows;

[0115] Overall Fatigue Index = Score BG ×1 / 6 + Score BL×1 / 6 +Score Protein ×1 / 6 +Score Hydration ×1 / 6 +Score SF ×1 / 6+Score MF ×1 / 6

[0116] Among them, Score BG Score for blood glucose supply BL Score for lactic acid accumulation Protein Score for protein levels Hydration Score for hydration status SF Score is used to rate subjective fatigue. MF Scoring muscle fatigue.

[0117] 3.4 Cloud Platform and App Deployment

[0118] Cloud platform setup: Alibaba Cloud ECS server (8 cores, 16GB memory) is used to deploy data storage module (MySQL database), model inference module (Docker container encapsulation) and API interface;

[0119] Data transmission protocol: The BLE module adopts the GATT protocol, with a transmission rate of 1Mbps, and data encryption uses the AES-128 algorithm;

[0120] APP Function Development: Develop a cross-platform APP based on the Flutter framework, including functions such as device connectivity, real-time data visualization (line chart), fatigue level prompts, intervention plan push notifications, and historical data queries, with a response latency of ≤1s.

[0121] The above method is used to train an evaluation model applicable to electronic fabric patches containing multimodal sensing modules, which can judge the wearer's physical condition based on the electrical signal information fed back by the electronic fabric patch.

[0122] Example 4: Application of a motion fatigue monitoring and management system based on adaptive sweat-regulated electronic fabrics

[0123] An e-ASRHT patch was assembled using the method described in Example 1. Multimodal sensing modules were then fabricated on the biochemical electrode surface of the e-ASRHT patch according to the steps in Example 2. The multimodal sensing modules were connected to a data transmission and interaction module, and the evaluation model was trained and deployed according to Example 3 to obtain a motion fatigue monitoring and management system (monitoring sensor).

[0124] 4.1 Wearing and Activation

[0125] Wipe the application area (front of the thigh or outer upper arm) with a 75% alcohol wipe. After the alcohol has completely evaporated, ensure the skin is dry and clean. Apply the e-ASRHT patch to the skin. Figure 8 As shown, secure the edges with medical tape (avoid obstructing the sensor area) to ensure close contact between the IP hydrogel and the skin (adhesion ≥ 0.55 MPa). Open the mobile app, identify and pair the device via Bluetooth, and synchronize the device ID and monitoring sensor status after successful pairing. When the user is at rest, activate the iontophoresis module to control the iontophoresis electrode to output a 50 μA pulse current at a frequency of 1 Hz for 60 seconds, promoting the release of carbacholine loaded in the conductive hydrogel, thereby inducing sweat secretion.

[0126] 4.2 Monitoring and Intervention Implementation

[0127] Real-time monitoring: Data transmitted from the multimodal sensing module is converted into various indicator detection values ​​via a mobile app; further analysis can be performed by combining these detection values, such as real-time display of multidimensional indicator curves, including Na. + K + Concentration, glucose concentration, MDF / MEF ratio, and comprehensive fatigue index (0-2 levels).

[0128] The AI ​​assessment module analyzes the obtained data and outputs a grading conclusion, providing suggestions and guidance based on the measured values. When the fatigue index is ≥1, the app pops up an intervention suggestion; when the hydration level is ≥1, it is recommended to drink electrolyte water (200ml each time); when the blood glucose level is ≥1, it is recommended to consume Gu energy gel (1 packet / time, containing 23g carbohydrates); when the protein level is ≥1, it is recommended to drink MYPROTEIN whey protein powder (1 bottle / time, containing 20g protein); when the muscle fatigue level is ≥1, it is recommended to reduce exercise intensity (reduce cycling resistance by 20%) or rest for 5-10 minutes.

[0129] Example 5: Overall Performance Verification of a Motion Fatigue Monitoring and Management System Based on Adaptive Sweat-Regulating Electronic Fabrics

[0130] Controlled trial design: Three subjects were selected for a self-controlled trial, with three intervention modes set up, each group 7 days apart; the control group (C) only wore the patch for monitoring, without any intervention suggestions; the self-perceived intervention group (SI) wore the patch for monitoring, and users intervened based on their own feelings; the systemic intervention group (MI) wore the patch for monitoring and strictly followed the intervention plan pushed by the APP. Key indicator detection: Fatigue relief rate: calculated by formula (fatigue index of control group - fatigue index of intervention group) / fatigue index of control group × 100%.

[0131] Verification Results: The fatigue relief effect was assessed using the comprehensive fatigue index. The fatigue relief rate in the MI group was 43.28%-51.82%, which was 1.82 times higher than that in the SI group (22.18%-31.35%). Recovery speed: The fatigue index in the MI group recovered to 97.94% of the baseline after 24 hours, which was 24 hours earlier than that in the SI group (75.48%). This demonstrates that the detection and intervention of the motion fatigue monitoring and management system based on adaptive sweat regulation electronic fabric can effectively alleviate fatigue.

[0132] The embodiments of the present invention have been described in detail above, but the content described is only a preferred embodiment of the present invention and should not be considered as limiting the scope of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the patent coverage of the present invention.

Claims

1. A sports-induced fatigue monitoring and management system, characterized in that: It can be used for full-cycle monitoring of the wearer before, during, and after exercise. include, The electronic fabric patch is formed by bonding an ion-electroosmotic hydrogel layer, an electronic textile fabric layer, a PDMS micropillar layer and an absorbent pad layer in sequence. The electronic textile fabric layer is printed with electromyographic sensing electrodes and biochemical sensing electrodes. The electronic textile fabric layer has a superhydrophobic substrate and also includes a unidirectional liquid transport channel formed after plasma treatment. The liquid transport channel includes sensing SP holes and wicking DP holes. The sensing SP holes are located in the center, and their number and position correspond to those of the biochemical sensing electrodes. The pore size is 1.6 mm. The wicking DP holes are distributed throughout the area outside the sensing SP holes, and their pore size is 0.8 mm. The spacing between adjacent holes is 3.0 mm. mm; The PDMS micropillar layer has multiple gradient micropillars, and the surface is hydrophilically modified with a PVA / SiO2 composite coating; The micropillars in the PDMS micropillar layer are arranged radially outward from the center, with the height of the outer micropillars being higher than that of the inner micropillars, and the circumferential spacing of the outer micropillars being smaller than that of the inner micropillars; The absorbent pad layer includes a circular absorbent paper and an annular PET fabric. The annular PET fabric surrounds the PDMS micropillar layer, and the circular absorbent paper is placed on the side of the PDMS micropillar layer without micropillars and is attached to the annular PET fabric. The outer edge of the annular PET fabric is aligned with the edge of the absorbent paper, and the pore size of the annular PET fabric is larger than that of the circular absorbent paper. The circular absorbent paper is detachably attached to the PDMS micropillar layer and the annular PET fabric; The multimodal sensing module includes an electromyography sensing electrode disposed on one side of the electronic textile fabric layer and a biochemical sensing electrode disposed on the other side of the electronic textile fabric layer. The data transmission and interaction module includes a Bluetooth Low Energy communication unit, a cloud storage unit, a mobile terminal APP, and an integrated circuit module. The integrated circuit module is connected to the multimodal sensing module and interacts with the multimodal sensing module.

2. The exercise fatigue monitoring and management system according to claim 1, characterized in that: Biochemical sensing electrodes include ion-selective electrodes (ISE), enzyme biosensing electrodes, and molecularly imprinted polymer (MIP) sensing electrodes.

3. The exercise fatigue monitoring and management system according to claim 2, characterized in that: The ISE includes the detection of Na + K + and NH4 + The enzyme biosensing electrode includes one or more of the following electrodes: the enzyme biosensing electrode includes an electrode for detecting one or more of glucose, lactic acid, and urea; the MIP sensing electrode includes an electrode for detecting cortisol and / or testosterone.

4. The exercise fatigue monitoring and management system according to any one of claims 1-3, characterized in that: The mobile terminal APP is deployed on a cloud platform with an AI assessment module, including a data preprocessing unit, a multi-algorithm fusion model, and an intervention plan generation unit.

5. A method for monitoring and managing exercise fatigue using the exercise fatigue monitoring and management system according to any one of claims 1-4, characterized in that: Includes the following steps; S1: The electronic fabric patch is attached to the user's skin. During exercise, sweat is collected directly. During rest, the iontophoresis module is activated to stimulate sweat secretion and collect sweat. S2: The data information collected by the multimodal sensing module is transmitted to the mobile terminal APP and uploaded to the cloud platform through the Bluetooth Low Energy communication unit. The preprocessing includes linear interpolation of missing values, Fourier transform of electromyography signals and feature standardization. S3: The AI ​​assessment module of the cloud platform calls the optimized model to classify and assess six fatigue-related states: subjective fatigue (SF), blood glucose supply (BG), lactate accumulation (BL), hydration status (Hydration), muscle fatigue (MF), and protein level (Protein), and calculate the comprehensive fatigue index.

6. The method for monitoring and managing exercise fatigue according to claim 5, characterized in that: The detection response time of the multimodal sensing module is: Na + / K + / NH4 + The data were aligned using linear interpolation for the following parameters: glucose / lactic acid / urea (6 s), electromyography (MEF / MDF) (1 min), and cortisol / testosterone (10 min).

7. The method for monitoring and managing exercise fatigue according to claim 5, characterized in that: The comprehensive fatigue index is calculated as follows; Overall Fatigue Index = Score BG ×1 / 6 + Score BL ×1 / 6 +Score Protein ×1 / 6 + Score Hydration ×1 / 6 +Score SF ×1 / 6+Score MF ×1 / 6 Among them, Score BG Score for blood glucose supply BL Score for lactic acid accumulation Protein Score for protein levels Hydration Score for hydration status SF Score is used to rate subjective fatigue. MF Scoring muscle fatigue.

Images