AIS patient self-correction system and method

Abstract

According to the AIS patient self-correction system and method provided by the invention, complete-cycle precise management of scoliosis is realized; through a 64-channel electrode array and a phase interference tactile feedback architecture, paravertebral muscle electrical activity can be captured in real time, subcutaneous precise stimulation can be realized, and the muscle imbalance detection sensitivity is obviously improved; according to the muscle imbalance index, a direct mathematical correlation between a muscle force difference value and a Cobb angle change is established, and the problem that biomechanical indexes and imaging parameters are disjointed in traditional evaluation is solved; based on a dynamic correction strategy of a PPO-Clip reinforcement learning algorithm, the training intensity can be automatically adjusted according to the real-time physiological status of the patient, so that the training compliance is further improved; the end-cloud cooperative processing architecture compresses key response delay to be extremely low, supports daily continuous pathological monitoring, realizes integrated closed-loop intervention, and remarkably reduces the development risk of the Cobb angle.

Description

Technical Field

[0001] This invention relates to the field of muscle quantitative assessment indicators and corrective training guidance for adolescent idiopathic scoliosis, and in particular to a self-correction system and method for AIS patients. Background Technology

[0002] Adolescent idiopathic scoliosis (AIS) is the most common three-dimensional spinal deformity, with a global incidence of 2-3%. Currently, clinical diagnosis mainly relies on Cobb angle measurement, but X-rays pose radiation exposure risks and cannot dynamically monitor changes in muscle function. Existing interventions include orthotic braces and physical therapy; however, traditional braces lack biofeedback mechanisms, leading to poor patient compliance (studies show that over 60% of patients do not meet the daily wearing time requirement).

[0003] In terms of muscle function assessment, although surface electromyography (sEMG) technology has been used to monitor paraspinal muscle activity, it faces three major technical bottlenecks: First, single-point EMG acquisition is difficult to capture the synergistic effect of muscles on both sides of the spine, and cannot quantify muscle imbalance on the convex and concave sides; Second, motion artifacts (such as signal drift during walking) cause distortion of home monitoring data (the literature reports a decrease in signal-to-noise ratio of more than 40%); Third, assessment and training are separated, and existing systems only provide static measurements and lack real-time biofeedback mechanisms.

[0004] Existing smart wearable solutions (such as the spinal posture monitoring garment in CN113729775A) integrate IMU sensors, but they have significant drawbacks: 1) They do not integrate biomechanical signals and cannot reflect muscle compensation mechanisms; 2) Data processing relies on the cloud, with latency exceeding 500ms, failing to meet real-time correction requirements; 3) Training guidance uses two-dimensional image prompts, resulting in spatial positioning errors >15°. More importantly, current technology lacks pathological progression prediction models, making it difficult to provide quantitative evidence for individualized intervention.

[0005] Therefore, there is an urgent need to develop an AIS self-correction system that integrates multimodal sensing, has real-time biofeedback, and can predict scoliosis progression, in order to address the three major clinical pain points of lack of dynamic assessment, low training compliance, and delayed intervention. Summary of the Invention

[0006] The purpose of this invention is to provide a self-correction system and method for AIS patients to solve the problems existing in the prior art.

[0007] To achieve the above objectives, the present invention provides the following solution:

[0008] This invention provides a self-correction method for AIS patients, comprising the following:

[0009] S1. The patient wears a smart wearable electromyography (EMG) testing device and performs a standard sequence of movements;

[0010] S2. Acquire 64-channel paravertebral electromyography (EMG) signals using an electrode array in a smart wearable electromyography testing device. i (t), the three-dimensional attitude angle of the spine [θ] is acquired by a 9-axis IMU. sag (t),θ cor (t)], the pressure P at the contact point of the brace is collected by a miniature pressure sensor. j (t);

[0011] S3. Execute the dynamic electromyography enhancement algorithm on the portable terminal. The formula is as follows:

[0012]

[0013] Where, Δθ=|θ cor (t)-θ ref |, k is the clinically calibrated curvature sensitivity coefficient;

[0014] S4. Calculate the muscle imbalance index based on enhanced electromyographic signals using a muscle imbalance quantification model. The formula is as follows:

[0015]

[0016] Where convex is the set of muscle group numbers on the convex side, concave is the set of muscle group numbers on the concave side, and f i The magnitude of the muscle force vector;

[0017] S5. When γ>γ thresh At that time, the haptic feedback device activates phase interference focusing stimulation, as shown in the formula:

[0018]

[0019] Wherein, λ is the subcutaneous standing wave wavelength;

[0020] S6. Posture correction guidelines are dynamically projected using AR devices, allowing patients to adjust their posture and reassess their muscle imbalance index according to the guidelines;

[0021] S7. After each day's training, the cloud platform compares the daytime data with the personalized digital twin model and updates the training parameters.

[0022] Preferably, in step S1, the standard action sequence includes:

[0023] Self-check actions: Standing still, forward bend, left and right lateral bends;

[0024] Training exercises: overhead squat with both hands, freehand lunge walk, and backward step swallow balance;

[0025] The sequence of actions is from self-check to training, and is repeated 3 times a day.

[0026] Preferably, in the standard action sequence,

[0027] The assessment parameters for overhead squats include: spinal lateral swing angle and hip flexion asymmetry;

[0028] The assessment parameters for freehand archery walking include: spinal-pelvic rotation angle and ankle trajectory deviation;

[0029] The evaluation parameters for the backward straddle balance include: the angle between the thigh and the horizontal plane, and the distance of the knee center swing.

[0030] Each parameter is calculated through spatiotemporal fusion of IMU and electromyographic signals.

[0031] Preferably, step S3 further includes: performing adaptive wavelet packet decomposition, with the formula:

[0032]

[0033] Where λ is the regularization coefficient. Here, represents the motion artifact transfer function, and IMU represents the inertial measurement unit data.

[0034] Preferably, in step S4, the muscle force vector magnitude f i The formula, calculated using a musculoskeletal dynamics model, is as follows:

[0035]

[0036] Where J(θ) is the Jacobian matrix of the spine reconstructed from the patient's CT scan, R is the muscle force transmission matrix calibrated by deep learning, and B is the viscous damping coefficient.

[0037] Preferably, in step S5, the triggering of the phase interference focusing stimulus adopts the PPO-Clip algorithm optimization strategy, and the formula is:

[0038]

[0039] Among them, s t Let a be the state space. t This is the action space.

[0040] Preferably, in step S7, the personalized digital twin model pathological progression prediction algorithm is formulated as follows:

[0041] h(θ,t)=h0(t)exp(STGCN(F kin +0.78·Cobb t0 );

[0042] Among them, Fkin Cobb extracts musculoskeletal dynamics features from a spatiotemporal graph convolutional network. t0 This is the initial Cobb angle.

[0043] The present invention also provides a self-correction system for AIS patients, comprising:

[0044] The data acquisition unit includes a 64-channel electrode array, a 9-axis IMU, and a miniature pressure sensor integrated into the smart wearable electromyography testing device.

[0045] An edge processing unit, deployed on a portable terminal, is used to perform adaptive wavelet packet decomposition;

[0046] The cloud analytics unit deploys a muscle loss quantification model and a personalized digital twin model.

[0047] A feedback control unit, which includes a haptic feedback device and an AR device;

[0048] A closed-loop management unit, which is used for coordinated control.

[0049] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when loaded onto the processor, implements the above-described self-correction method for AIS patients.

[0050] The present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described self-correction method for AIS patients.

[0051] The present invention achieves the following beneficial technical effects compared to the prior art:

[0052] This invention provides a self-correction system and method for AIS patients, achieving precise management of scoliosis throughout its entire lifecycle. Through a 64-channel electrode array and a phase interference tactile feedback architecture, it can capture paraspinal electromyographic activity in real time and achieve precise subcutaneous stimulation, significantly improving the sensitivity of muscle imbalance detection. The muscle imbalance index establishes a direct mathematical correlation between muscle strength difference and Cobb angle changes, solving the problem of disconnect between biomechanical indicators and imaging parameters in traditional assessments. A dynamic correction strategy based on the PPO-Clip reinforcement learning algorithm can automatically adjust training intensity according to the patient's real-time physiological state, further improving training compliance. The edge-cloud collaborative processing architecture compresses critical response latency to extremely low levels, supports daily continuous pathological monitoring, and achieves integrated closed-loop intervention of "assessment-training-prediction," significantly reducing the risk of Cobb angle progression. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0054] Figure 1 This invention provides a flowchart of a self-correction method for AIS patients. Detailed Implementation

[0055] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] The purpose of this invention is to provide a solution to the problems existing in the prior art.

[0057] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0058] Example 1:

[0059] This embodiment provides a self-correction method for AIS patients, such as... Figure 1 As shown, it includes the following:

[0060] S1. The patient wears a smart wearable electromyography (EMG) testing device and performs a standard sequence of actions; the standard sequence of actions includes:

[0061] Self-check actions: standing at rest, forward flexion, and lateral flexion; the patient completes the standard actions for 3 minutes while standing, generating the "Muscle Status Report" for the day;

[0062] Training exercises: overhead squat with both hands, freehand lunge walk, and backward step swallow balance;

[0063] The sequence of actions is from self-check to training, repeated 3 times daily;

[0064] In the above standard action sequence,

[0065] The assessment parameters for overhead squats include: spinal lateral swing angle and hip flexion asymmetry;

[0066] The assessment parameters for freehand archery walking include: spinal-pelvic rotation angle and ankle trajectory deviation;

[0067] The evaluation parameters for the backward straddle balance include: the angle between the thigh and the horizontal plane, and the distance of the knee center swing.

[0068] Each parameter is calculated through spatiotemporal fusion of IMU and electromyographic signals.

[0069] S2. Acquire 64-channel paravertebral electromyography (EMG) signals using an electrode array in a smart wearable electromyography testing device. i (t), the three-dimensional attitude angle of the spine [θ] is acquired by a 9-axis IMU. sag (t),θ cor (t)], the pressure P at the contact point of the brace is collected by a miniature pressure sensor. j (t); Real-time acquisition of paraspinal electromyographic signals, three-dimensional spinal posture, and brace pressure distribution.

[0070] S3. Execute the dynamic electromyography enhancement algorithm on the portable terminal. The formula is as follows:

[0071]

[0072] Where, Δθ=|θ cor (t)-θ ref |, k is the clinically calibrated curvature sensitivity coefficient;

[0073] It also includes: performing adaptive wavelet packet decomposition, with the formula:

[0074]

[0075] Where λ is the regularization coefficient. Here, represents the motion artifact transfer function, and IMU represents the inertial measurement unit data.

[0076] S4. Calculate the muscle imbalance index based on enhanced electromyographic signals using a muscle imbalance quantification model. The formula is as follows:

[0077]

[0078] Where convex is the set of muscle group numbers on the convex side, concave is the set of muscle group numbers on the concave side, and f i The magnitude of the muscle force vector is calculated using a musculoskeletal dynamics model, and the formula is:

[0079]

[0080] Where J(θ) is the Jacobian matrix of the spine reconstructed from the patient's CT scan, R is the muscle force transmission matrix calibrated by deep learning, and B is the viscous damping coefficient.

[0081] S5. When γ>γ threshAt that time, the haptic feedback device activates phase interference focusing stimulation, as shown in the formula:

[0082]

[0083] Wherein, λ is the subcutaneous standing wave wavelength;

[0084] The triggering of the phase interference focusing stimulus employs the PPO-Clip algorithm optimization strategy, with the following formula:

[0085]

[0086] Among them, s t Let a be the state space. t This is the action space.

[0087] S6. Posture correction guidelines are dynamically projected using AR devices, allowing patients to adjust their posture and reassess their muscle imbalance index according to the guidelines;

[0088] S7. After each day's training, the cloud platform compares the daytime data with the personalized digital twin model and generates personalized training parameter updates.

[0089] The personalized digital twin model algorithm for predicting pathological progression has the following formula:

[0090] h(θ,t)=h0(t)exp(STGCN(F kin +0.78·Cobb t0 );

[0091] Among them, F kin Cobb extracts musculoskeletal dynamics features from a spatiotemporal graph convolutional network. t0 This is the initial Cobb angle.

[0092] Example 2:

[0093] This embodiment provides a self-correction system for AIS patients, including:

[0094] The data acquisition unit includes a 64-channel electrode array (using a 16-channel sEMG sensor), a 9-axis IMU (integrating an accelerometer, gyroscope, and magnetometer), and a miniature pressure sensor integrated into the smart wearable electromyography testing device.

[0095] An edge processing unit, deployed on a portable terminal, is used to perform adaptive wavelet packet decomposition and transmit data using Bluetooth Low Energy 5.2, thereby achieving data preprocessing and local feedback control.

[0096] The cloud analytics unit deploys a muscle loss quantification model and a personalized digital twin model for pathological modeling and training strategy generation.

[0097] The feedback control unit includes a haptic feedback device and an AR device, thereby realizing real-time biofeedback and motion correction;

[0098] A closed-loop management unit, which is used for coordinated control.

[0099] Example 3:

[0100] This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is loaded onto the processor, it implements the above-described self-correction method for AIS patients.

[0101] Example 4:

[0102] This embodiment provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described self-correction method for AIS patients.

[0103] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0104] It should be noted that the components mentioned in the above embodiments are all general standard parts or components known to those skilled in the art. Their structures and principles can be learned by those skilled in the art through technical manuals or conventional experimental methods.

[0105] This invention has illustrated its principles and implementation methods using specific examples. The descriptions of these embodiments are merely illustrative of the method and its core ideas; furthermore, those skilled in the art will recognize that modifications may be made to the specific implementation methods and application scope based on the principles of this invention. Therefore, the content of this specification should not be construed as limiting the invention.

Claims

1. A self-correction method for AIS patients, characterized by: Including the following: S1. The patient wears a smart wearable electromyography (EMG) testing device and performs a standard sequence of movements; S2. Acquire 64-channel paravertebral electromyography (EMG) signals using an electrode array in a smart wearable electromyography testing device. i (t), the three-dimensional attitude angle of the spine [θ] is acquired by a 9-axis IMU. sag (t),θ cor (t)], the pressure P at the contact point of the brace is collected by a miniature pressure sensor. j (t); S3. Execute the dynamic electromyography enhancement algorithm on the portable terminal. The formula is as follows: Where, Δθ=|θ cor (t)-θ ref |, k is the clinically calibrated curvature sensitivity coefficient; S4. Calculate the muscle imbalance index based on enhanced electromyographic signals using a muscle imbalance quantification model. The formula is as follows: Where convex is the set of muscle group numbers on the convex side, concave is the set of muscle group numbers on the concave side, and f i The magnitude of the muscle force vector; S5. When γ>γ thresh At that time, the haptic feedback device activates phase interference focusing stimulation, as shown in the formula: Wherein, λ is the subcutaneous standing wave wavelength; S6. Posture correction guidelines are dynamically projected using AR devices, allowing patients to adjust their posture and reassess their muscle imbalance index according to the guidelines; S7. After each day's training, the cloud platform compares the daytime data with the personalized digital twin model and updates the training parameters.

2. The self-correction method for AIS patients according to claim 1, characterized in that: In step S1, the standard action sequence includes: Self-check actions: Standing still, forward bend, left and right lateral bends; Training exercises: overhead squat with both hands, freehand lunge walk, and backward step swallow balance; The sequence of actions is from self-check to training, and is repeated 3 times a day.

3. The self-correction method for AIS patients according to claim 2, characterized in that: In the standard action sequence, The assessment parameters for overhead squats include: spinal lateral swing angle and hip flexion asymmetry; The assessment parameters for freehand archery walking include: spinal-pelvic rotation angle and ankle trajectory deviation; The evaluation parameters for the backward straddle balance include: the angle between the thigh and the horizontal plane, and the distance of the knee center swing. Each parameter is calculated through spatiotemporal fusion of IMU and electromyographic signals.

4. The self-correction method for AIS patients according to claim 1, characterized in that: Step S3 also includes: performing adaptive wavelet packet decomposition, with the following formula: Where λ is the regularization coefficient. Here, represents the motion artifact transfer function, and IMU represents the inertial measurement unit data.

5. The self-correction method for AIS patients according to claim 1, characterized in that: In step S4, the muscle force vector magnitude f i The formula, calculated using a musculoskeletal dynamics model, is as follows: Where J(θ) is the Jacobian matrix of the spine reconstructed from the patient's CT scan, R is the muscle force transmission matrix calibrated by deep learning, and B is the viscous damping coefficient.

6. The self-correction method for AIS patients according to claim 1, characterized in that: In step S5, the triggering of the phase interference focusing stimulus adopts the PPO-Clip algorithm optimization strategy, and the formula is as follows: Among them, s t Let a be the state space. t This is the action space.

7. The self-correction method for AIS patients according to claim 1, characterized in that: In step S7, the personalized digital twin model pathology progression prediction algorithm has the following formula: h(θ,t)=h0(t)exp(STGCN(F kin )+0.78·Cobb t0 ); Among them, F kin Cobb extracts musculoskeletal dynamics features from a spatiotemporal graph convolutional network. t0 This is the initial Cobb angle.

8. A self-correction system for AIS patients, characterized in that: include: The data acquisition unit includes a 64-channel electrode array, a 9-axis IMU, and a miniature pressure sensor integrated into the smart wearable electromyography testing device. An edge processing unit, deployed on a portable terminal, is used to perform adaptive wavelet packet decomposition; The cloud analytics unit deploys a muscle loss quantification model and a personalized digital twin model. A feedback control unit, which includes a haptic feedback device and an AR device; A closed-loop management unit, which is used for coordinated control.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the computer program is loaded into the processor, it implements the AIS patient self-correction method according to any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, it implements the AIS patient self-correction method according to any one of claims 1-7.

Images