Personalized acupoint physiotherapy wearing system and neck acupoint personalized positioning method

Abstract

The invention discloses a personalized acupoint physiotherapy wearing system and a neck acupoint personalized positioning method. The method mainly comprises the steps that body surface feature data such as the height and the neck length of a user are acquired; calculating the body surface distance of the internal spine key bony marker through a prediction model constructed by a regression algorithm with a regularization item; calculating an individual same-body cun unit according to a traditional Chinese medicine bone degree cun division method; and constructing a two-dimensional coordinate system and calculating personalized neck acupoint coordinates by combining the position of the bony marker and the same body cun. The system comprises a data acquisition module, an operation processing module used for executing the method, and a flexible wearable device capable of activating the specific physiotherapy unit according to the coordinate matching. The problem of inaccurate acupoint positioning caused by individual dissection difference in non-specific cervical pain family intervention is solved, low-cost and high-precision acupoint positioning and targeted physiotherapy can be achieved only through a small number of human body measurement parameters, and the method is suitable for family personalized rehabilitation application.

Description

Technical Field

[0001] This invention relates to medical rehabilitation engineering and intelligent wearable system technology, and in particular to a personalized acupoint physiotherapy wearable system and a method for locating acupoints on the neck. Background Technology

[0002] Nonspecific neck pain is a musculoskeletal disorder characterized by high incidence and long-term recurrence in clinical practice. The effectiveness of interventions largely depends on the accuracy of treatment placement and individual adaptability. With the development of wearable medical devices and home rehabilitation technologies, achieving precise neck physiotherapy interventions in non-clinical settings has become a crucial technical challenge in current research and product development. Existing related technologies mainly focus on the following two areas.

[0003] The first category of technologies is based on predicting individual morphology using human body parameters. These methods typically collect anthropometric parameters such as height, weight, trunk length, and shoulder width to construct a mapping relationship between these parameters and body shape, which is then used in applications such as anthropometric estimation, clothing customization, or human model reconstruction. The primary focus of this type of technology is on morphological reconstruction, and it is not further used to predict specific treatment locations or stimulation points, thus making it difficult to directly support precise physical therapy interventions based on anatomical or functional locations.

[0004] The second category of technology involves solutions for predicting acupoint locations. Existing solutions largely rely on manually calibrated acupoints as training samples, combined with image acquisition equipment to obtain images of the human body surface, and then use image processing or deep learning algorithms to identify and predict acupoints. While this type of method achieves a certain degree of automated acupoint location identification, its prediction process is highly dependent on image quality, shooting angle, lighting conditions, and the stability of the human body's posture, and typically requires complex acquisition environments or specialized equipment. Furthermore, the subjective differences inherent in the manual calibration process further affect the consistency of model training and the stability of prediction results.

[0005] Therefore, existing technologies lack a technical solution that can directly predict individual acupoint locations based on human body parameters without relying on high-quality image acquisition, and further work in conjunction with wearable physiotherapy devices to meet the application needs of home-based and personalized intervention for non-specific neck pain.

[0006] Medical-grade products (orthopedic braces) utilize 3D scanning or plaster molding, which, while precise, are extremely expensive and only suitable for severe cases, not for routine physical therapy. Consumer-grade products (such as SKG and Breo) focus on optimizing massage head materials or electrical pulse waveforms, neglecting precise positional matching. They often employ a "one-size-fits-all" design or rely on simple mechanical structures (such as elastic arms or manual adjustment knobs) to allow users to adjust the position by feel, resulting in low precision and a poor user experience.

[0007] Existing home-use neck therapy products mainly suffer from the following defects: (1) Lack of personalized adaptation and large positioning deviation: Most of the cervical spine physiotherapy devices on the market adopt standardized size design, ignoring the individual anatomical differences of users. Due to the different skeletal landmark positions of people with different heights and body types, the physiotherapy contact points cannot accurately cover key acupoints such as Dazhui, Jianjing, and Jiaji, often resulting in "position deviation" phenomenon, which seriously affects the intervention effect of non-specific neck pain.

[0008] (2) Traditional customization is costly and difficult to popularize: Although medical-grade braces or high-end customization can achieve precise positioning through 3D scanning and CT / X-ray imaging, this process relies on expensive professional equipment and doctor intervention, and cannot be achieved at low cost in home settings.

[0009] (3) Lack of effective prediction methods from the body surface to the internal skeleton: Existing technologies cannot infer the specific location of the spinous processes of the spine (such as the spinous process T1 of the first thoracic vertebra and the spinous process T3 of the third thoracic vertebra) deep under the skin based solely on simple body surface data (such as height and forearm length) measured by the user, resulting in a lack of anatomical basis for the acupoint positioning of home devices.

[0010] It should be noted that the information disclosed in the background section above is only for understanding the background of this application, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0011] The main objective of this invention is to overcome the deficiencies in the aforementioned background technology and provide a personalized acupoint therapy wearable system and a method for locating acupoints on the neck.

[0012] To achieve the above objectives, the present invention adopts the following technical solution: A personalized method for locating acupoints on the neck includes the following steps: S1. Obtain multiple external anthropometric feature data of the target user; S2. Input the feature data into a pre-trained spinal bony landmark prediction model, calculate it using a regression algorithm with a regularization term, and output the predicted distance of the target user's internal spinal key bony landmarks relative to a preset reference point on the body surface. S3. Obtain the forearm length of the target user and calculate the user's individualized body length unit according to the traditional Chinese medicine bone measurement method. S4. Using the predicted spinal bony landmarks as the longitudinal reference and the calculated body length as the lateral measurement unit, a two-dimensional coordinate system is constructed. Based on the predetermined conversion relationship between acupoints, bony landmarks, and body length, the personalized planar coordinates of the target cervical acupoints are calculated.

[0013] Further, in step S2, the spinal bony landmark prediction model is constructed in the following manner: S21. Model training sample generation: Using a parametric digital human body model, multiple sets of virtual human body 3D model samples are generated by sampling in its shape parameter space, and external body surface measurement size data and actual distance data between specific bony landmarks of the internal spine are extracted from each sample as training samples. S22. Establishing the model mapping relationship: To address the multicollinearity problem among the body surface measurement data in the training samples, a loss function containing an L2 regularization term is constructed. By minimizing this loss function, a set of regression coefficients is obtained, thereby establishing a stable linear mapping model from the body surface measurement feature vector to the distance values ​​of internal bony landmarks.

[0014] Further, in step S2, the input feature data vector includes gender, height, neck length, and back length; the predicted body surface distance output by the model includes the body surface distance from the spinous process of the seventh cervical vertebra to the spinous process of the first thoracic vertebra, and the body surface distance from the spinous process of the seventh cervical vertebra to the spinous process of the third thoracic vertebra.

[0015] Furthermore, the regression algorithm with regularization term used in step S2 is either Ridge Regression, Lasso Regression, or Elastic Net.

[0016] Furthermore, in step S3, the individualized body length unit is obtained by calculating the user's forearm length by dividing it into specific equal parts.

[0017] Furthermore, step S4 specifically includes: S41. Coordinate system establishment: The predicted position of the spinous process of the seventh cervical vertebra is taken as the origin of the two-dimensional coordinate system, and the direction of the body surface downward along the spine is taken as the positive direction of the vertical axis. S42. Calculation of acupoint coordinates: For each target acupoint, its vertical coordinate value is determined based on the predicted surface distance of the bony landmark of the specific spinal segment associated with the acupoint, output by the model; its horizontal coordinate value is determined by multiplying the calculated individualized body length unit with the predefined lateral dimension of the acupoint relative to the midline of the spine, so as to characterize the symmetrical distribution of acupoints on both sides of the spine.

[0018] Furthermore, the target cervical acupoints include at least one of the following: Dazhui (GV14), Jianzhongshu (GB2), Dazhu (BL11), Jianwaishu (GB2), and Jiaji (EX-B1). The ordinates of Dazhu (BL11) and Jianwaishu (GB2) are determined based on the predicted surface distance from the spinous process of the seventh cervical vertebra to the spinous process of the first thoracic vertebra, and the ordinate of Jiaji (EX-B1) is determined based on the predicted surface distance from the spinous process of the seventh cervical vertebra to the spinous process of the third thoracic vertebra.

[0019] A personalized acupoint therapy wearable system for non-specific neck pain intervention includes: The data acquisition module is used to acquire the user's anthropometric data. The processing module is configured to execute the personalized neck acupoint positioning method, receive data acquired by the data acquisition module, and calculate and output personalized neck acupoint distribution coordinates. The physiotherapy execution device includes a flexible wearable carrier and multiple physiotherapy units disposed thereon; the spatial distribution position or activatable area of ​​the physiotherapy units on the flexible wearable carrier is adapted and set according to the acupoint distribution coordinates output by the computing module.

[0020] Furthermore, the physiotherapy execution device also includes a control unit and a flexible circuit layer; the flexible circuit layer integrates multiple independently controllable physiotherapy micro-elements distributed in an array; the control unit is used to receive the acupoint distribution coordinates, and select the corresponding physiotherapy micro-elements according to the coordinate information, so as to activate them to cover the predicted target acupoint area in physical space.

[0021] Furthermore, it also includes a mobile terminal application; the mobile terminal application is used to guide the user to input or automatically collect human body measurement feature data, and to visualize the acupoint distribution coordinates calculated by the calculation and processing module, while generating control commands based on the coordinates and sending them to the physiotherapy execution device via wireless communication.

[0022] The present invention has the following beneficial effects: This invention addresses the technical problem of inaccurate acupoint positioning and poor treatment effects in existing home-use neck physiotherapy products due to neglecting individual anatomical differences. It provides an innovative solution that achieves high-precision acupoint positioning using only key anthropometric data. The core of this solution lies in establishing a mapping model between surface features (such as height, neck length, and back length) and key bony landmarks of the internal spine (such as the spinous processes of the seventh cervical vertebra (C7), the first thoracic vertebra (T1), and the third thoracic vertebra (T3). Combined with traditional Chinese medicine bone measurement methods, the predicted bony landmark locations are transformed into personalized two-dimensional coordinates for key acupoints such as Dazhui, Jianjing, and Jiaji. Based on these coordinates, customized physiotherapy plans can be generated, driving the flexible heating module of a wearable terminal to precisely intervene in specific areas. This achieves individualized acupoint prediction based on human parameters without relying on high-quality image acquisition equipment, and works in conjunction with wearable physiotherapy devices, effectively meeting the application needs of home-based, personalized intervention for non-specific neck pain.

[0023] Specifically, this invention effectively solves the problem that existing home-use devices fail to accurately cover key acupoints due to neglecting individual differences (such as neck length and bone position), resulting in poor or even ineffective intervention. Simultaneously, it overcomes the difficulty of traditional medical-grade customized treatments relying on expensive imaging equipment (such as CT / MRI) and being difficult to popularize at low cost. Its preferred implementation innovatively combines a regression algorithm with a regularization term (such as ridge regression) with traditional Chinese medicine bone measurement methods. The algorithm infers the location of internal spinous processes from body surface dimensions (solving the anatomical problem of "where are the bones?"), and then determines the location of acupoints based on traditional Chinese medicine theory (solving the meridian problem of "where are the acupoints?"). Thus, it achieves near-medical-grade positioning accuracy with only a few simple measurement data points.

[0024] The beneficial effects of this invention are mainly reflected in the following aspects: First, it possesses high-precision acupoint location capabilities. By employing algorithms such as ridge regression to process multicollinearity in anthropometric data, the model can accurately predict the location of invisible bony landmarks. Experiments show that the model's coefficient of determination R² on the test set exceeds 0.97, and the mean absolute percentage error (MAPE) is less than 1%, ensuring that even for users of different body types, the system can accurately calculate the positions of T1 and T3 spinous processes, allowing the therapeutic patches to accurately adhere to acupoints such as Jianwaiyu and Jiaji, significantly improving the therapeutic effect. Second, it has the advantages of low cost and high ease of use. Compared to customized medical solutions that rely on MRI / CT or 3D scanning, this invention only requires users to provide simple external data such as height, neck length, and forearm length, and the algorithm can reverse-engineer the internal structure. This "small data-driven" mode greatly reduces the customization threshold, enabling high-precision personalized therapeutic products to enter ordinary households. Third, it achieves a scientific integration of traditional Chinese and Western medicine. This invention organically integrates modern statistical learning algorithms (optimal ridge regression) with traditional Chinese medicine bone measurement methods, forming a complete and digitizable closed-loop acupoint positioning technology. Finally, it provides a customized physiotherapy experience. The system generates a personalized physiotherapy layout based on calculated coordinates, avoiding energy waste from ineffective areas and preventing discomfort that may be caused by incorrect stimulation of non-acupoint areas, thus achieving precise intervention for non-specific neck pain.

[0025] In summary, this invention is mainly used for home-based adjunctive treatment of nonspecific neck pain patients, and can achieve personalized customization and precise treatment of physiotherapy equipment at low cost.

[0026] Other beneficial effects of the embodiments of the present invention will be further described below. Attached Figure Description

[0027] Figure 1 This is the overall flowchart of the personalized neck acupoint positioning method of the present invention.

[0028] Figure 2This is a technical roadmap for an embodiment of the present invention.

[0029] Figure 3 This is a schematic diagram illustrating the extraction of body inch units and acupoint location in an embodiment of the present invention.

[0030] Figure 4 This is an overall block diagram of the personalized acupoint therapy wearable system of the present invention. Detailed Implementation

[0031] The embodiments of the present invention will be described in detail below. It should be emphasized that the following description is merely exemplary and not intended to limit the scope and application of the present invention.

[0032] It should be noted that when a component is referred to as "fixed to" or "set on" another component, it can be directly on or indirectly on that other component. When a component is referred to as "connected to" another component, it can be directly connected to or indirectly connected to that other component. Furthermore, a connection can be used for fixing, coupling, or communication.

[0033] It should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", and "outer" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the embodiments of the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the present invention.

[0034] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of embodiments of the present invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0035] This invention aims to solve the problem of inaccurate acupoint positioning caused by neglecting individual anatomical differences in home-use neck physiotherapy devices. It proposes a technical solution that uses a ridge regression algorithm to construct a mapping model between body surface features and spinal bony landmarks, and combines it with the traditional Chinese medicine bone measurement method to calculate personalized acupoint coordinates. With only a small amount of anthropometric data, it can achieve low-cost, high-precision acupoint positioning and targeted physiotherapy intervention without the need for expensive imaging equipment.

[0036] See Figure 1 and Figure 3 This invention provides a personalized method for locating acupoints on the neck, comprising the following steps: Step S1: Obtain multiple external anthropometric feature data of the target user.

[0037] Step S2: Input the feature data into a pre-trained spinal bony landmark prediction model, calculate using a regression algorithm with regularization term, and output the predicted distance of the target user's internal spinal key bony landmarks relative to a preset reference point on the body surface.

[0038] In some embodiments, in step S2, the spinal bony landmark prediction model is constructed in the following manner: S21. Model training sample generation: Using a parametric digital human body model, multiple sets of virtual human body 3D model samples are generated by sampling in its shape parameter space, and external body surface measurement size data and actual distance data between specific bony landmarks of the internal spine are extracted from each sample as training samples. S22. Establishing the model mapping relationship: To address the multicollinearity problem among the body surface measurement data in the training samples, a loss function containing an L2 regularization term is constructed. By minimizing this loss function, a set of regression coefficients is obtained, thereby establishing a stable linear mapping model from the body surface measurement feature vector to the distance values ​​of internal bony landmarks.

[0039] In some embodiments, in step S2, the input feature data vector includes gender, height, neck length, and back length; the predicted body surface distance output by the model includes the body surface distance from the spinous process of the seventh cervical vertebra to the spinous process of the first thoracic vertebra, and the body surface distance from the spinous process of the seventh cervical vertebra to the spinous process of the third thoracic vertebra.

[0040] In some embodiments, the regression algorithm with regularization term used in step S2 is the Ridge Regression Algorithm, the Lasso Regression Algorithm, or the Elastic Net Algorithm.

[0041] Step S3: Obtain the forearm length of the target user and calculate the individualized body length unit of the user according to the traditional Chinese medicine bone measurement method.

[0042] In some embodiments, in step S3, the individualized body length unit is obtained by calculating the user's forearm length by dividing it into specific equal parts.

[0043] Step S4: Using the predicted spinal bony landmarks as the longitudinal reference and the calculated body length as the lateral measurement unit, construct a two-dimensional coordinate system, and calculate the personalized planar coordinates of the target cervical acupoints based on the predetermined conversion relationship between acupoints, bony landmarks, and body length.

[0044] In some embodiments, step S4 specifically includes: S41. Coordinate system establishment: The predicted position of the spinous process of the seventh cervical vertebra is taken as the origin of the two-dimensional coordinate system, and the direction of the body surface downward along the spine is taken as the positive direction of the vertical axis. S42. Calculation of acupoint coordinates: For each target acupoint, its vertical coordinate value is determined based on the predicted surface distance of the bony landmark of the specific spinal segment associated with the acupoint, output by the model; its horizontal coordinate value is determined by multiplying the calculated individualized body length unit with the predefined lateral dimension of the acupoint relative to the midline of the spine, so as to characterize the symmetrical distribution of acupoints on both sides of the spine.

[0045] In some embodiments, the target cervical acupoints include at least one of Dazhui (GV14), Jianzhongshu (GB2), Dazhu (BL11), Jianwaishu (GB2), and Jiaji (EX-B1). The ordinates of Dazhu (BL11) and Jianwaishu (GB2) are determined based on the predicted surface distance from the spinous process of the seventh cervical vertebra to the spinous process of the first thoracic vertebra, and the ordinate of Jiaji (EX-B1) is determined based on the predicted surface distance from the spinous process of the seventh cervical vertebra to the spinous process of the third thoracic vertebra.

[0046] See Figure 4 This invention also provides a personalized acupoint therapy wearable system for non-specific neck pain intervention, comprising a data acquisition module, a processing module, and a therapy execution device. The data acquisition module acquires the user's anthropometric data. The processing module is configured to execute the personalized neck acupoint positioning method, receive the data acquired by the data acquisition module, and calculate and output personalized neck acupoint distribution coordinates. The therapy execution device includes a flexible wearable carrier and multiple therapy units disposed thereon. The spatial distribution position or activatable area of ​​the therapy units on the flexible wearable carrier is adapted and set according to the acupoint distribution coordinates output by the processing module.

[0047] In some embodiments, the physiotherapy execution device further includes a control unit and a flexible circuit layer; the flexible circuit layer integrates multiple independently controllable physiotherapy micro-elements distributed in an array; the control unit is used to receive the acupoint distribution coordinates, and select the corresponding physiotherapy micro-elements according to the coordinate information, so as to activate them to cover the predicted target acupoint area in physical space.

[0048] In some embodiments, the personalized acupoint therapy wearable system further includes a mobile terminal application; the mobile terminal application is used to guide the user to input or automatically collect human body measurement feature data, and to visualize the acupoint distribution coordinates calculated by the computing module, while generating control commands based on the coordinates and sending them to the therapy execution device via wireless communication.

[0049] This invention proposes a personalized acupoint positioning method for the neck, innovatively integrating modern statistical learning algorithms with traditional Chinese medicine theory. By constructing a mapping model between body surface features and spinal bony landmarks, it effectively solves the problem of multicollinearity in anthropometric data. This allows for high-precision inference of the location of key internal spinous processes based on only a limited amount of external data such as height and neck length. Furthermore, by combining this with traditional Chinese medicine bone measurement methods, completely personalized acupoint coordinates are calculated, achieving a complete technical loop from "anatomical positioning" to "meridian positioning." This solution fundamentally overcomes the positioning errors caused by the "one-size-fits-all" design of existing home products and the difficulty in widespread adoption of medical-grade customization due to the reliance on expensive imaging equipment. It achieves positioning accuracy comparable to professional-grade products at a very low cost. Finally, by mapping personalized coordinates to a matrix-style physiotherapy unit on a wearable device, the system can achieve on-demand, precise energy delivery or physical stimulation. This not only significantly improves the effectiveness of home intervention for non-specific neck pain but also avoids ineffective treatment or resource waste caused by inaccurate positioning, successfully achieving an excellent balance between low cost, ease of use, high precision, and personalization.

[0050] The following further describes specific embodiments and experimental verifications of the present invention.

[0051] This invention relates to a personalized neck acupoint location method based on human body measurement features. It is a method for predicting neck acupoint locations based on human body parameters, realizing the mapping from human body surface measurement data to acupoint coordinates, and providing a personalized physiotherapy system built based on this method.

[0052] A method for predicting the location of acupoints in the neck based on human body parameters, the overall technical approach is as follows: Figure 2 As shown, it includes the following steps: Step 1: Feature Parameter Acquisition: Obtain the external anthropometric feature data of the target user. This feature data is the basic input to the prediction model and must include features that significantly affect spinal length: height, neck length (distance from the C2 spinous process to the C7 spinous process), and back length (distance from the C2 spinous process to the L3 spinous process). Preferably, it also includes gender, sitting height, and total spinal length. This data can be obtained by guiding the user to perform measurements with a measuring tape or through image recognition via a mobile app.

[0053] Step 2, Bony Landmark Mapping: Input the feature data obtained in Step 1 into the pre-built spinal bony landmark prediction model, and calculate and output the surface distance of the key spinal bony landmarks of the target user relative to the reference point (C7 spinous process) through the ridge regression algorithm.

[0054] Step 21, Model Training Phase: To build the model, a large number of virtual human body samples (e.g., 600 sets) are first generated using a parametric digital human body model (such as SKEL or SMPL). Truncated Gaussian sampling is performed in the shape latent space to generate a 3D mesh model covering different body types (fat, thin, tall, short). Two types of data are extracted from each virtual sample: input... (Body surface measurement dimensions) and output end (The actual geodesic distance between key points of the skeleton) thus forms multiple sets of sample data containing the correspondence between external body surface dimensions and internal skeletal positions.

[0055] Step 22, Model Construction and Ridge Regression Algorithm: To address the strong multicollinearity among anthropometric data (e.g., the correlation between height and spinal length, where directly using ordinary least squares can lead to overfitting or parameter instability), this invention constructs a loss function with an L2 regularization term: in, This represents the true value of the distance to bony landmarks. The input is the body surface feature vector. For regression coefficients, The regularization parameter is used; the regression coefficients are solved by minimizing the loss function. Establish a linear mapping relationship from external surface features to internal bony landmarks: .

[0056] Step 23, Predicted Output: The surface distance of the body output by the model. Specifically, this includes: The distance on the body surface from the spinous process of the seventh cervical vertebra (C7) to the spinous process of the first thoracic vertebra (T1); : The distance on the body surface from the spinous process of the seventh cervical vertebra (C7) to the spinous process of the third thoracic vertebra (T3).

[0057] The input feature data vector x can be standardized and includes, but is not limited to, gender, height, sitting height, neck length, back length, and total spine length.

[0058] Step 3, Body Measurement Calculation: To locate acupoints using traditional Chinese medicine bone measurement methods, it is necessary to calculate the individual's "cun" length. Obtain the target user's forearm length. (The straight-line distance from the elbow crease to the wrist crease), the calculation formula is: This value It will be used as the reference unit for horizontal coordinate calculation.

[0059] Step 4: Acupoint Coordinate Generation: Establish a coordinate system with the C7 spinous process as the origin. The coordinate system is defined with the positive Y-axis pointing vertically downwards and the positive X-axis pointing horizontally to the right. Based on the predicted... and the calculated Using the predicted spinal bony landmarks as the longitudinal reference and the corresponding body measurements as the lateral unit, the planar coordinates of key acupoints are calculated. The calculation logic for the center coordinates (x, y) of each key acupoint is as follows: Dazhui acupoint (GV14): Located at the origin, with coordinates (0, 0); Jianzhongshu (SI15): Located 2 cun lateral to the lower border of the C7 spinous process, with coordinates as follows: ; Dazhu (BL11): Located on the horizontal line below the spinous process of T1 (i.e., downwards from the origin). ), 1.5 inches to the side, coordinates are ; Jianwaiyu (SI14): Located 3 cun lateral to the lower border of the T1 spinous process, at coordinates [missing information]. ; Jiaji point (EX-B2, T3 segment): Located on the horizontal line below the lower edge of the T3 spinous process (i.e., downwards from the origin). ), 0.5 inches to the side, coordinates are ; in the formula These represent acupoints symmetrically distributed on both sides of the spine.

[0060] Extraction of the same body size unit Bcun as follows Figure 3 As shown in (a), the acupoints are illustrated as follows: Figure 3 As shown in (b).

[0061] Based on the above method, a personalized acupoint therapy wearable system for non-specific neck pain intervention, such as... Figure 4 As shown, it includes: Data acquisition module: used to collect users' physical characteristics such as height, neck length, back length, sitting height, and forearm length.

[0062] The calculation and processing module is equipped with the prediction method described above, receives data, and calculates the set of acupoint coordinates (x, y).

[0063] Physiotherapy execution device (smart physiotherapy garment): Flexible wearable carrier: Made of highly elastic fabric to ensure a close fit to the curves of the human back.

[0064] Flexible circuit layer and heating micro-element: Several physiotherapy units are arranged on the carrier, such as integrated matrix-distributed graphene heating micro-element (or heating pads in preset positions).

[0065] Control logic: The physical layout or electrically activated area of ​​the physiotherapy unit on the flexible wearable carrier is matched and set according to the (x,y) coordinates output by the computing module. That is, the system only activates the heating unit covering the area of ​​the predicted acupoint coordinates, thereby achieving precise physiotherapy.

[0066] Optionally, the system further includes a mobile terminal application, which guides the user to input measurement data and projects the calculation results visually onto a virtual human back model, while simultaneously sending control commands to the physiotherapy execution device via a wireless communication protocol.

[0067] See also Figure 2 In some specific embodiments, the method mainly includes three stages: model building and training, user data processing and coordinate calculation (APP side), and hardware execution and feedback (device side). Model Building and Training: This stage first constructs the data source, using parametric digital human bodies as samples. A shape latent space truncated Gaussian sampling strategy is employed to output 600 sets of 3D digital human body meshes. Next, feature engineering is performed, inputting height, neck length, sitting height, spine length, and gender as features, and setting Y1 (C7-T1 interval) and Y2 (C7-T3 interval) as prediction targets. Finally, algorithm training is conducted, with ridge regression being the preferred algorithm. After training, the model performance reaches R0. 2 =0.9766, MAPE(%)=0.93.

[0068] User data processing and coordinate calculation (APP side): The first step is user data collection. This step collects key parameters such as height, neck length, back length, sitting height, spine length, gender, and forearm length through measurement or APP recognition interaction, providing basic data support for subsequent calculations. Next, the core logic calculation stage begins. Height, sitting height, neck length, back length, and gender are input into the Ridge Regression Model to calculate the longitudinal anchor points Y1 (length of segment C7-T1) and Y2 (length of segment C7-T3). At the same time, using the calculation method of "length from elbow crease to wrist crease ÷ 12", the individual inch Bcun (individualized body inch) is obtained as the horizontal scale. In addition, the coordinate synthesis formula is established, specifying that the coordinate P(x,y)=(Bcun×n,Ypred), where the horizontal coordinate x is the lateral inch number n multiplied by Bcun, and the vertical coordinate y is the pre-assigned segment length (i.e., Y1 or Y2). Finally, the six-dimensional coordinate generation step outputs the personalized coordinates of specific acupoints based on the results of the core logic operation. The coordinates of Dazhui are (0,0), Jianzhongshu are (±2Bcun,0), Dazhu are (±1.5Bcun,Y1), Jianwaishu are (±3Bcun,Y1), and Jiaji are (±0.5Bcun,Y2).

[0069] Hardware execution and feedback (device side): First, the communication and mapping module receives the instructions of activation unit ID and PWM duty cycle, and matches and maps the calculated coordinates (x,y) to the center of the physiotherapy unit; then, it enters the precise intervention stage, using the wearable physiotherapy device as a carrier, and uses heating, electric stimulation, massage and other units to intervene at the target position; the final effect is to effectively relieve non-specific neck pain.

[0070] Examples and experimental verification Details of predictive model construction and validation data The specific training parameters and performance verification of the model described in step 2 of the technical solution are explained in detail.

[0071] 1. Sample Generation Parameters: Using the SKEL parametric human body model, when sampling in the shape parameter space, the height range was set to 150cm-190cm, and the BMI range was set to 18.5-30, generating a total of 600 virtual samples (male to female ratio 1:1). This sample size covers more than 95% of the body shape characteristics of adults.

[0072] 2. Model Parameter Optimization: In the training of the ridge regression model, the optimal regularization parameters are determined by leave-one-out cross-validation (LOOCV). Experiments show that when The model performs best on the test set when the value is 0.08.

[0073] Typical application scenarios and coordinate calculation examples This embodiment demonstrates how the system transforms abstract algorithms into concrete physical therapy plans through a specific user's actual usage process. Assume user A (male) uses this system; their operation process is as follows: 1. Data Input: User A measures and inputs the following data through the APP: height 175cm, sitting height 90cm, spine length 70cm, forearm length 24cm, neck length 9.5cm, back length 45cm, gender male.

[0074] 2. Background calculation: The system first calculates according to the formula. Calculate user A's body size = 2cm. After normalizing the height, neck length, and other data, input them into the ridge regression model. The model outputs the predicted value: = 4.8cm (distance of T1 relative to C7), = 9.6cm (distance of T3 relative to C7).

[0075] 3. Coordinate Generation: The system automatically generates a unique acupoint coordinate table for user A (with Dazhui as the origin): Jianwaiyu (right): ( 6, 4.8). Jiaji point (right): ( 1, 9.6). (The same applies to the other acupoints).

[0076] 4. Execution Result: At this point, control commands are sent to the physiotherapy garment. Since the heating matrix spacing of this physiotherapy garment is 1cm, the system will automatically activate the heating micro-elements located near the coordinates, precisely covering the calculated anatomical positions.

[0077] Hardware construction details of wearable physiotherapy devices Flexible carrier material selection: High-elasticity Lycra fabric blended with 75% nylon and 25% spandex is selected. This blended fabric has excellent four-way elasticity and can maintain close contact with the skin of the back during large dynamic movements (such as looking down and turning the head) to prevent electrode displacement.

[0078] Electrode layout: Heating units are arranged according to the acupoint locations, with each unit measuring 2.5cm × 2.5cm.

[0079] Control unit design: The main control chip uses either ESP32 or nRF52832 (Bluetooth Low Power SoC). The circuit design includes multiple analog switches to dynamically select specific heating units in the matrix according to APP commands.

[0080] The power source is a 1000mAh lithium polymer battery, sewn into the hem of the garment.

[0081] Alternative embodiments: Alternatives to the prediction algorithm: Besides the preferred ridge regression, step 2 of the method in this invention can also use Lasso regression or Elastic Neural Networks to establish the mapping model. However, considering the small sample size and strong collinearity of anthropometric data, ridge regression has the best robustness and computational efficiency in this scenario, and is therefore the preferred choice.

[0082] Alternatives to therapeutic media: The examples primarily describe graphene thermotherapy. However, based on the acupoint coordinates calculated according to this invention, the therapeutic unit can also be replaced with: TENS / EMS electrostimulation electrodes for low-frequency pulse massage; micro-vibration motors for physical massage; magnetotherapy magnets for static magnetic field therapy, and other therapeutic media.

[0083] Alternatives to data acquisition: In addition to manual measurement input, the data acquisition module S1 can also be achieved through computer vision technology: the user takes photos of the back and forearm, and the APP automatically extracts key points of the skeleton and estimates the neck length, back length and forearm length data through image recognition algorithms.

[0084] Performance Analysis (1) Overall predictive performance analysis Table 1 shows the prediction results of the three models on the overall dataset. The results from both the training and test sets demonstrate extremely high fit and stability. The R² values ​​are all between 0.976 and 0.978, with only a small difference between the adjusted R² and these values ​​(R² < 0.001), indicating that the models have high explanatory power for the input features and do not exhibit significant overfitting. On the test set, which serves as the core performance indicator, the MAE is 0.00029, the RMSE is 0.00038, and the MAPE for all datasets is below 1%, indicating extremely low prediction errors.

[0085] Table 1 Overall Performance Evaluation (2) Predictive performance analysis of Y1 (C7–T1 interspinous distance) Table 2 shows the performance of each model in the Y1 (C7-T1 interspinous distance) prediction task. The R² values ​​for the training and test sets are approximately 0.973 and 0.974, respectively. The difference in adjusted R² values ​​is slight, indicating that the models have high explanatory power. On the test set, the MAE is 0.00022, the RMSE is 0.00027, and the MAPE is less than 1%, indicating extremely small prediction errors that meet the tolerance for anthropometric errors.

[0086] (3) Predictive performance analysis of Y2 (C7–T3 spinous process distance) Table 3 shows the performance of each model in the Y2 (C7–T3 spinous process distance) prediction task. The R² values ​​for the training and test sets are approximately 0.979 and 0.982, respectively, and the error metrics are better than those for Y1. The MAPE on the test set is only 0.76%, indicating that the model performs more stably in predicting long-segment spinous process distances.

[0087] Table 2Y1 Predictive Performance Evaluation Table 3 Y2 Predictive Performance Evaluation

[0088] Feedback from real-person physical therapy experiments This section compares the therapeutic effects of the system before and after use by conducting a two-course wearing experiment (each course lasting 5 days, with each day lasting 20 minutes).

[0089] 1. Experimental Design Intervention cycle: 2 treatment courses, once a day, 20 minutes each time.

[0090] Physiotherapy parameters: set to 55℃-65℃ heating.

[0091] Assessment scale: VAS (Visual Analog Scale): Visual Analog Pain Scale (0-10 points, the higher the score, the more painful).

[0092] CROM (Cervical Range of Motion): Cervical spine range of motion (flexion, lateral flexion, rotation).

[0093] 2. Experimental Results (1) CROM analysis This study used repeated measures ANOVA data. As shown in Table 4, significant changes in cervical spine mobility in all six directions were observed between the three measurement time points, indicating a systematic change in the functional range of motion of the cervical spine as the intervention progressed. These results demonstrate that two courses of personalized acupoint therapy intervention can effectively promote the recovery of multi-directional cervical spine mobility in a short period and has a positive effect on improving overall cervical spine function.

[0094] Table 4 Results of CROM Intra-subjective Effect Test Furthermore, the results of the trend analysis are shown in Table 5. CROM in all six directions showed significant linear growth, indicating continuous improvement over time and a certain cumulative effect of the intervention. A significant secondary trend was observed only in the forward flexion direction. This indicates that the rate of improvement in the forward flexion direction slowed slightly after the first course of treatment. No significant non-linear changes were observed in other directions, further suggesting that personalized acupoint therapy mainly improves cervical spine mobility through a stable and continuous linear recovery process.

[0095] Table 5. Results of intra-module comparison test of CROM

[0096] (2) VAS Result Analysis Repeated measures ANOVA showed a significant main effect of time on the pain scores (VAS) of the subjects at the three measurement time points. This indicates that the intervention process has a significant and stable effect on pain relief.

[0097] In summary, this invention proposes a personalized neck acupoint positioning method and a personalized acupoint physiotherapy wearable system based on human body parameters. The key innovative contributions and prominent features of this invention include: the innovative deep integration of machine learning algorithms (such as ridge regression) with traditional Chinese medicine bone measurement methods, constructing a precise mapping model from easily obtainable human surface features (such as height, neck length, and back length) to the locations of key bony landmarks of the internal spine (such as the spinous processes of C7, T1, and T3), which is further transformed into personalized acupoint coordinates; and then deeply integrating this coordinate information with a wearable physiotherapy device, realizing a complete technical closed loop of "positioning-computation-intervention," providing a novel solution for achieving low-cost, high-precision personalized physiotherapy intervention in non-clinical environments. The system is flexibly designed; the input end can use manual measurement or automatic feature extraction via computer vision, and the output physiotherapy medium can be flexibly replaced with various forms such as heat therapy, electrical stimulation, and vibration massage, and the carrier can also be adapted to different wearing forms.

[0098] Compared with the prior art, the significant technical advantages of the present invention are reflected in the following aspects: (1) High-precision acupoint location capability: This invention abandons the traditional "average code" design and effectively solves the multicollinearity problem in human body measurement data through the ridge regression algorithm, which can accurately predict the location of bony landmarks under the body surface. Experimental data show that the model has high goodness of fit and low error, which can ensure accurate calculation of the key spinous process positions for users of different body types, so that the physiotherapy unit can accurately cover the target acupoints and effectively improve the intervention effect.

[0099] (2) Low cost and high ease of use: Compared with customized solutions that rely on medical imaging or 3D scanning, this invention only requires users to provide a few simple body surface measurement data, and the internal anatomical structure can be deduced through the algorithm model. This "small data driven" paradigm greatly reduces the technical threshold and cost of precise customization, making personalized physiotherapy products accessible to home settings.

[0100] (3) Scientific integration of traditional Chinese and Western medicine: This invention innovatively combines modern statistical learning algorithms with traditional Chinese medicine bone measurement method. The algorithm is used to solve the anatomical location problem of "where the bones are", and the bone measurement method is used to solve the meridian location problem of "where the acupoints are", thus forming a complete and quantifiable digital acupoint location technology system.

[0101] (4) Customized physiotherapy experience: The system generates customized physiotherapy plans based on the calculated personalized coordinates, and drives the execution device to perform precise intervention on specific areas. This not only avoids energy waste or ineffective stimulation caused by positional deviation, but also improves the safety and comfort of treatment, and achieves precise and efficient intervention for non-specific neck pain.

[0102] This invention also provides a storage medium for storing a computer program, which, when executed, performs at least the methods described above.

[0103] This invention also provides a control device, including a processor and a storage medium for storing a computer program; wherein the processor executes the computer program by performing at least the method described above.

[0104] This invention also provides a processor that executes a computer program, at least performing the methods described above.

[0105] The storage medium can be implemented by any type of non-volatile storage device, or a combination thereof. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), magnetic random access memory (FRAM), flash memory, magnetic surface memory, optical disc or CD-ROM; magnetic surface memory can be disk storage or magnetic tape storage. The storage media described in the embodiments of this invention are intended to include, but are not limited to, these and any other suitable types of memory.

[0106] In the several embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0107] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0108] In addition, in the various embodiments of the present invention, each functional unit can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0109] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0110] Alternatively, if the integrated units of this invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.

[0111] The methods disclosed in the several method embodiments provided by this invention can be arbitrarily combined without conflict to obtain new method embodiments.

[0112] The features disclosed in the several product embodiments provided by this invention can be arbitrarily combined without conflict to obtain new product embodiments.

[0113] The features disclosed in the several method or device embodiments provided by the present invention can be arbitrarily combined without conflict to obtain new method or device embodiments.

[0114] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various equivalent substitutions or obvious modifications can be made without departing from the concept of the present invention, and all such modifications, achieving the same performance or application, should be considered within the scope of protection of the present invention.

Claims

1. A personalized method for locating acupoints on the neck, characterized in that, Includes the following steps: S1. Obtain multiple external anthropometric feature data of the target user; S2. Input the feature data into a pre-trained spinal bony landmark prediction model, calculate it using a regression algorithm with a regularization term, and output the predicted distance of the target user's internal spinal key bony landmarks relative to a preset reference point on the body surface. S3. Obtain the forearm length of the target user and calculate the user's individualized body length unit according to the traditional Chinese medicine bone measurement method. S4. Using the predicted spinal bony landmarks as the longitudinal reference and the calculated body length as the lateral measurement unit, a two-dimensional coordinate system is constructed. Based on the predetermined conversion relationship between acupoints, bony landmarks, and body length, the personalized planar coordinates of the target cervical acupoints are calculated.

2. The personalized neck acupoint positioning method as described in claim 1, characterized in that, In step S2, the spinal bony landmark prediction model is constructed in the following manner: S21. Model training sample generation: Using a parametric digital human body model, multiple sets of virtual human body 3D model samples are generated by sampling in its shape parameter space, and external body surface measurement size data and actual distance data between specific bony landmarks of the internal spine are extracted from each sample as training samples. S22. Establishing the model mapping relationship: To address the multicollinearity problem among the body surface measurement data in the training samples, a loss function containing an L2 regularization term is constructed. By minimizing this loss function, a set of regression coefficients is obtained, thereby establishing a stable linear mapping model from the body surface measurement feature vector to the distance values ​​of internal bony landmarks.

3. The personalized neck acupoint positioning method as described in claim 2, characterized in that, In step S2, the input feature data vector includes gender, height, neck length, and back length; the predicted body surface distance output by the model includes the body surface distance from the spinous process of the seventh cervical vertebra to the spinous process of the first thoracic vertebra, and the body surface distance from the spinous process of the seventh cervical vertebra to the spinous process of the third thoracic vertebra.

4. The personalized neck acupoint positioning method as described in claim 1, characterized in that, The regression algorithm with regularization term used in step S2 is either Ridge Regression, Lasso Regression, or Elastic Net.

5. The personalized neck acupoint positioning method as described in claim 1, characterized in that, In step S3, the individualized body length unit is obtained by calculating the user's forearm length by dividing it into specific equal parts.

6. The personalized neck acupoint positioning method as described in claim 1, characterized in that, Step S4 specifically includes: S41. Coordinate system establishment: The predicted position of the spinous process of the seventh cervical vertebra is taken as the origin of the two-dimensional coordinate system, and the direction of the body surface downward along the spine is taken as the positive direction of the vertical axis. S42. Calculation of acupoint coordinates: For each target acupoint, its vertical coordinate value is determined based on the predicted surface distance of the bony landmark of the specific spinal segment associated with the acupoint, output by the model; its horizontal coordinate value is determined by multiplying the calculated individualized body length unit with the predefined lateral dimension of the acupoint relative to the midline of the spine, so as to characterize the symmetrical distribution of acupoints on both sides of the spine.

7. The personalized neck acupoint positioning method as described in claim 1, characterized in that, The target neck acupoints include at least one of the following: Dazhui (GV14), Jianzhongshu (GB2), Dazhu (BL11), Jianwaishu (GB2), and Jiaji (EX-B1). The ordinates of Dazhu (BL11) and Jianwaishu (GB2) are determined based on the predicted distance from the spinous process of the seventh cervical vertebra to the spinous process of the first thoracic vertebra, and the ordinate of Jiaji (EX-B1) is determined based on the predicted distance from the spinous process of the seventh cervical vertebra to the spinous process of the third thoracic vertebra.

8. A personalized acupoint therapy wearable system for non-specific neck pain intervention, characterized in that, include: The data acquisition module is used to acquire the user's anthropometric data. The processing module is configured to execute the personalized neck acupoint positioning method as described in any one of claims 1 to 7, receive data acquired by the data acquisition module, and calculate and output personalized neck acupoint distribution coordinates. The physiotherapy execution device includes a flexible wearable carrier and multiple physiotherapy units disposed thereon; the spatial distribution position or activatable area of ​​the physiotherapy units on the flexible wearable carrier is adapted and set according to the acupoint distribution coordinates output by the computing module.

9. The personalized acupoint therapy wearable system as described in claim 8, characterized in that, The physiotherapy execution device also includes a control unit and a flexible circuit layer; the flexible circuit layer integrates multiple independently controllable physiotherapy micro-elements distributed in an array; the control unit is used to receive the acupoint distribution coordinates, and determine the specific position of the physiotherapy micro-element to be activated in the flexible circuit layer according to the coordinates, and then select one or more physiotherapy micro-elements corresponding to the position to activate them so as to cover the predicted target acupoint area in physical space.

10. The personalized acupoint therapy wearable system as described in claim 8, characterized in that, It also includes a mobile terminal application; the mobile terminal application is used to guide the user to input or automatically collect human body measurement feature data, and to visualize the acupoint distribution coordinates calculated by the calculation and processing module. At the same time, it generates control commands based on the coordinates and sends them to the physiotherapy execution device via wireless communication.

Images