Acupuncture depth sensor, acupuncture posture detection system and acupuncture teaching tool

Abstract

The present invention belongs to the technical field of linear dimension measurement, and specifically relates to an acupuncture depth sensor, an acupuncture posture detection system, and an acupuncture teaching tool. The acupuncture depth sensor includes a support body and a plurality of layered electrode assemblies arranged in the support body; the support body has a storage space inside, and a plurality of electrode assemblies are arranged in an interval arrangement in the storage space; the electrode assembly is connected to the detection circuit, and each electrode assembly includes a flexible conductive layer and a conductive frame layer sandwiched on both sides of the flexible conductive layer; at least one end of the support body has an opening for a conductive needle to penetrate into the support body and electrically contact the electrode assembly, thereby changing the electrical connection state of the electrode assembly within the acupuncture depth range. This acupuncture depth sensor can accurately measure the insertion depth of acupuncture. On this basis, the present invention also provides an acupuncture posture detection system and an acupuncture teaching tool, which can improve the standardization level of acupuncture operations, as well as the convenience and safety of acupuncture teaching.

Description

Technical Field

[0001] The invention belongs to the technical field of linear dimension measurement, and in particular relates to an acupuncture depth sensor, an acupuncture posture detection system and an acupuncture teaching tool. Background Art

[0002] Acupuncture, as an ancient treatment method, still occupies an important position in modern medicine. However, traditional acupuncture operation is highly dependent on the operator's experience and touch, which makes the effect of acupuncture treatment largely depend on the individual's technical level. In particular, the lack of a tool that can provide accurate measurement and real-time feedback in the control of acupuncture depth and angle leads to uncontrollability and uncertainty of treatment results.

[0003] Traditional acupuncture teaching mostly relies on real models or animals for practice. This method not only has safety risks, but also involves ethical disputes, making it difficult to conduct large-scale learning and training. Using real models for acupuncture training can easily cause discomfort to the model's body, and mispuncture can easily cause physical injuries. Training on real models can also cause great psychological pressure on both students and models. Training on animals is limited in training effectiveness due to the differences in structure between animals and humans, and faces ethical and legal disputes.

[0004] There are also some acupuncture training auxiliary equipment on the market, such as opening a series of holes on the human body model as simulated acupoints. Although this type of auxiliary model can help students become familiar with the location of acupoints to a certain extent, it cannot further evaluate the accuracy and standardization of acupuncture. For example, it cannot provide accurate measurement data of acupuncture depth and angle, nor can it judge and evaluate whether the acupuncture trajectory during acupuncture is reasonable. On the other hand, when using this type of auxiliary model, the touch and feel experienced by students during acupuncture are quite different from the actual situation, resulting in a disconnection between actual operations and training operations. Therefore, the existing acupuncture training auxiliary equipment has limited effect in improving the accuracy and standardization of acupuncture operations, and it is difficult to meet the needs of modern acupuncture education for high-standard operating skills. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention provides an acupuncture depth sensor, an acupuncture posture detection system and an acupuncture teaching tool. The present invention aims to provide an acupuncture depth sensor capable of accurately measuring acupuncture depth and an acupuncture posture detection system capable of real-time feedback of acupuncture posture, as well as an acupuncture teaching tool based on the system, in order to improve the standardization level of acupuncture operation and the convenience of acupuncture teaching, while reducing the dependence on real models or animals, and improving the safety and ethical compliance of teaching and practice.

[0006] The first aspect of the present invention is to provide a needle puncture depth sensor, including a support body and a plurality of layered electrode assemblies arranged in the support body; the support body has a accommodating space inside, and a plurality of electrode assemblies are arranged in an interval arrangement in the accommodating space; the electrode assembly is connected to a detection circuit, and each electrode assembly includes a flexible conductive layer and a conductive frame layer sandwiched on both sides of the flexible conductive layer; at least one end of the support body has an opening for a conductive needle to penetrate into the support body and electrically contact the electrode assembly, thereby changing the electrical connection state of the electrode assembly within the range of the needle puncture depth.

[0007] As a further optimization scheme for the acupuncture depth sensor, the flexible conductive layer is an elastic foamed composite material layer, including a foamed material matrix and a conductive embedding body distributed in the foamed material matrix, and the conductive embedding body is selected from one or more of conductive fibers and conductive particles.

[0008] As a further optimization solution for the needle puncture depth sensor, the flexible conductive layer is a conductive textile, which is a layered material formed by connecting fibers through weaving or non-woven processes; the fibers forming the conductive textile are at least partially conductive.

[0009] As a further optimization solution for the acupuncture depth sensor, the conductive frame layer includes a frame and a contact network fixed within the frame; the frame is electrically conductive with the contact network, and the contact network is electrically conductive with the flexible conductive layer.

[0010] As a further optimization scheme for the puncture depth sensor, on the side of the conductive frame layer adjacent to the flexible conductive layer, a frame protrudes from the contact network, thereby forming a recessed accommodation area within the range enclosed by the frame; the flexible conductive layer is clamped in the accommodation areas on both sides.

[0011] As a further optimization solution for the needle penetration depth sensor, the conductive frame layer is an integrally formed copper mesh.

[0012] As a further optimization scheme of the acupuncture depth sensor, it also includes a coding circuit board, and each electrode assembly is electrically connected to the coding circuit board; at least one electrical parameter between the detection ends of the coding circuit board changes as the electrode assemblies are punctured to different depths.

[0013] As a further optimization scheme of the acupuncture depth sensor, at least one resistor is provided on the encoding circuit board; resistors are connected between two adjacent ones of the multiple electrode assemblies arranged at intervals, and two of the multiple electrode assemblies arranged at both ends each form a detection end; the resistance value between the two detection ends changes as the acupuncture to different depths causes different numbers of resistor short circuits.

[0014] As a further optimization scheme of the acupuncture depth sensor, at least one capacitor is provided on the encoding circuit board; capacitors are connected between two adjacent ones of the multiple electrode assemblies arranged at intervals, and two of the multiple electrode assemblies arranged at both ends each form a detection end; the capacitance value between the two detection ends changes as the acupuncture to different depths causes different numbers of capacitor short circuits.

[0015] The second aspect of the present invention is to provide an acupuncture posture detection system, comprising the above-mentioned acupuncture depth sensor and an acupuncture training needle; an angle detection element is provided at the end of the acupuncture training needle, and the angle detection element has a built-in angle detection module to detect the acupuncture angle of the acupuncture training needle; the acupuncture depth sensor detects the acupuncture depth of the acupuncture training needle.

[0016] As a further optimization scheme of the acupuncture posture detection system, the acupuncture training needle includes a needle body and an angle detection module arranged on the top of the needle body, the angle detection module includes a gyroscope sensor, a power supply, a wireless charging receiving unit and a wireless transmission unit; the gyroscope sensor monitors the orientation angle data, the power supply supplies power to the gyroscope sensor and the wireless transmission unit, the wireless charging receiving unit receives energy from an external wireless charging transmitting unit to charge the power supply, and the wireless transmission unit outputs the orientation angle data monitored by the gyroscope sensor.

[0017] The third aspect of the present invention is to provide an acupuncture teaching tool, including a simulated human body model and the above-mentioned acupuncture posture detection system; the surface of the simulated human body model has a simulated skin layer, the simulated skin layer is elastic and penetrable; an acupuncture depth sensor is buried under the simulated skin layer.

[0018] As a further optimization solution for acupuncture teaching aids, the acupuncture posture detection system feeds back the sampling data measured by the gyroscope sensor after amplification and unscented Kalman filtering.

[0019] As a further optimization solution for acupuncture teaching aids, the acupuncture posture detection system feeds back the data measured by the gyroscope sensor after amplification and optimized unscented Kalman filtering.

[0020] The optimized unscented Kalman filter process adjusts the hyperparameters by constructing a loss function that follows the control equations of the analyzed system, thereby establishing a physically aware optimization framework; the optimizer is used to force the estimation of states and parameters to satisfy the control equations of the dynamic system.

[0021] The governing equations of a dynamic system are represented as a noisy discrete-time state-space system. The state space contains the state transition function M and the measurement function H. At time step k, the state-space model provides a one-step-ahead state prediction x k+1 and the predicted measurement y ’ k , the model is as follows:

[0022] x k+1 = M(x k , u k , θ) + v k , v k ~ N(0, Q)

[0023] y ’ k = H(x k , u k , θ) + μ k , μ k ~ N(0, R)

[0024] Among them, x k is the state vector at time step k, u k is an external input variable, the random variable v k is the process error vector, sampled from a central Gaussian distribution with covariance Q, and the random variable μ k is the measurement error vector, sampled from a centered Gaussian distribution with covariance R, and the vector θ contains a set of dynamic parameters that characterize the physical system.

[0025] The goal of algorithmic inverse modeling is to reconstruct the state space as described above through a set of observations. To this end, it is necessary not only to recover the state but also the parameters at the same time. The solution to this problem is to use existing state estimation techniques to expand the state into an enhanced state, that is, to process it by combining the state and the parameters in one vector. For time-invariant systems, the enhanced state transfer function is written as M^a=[M,I], where I is a unit vector; since the dynamic parameters are implicit, the measurement function and the measurement error covariance remain unchanged.

[0026] During the test optimization process, the NOMAD derivative-free optimization algorithm is used to solve the black box problem of the objective function.

[0027] The loss function based on mean square error and L2-norm is used for optimization as follows:

[0028] L data = ∑(from k1 to k2) ||y k -y ’ k ||2

[0029] Using the state transfer function to construct the loss function, the resulting physical perception loss function is expressed as:

[0030] L physics = ∑(from k1 to k2) ||x ’ k+1 -M(x’ k , u k , θ)||2.

[0031] Beneficial Effects

[0032] The acupuncture depth sensor provided by the present invention has excellent durability and adaptability, can achieve high-precision measurement of acupuncture depth, has stable and reliable electrical connection during acupuncture measurement, ensures stability and accuracy of signal transmission, and can be used to improve the standardization level of acupuncture operations and the convenience of acupuncture teaching.

[0033] The acupuncture posture detection system provided by the present invention can provide real-time feedback of acupuncture depth and angle information through the cooperation of acupuncture depth sensors and acupuncture training needles with angle detection modules, and can be used to evaluate and provide feedback on the standardization of acupuncture operations. The reliability and accuracy of feedback data can also be improved through effective data processing methods.

[0034] The acupuncture teaching tool provided by the present invention can reproduce a real acupuncture environment, allowing trainees to practice repeatedly without worrying about patient safety. The simulated human body model uses simulated skin and subcutaneous tissue materials to provide a real acupuncture experience. Combined with the depth and angle data provided by the acupuncture posture detection system, the trainee can monitor his or her acupuncture operation in real time through the interactive software of the host computer, adjust the technique in time, and effectively improve the acupuncture skills. The acupuncture teaching tool not only helps to improve the quality of acupuncture teaching, but also helps to promote the standardized development of acupuncture skills.

[0035] The present invention also provides an optimized UKF data processing solution, which performs better than the traditional UKF when data is limited, and adjusts hyperparameters by constructing a loss function to improve performance; it can effectively utilize the underlying physical laws of the system, and is particularly suitable for complex systems; and it can automatically select the value and structure of the covariance matrix during the filtering process, thereby increasing the flexibility and adaptability of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 Schematic diagram of the structure of the acupuncture depth sensor.

[0037] Figure 2 Schematic diagram of the internal structure of the acupuncture depth sensor.

[0038] Figure 3 and Figure 4 Schematic diagram of the structure of a single electrode assembly.

[0039] Figure 5 Schematic diagram of the structure of the foam material matrix.

[0040] Figure 6 Schematic diagram of the contact between the conductive path formed by the conductive fiber and the needle.

[0041] Figure 7 Schematic diagram of the contact between the conductive path formed by the conductive particles and the needle.

[0042] Figure 8 Schematic diagram of a layered material formed by weaving fibers.

[0043] Fig. 9 Schematic diagram of a layered material formed by nonwoven fibers.

[0044] Fig.10 Schematic diagram of the structure of the conductive frame layer.

[0045] Fig.11 This is a schematic diagram of the structure of an acupuncture training needle.

[0046] Fig.12 and Fig.13 Schematic diagram of the structure of the simulated human body model.

[0047] Fig.14 This is a schematic diagram of the structure of simulated acupuncture points on a human body model.

[0048] In the figure: 1. support body; 2. electrode assembly; 3. coding circuit board; 5. acupuncture training needle; 6. simulated human body model; 21. flexible conductive layer; 22. conductive frame layer; 51. needle body; 52. angle detection module; 61. simulated skin layer; 62. simulated subcutaneous tissue; 211. foam material matrix; 212. conductive embedding body; 221. frame; 222. contact network; 601. Yintang; 602. Shuigou; 603. Tanzhong; 604. Qihai; 605. Quze; 606. Laogong; 607. Zusanli; 608. Taichong; 609. Baihui; 610. Dazhui; 611. Huantiao; 612. Quchi; 613. Hegu; 614. Weizhong; 615. Kunlun. DETAILED DESCRIPTION

[0049] The present invention is further illustrated by specific examples below. These examples are exemplary and are intended to illustrate the problem and explain the present invention, but are not intended to be limiting.

[0050] First, the acupuncture depth sensor.

[0051] like Figure 1 and Figure 2As shown, the acupuncture depth sensor includes a support body 1 and a plurality of layered electrode assemblies 2 arranged in the support body 1. The support body 1 is roughly cylindrical and has a storage space inside. The support body 1 has an opening at one end for detecting the acupuncture depth for the needle to penetrate. A plurality of electrode assemblies 2 are arranged in an interval in the storage space of the support body 1. The interval of the electrode assemblies 2 can be adjusted according to the accuracy requirements of acupuncture monitoring. For example, an electrode assembly 2 is arranged at intervals of 2mm, 3mm, 4mm or 5mm. The smaller the interval, the higher the test accuracy. Figure 3 and Figure 4 As shown, each electrode assembly 2 includes a flexible conductive layer 21 and a conductive frame layer 22 sandwiched between the two sides of the flexible conductive layer 21. The electrode assembly 2 is connected to a detection circuit. When a conductive needle, such as an acupuncture needle, is inserted into the support 1 and electrically contacts the electrode assembly 2, the electrical connection state of the electrode assembly 2 within the acupuncture depth range will be changed, and the conductive needles connected between the electrode assemblies 2 will form a new conductive path. The detection circuit detects this change and the acupuncture depth can be determined.

[0052] Each electrode assembly 2 includes a flexible conductive layer 21 and a conductive frame layer 22 sandwiched between the two sides of the flexible conductive layer 21. The conductive frame layer 22 provides support for the deformable and elastic flexible conductive layer 21 on both sides to prevent the flexible conductive layer 21 from excessive deformation and maintain the overall shape of the electrode assembly 2 stable; and the conductive frame layer 22 has a large number of through holes or through grooves for the needle to pass through. In this way, during the acupuncture process, the electrode assembly 2 can accurately reflect the position change of the needle, and the needle always maintains close contact with the electrode assembly 2 during the process of inserting and removing the sensor, ensuring the continuity and stability of the electrical signal, which helps to improve the accuracy of acupuncture depth measurement. The flexible conductive layer 21 is sandwiched between the conductive frame layers 22 and is in full electrical contact with them. As the needle is inserted, the flexible conductive layer 21 is penetrated and adaptively deforms near the penetration site to maintain close contact with the needle and ensure a stable electrical connection. The flexible conductive layer can reduce the mechanical wear caused by repeated insertion and withdrawal of the needle, and the conductive frame layer 22 provides high stability and mechanical strength. The combination of the two ensures good contact, high detection accuracy, and excellent durability.

[0053] In some embodiments, the flexible conductive layer 21 is an elastic foamed composite material layer, including Figure 5 The foam material matrix 211 and the conductive embedding body 212 distributed in the foam material matrix 211 are shown. For example, the foam material matrix 211 is made of a polymer material such as polyaniline / polyurethane, and a conductive component, i.e., the conductive embedding body 212, is pre-mixed into the material before foaming.

[0054] Conductive embedding body 212 may be selected from conductive fibers, such as metal fibers, carbon fibers, conductive polymer fibers, etc. A large number of conductive fibers are distributed in the foam material, which enhances the strength and integrity of the material. Some of the conductive fibers may be scattered in the foam material matrix 211, and some of them are interconnected to form a network-like conductive path. When the needle is inserted into the composite material, the network-like conductive path forms a conductive path with the needle surface. Figure 6 At the same time, this network-like conductive path also forms multiple points of dense contact with the conductive frame layer 22 on both sides of the flexible conductive layer 21. Thus, the network-like conductive path tightly connects the needle and the conductive frame layer 22 together. The conductive fiber has a large aspect ratio, which makes it easier to form a continuous conductive path in the flexible conductive layer 21, which helps the stability of conduction.

[0055] The conductive embedding body 212 may be selected from conductive particles, such as graphite particles, metal particles, etc. A large number of conductive particles are distributed in the foamed material, and some of them are interconnected to form a network-like conductive path. When the needle is inserted into the composite material, the network-like conductive path forms a conductive path with the needle surface. Figure 7 At the same time, this network-like conductive path also forms multiple points of dense contact with the conductive frame layer 22 on both sides of the flexible conductive layer 21. Thus, the network-like conductive path tightly connects the needle and the conductive frame layer 22 together electrically. The conductive particles are easy to process and have low cost, and have better dispersion, and are easier to be densely dispersed in the flexible conductive layer 21, thereby strengthening the contact with the needle.

[0056] The conductive embedding body 212 may also be a mixture of conductive fibers and conductive particles.

[0057] In some embodiments, the flexible conductive layer 21 is a conductive textile. Figure 8 As shown, the conductive textile can be a layered material formed by weaving fibers; Fig. 9 As shown, the conductive textile can also be a layered material formed by connecting fibers via non-woven fabrics. The fibers forming the conductive textile are at least partially conductive. The fibers in the conductive textile are interlaced and woven or non-woven together to form a three-dimensional network structure. In this structure, the direct physical contact between the fibers constitutes a large number of conductive connection nodes. In addition, the conductivity of the conductive fibers themselves can form a good electrical contact with the needle surface after the needle is inserted, thereby connecting the needle to the conductive frame layer 22. The conductive textile retains the soft characteristics of traditional textiles. Not only is it not easily damaged during acupuncture, it can also adapt to needles of different diameters and maintain good contact.

[0058] In some embodiments, Fig.10As shown, the conductive frame layer 22 includes a frame 221 and a contact net 222 fixed within the range of the frame 221; the frame 221 is electrically conductive with the contact net 222, and the contact net 222 is electrically conductive with the flexible conductive layer 21. The flexible conductive layer 21 is in full contact with the contact net 222, and the contact net 222 is electrically conductive with the frame 221, so that the current is fully collected to the frame 221. In this way, the frame 221 not only provides mechanical support, but also can be used as a wiring carrier to facilitate connection with the detection circuit to ensure the stability and accuracy of signal transmission. Moreover, the contact net 222 is in a grid shape, which can increase the contact area with the flexible conductive layer 21 and enhance the reliability and conductive performance of the electrical connection. In particular, during the acupuncture process, the contact net 222 can provide multiple contact points, and even if some contacts fail, there are still enough conductive paths to ensure signal transmission. On the other hand, the contact net 222 is in a grid shape and has stronger rigidity than the flexible conductive layer 21, which can improve the durability of the flexible conductive layer 21. Since students' hands inevitably shake during acupuncture practice, the shaking component perpendicular to the acupuncture direction will expand the needle holes on the flexible conductive layer 21. The larger the shaking component, the more obvious the hole expansion effect, and the more difficult it is to restore the holes in the flexible conductive layer 21 after the needle is pulled out. The grid-like contact net 222 can just limit the shaking component perpendicular to the acupuncture direction, thereby preventing large holes that are difficult to restore from being generated in the flexible conductive layer 21, ensuring that the flexible conductive layer 21 can still maintain good and stable contact with the needle after repeated acupuncture, thereby improving the durability of the flexible conductive layer 21. The mesh holes on the contact net 222 can be rectangular, circular, triangular, pentagonal, hexagonal, etc.; preferably, the area of ​​a single mesh hole is 1.0mm 2 ~10.0mm 2 within the range.

[0059] In some embodiments, the conductive frame layer 22 is connected in one piece, that is, the frame 221 and the contact network 222 are a whole. Preferably, the material can be copper, stainless steel, conductive plastic, galvanized steel, etc.

[0060] In some embodiments, Fig.10 As shown, on the side of the conductive frame layer 22 adjacent to the flexible conductive layer 21, the frame 221 protrudes from the contact mesh 222, thereby forming a recessed accommodation area within the range enclosed by the frame 221; the flexible conductive layer 21 is clamped in the accommodation area on both sides. The accommodation area within the frame 221 prevents the flexible conductive layer 21 from unnecessary displacement or loosening during use, ensuring the structural stability of the entire electrode assembly 2. The accommodation area can also better fix the flexible conductive layer 21 in its position, simplify the assembly steps in the manufacturing process, and improve production efficiency.

[0061] In some embodiments, Figure 1 and Figure 2As shown, the acupuncture depth sensor itself is integrated with a coding circuit board 3, which is preferably arranged on the side wall of the support body 1. Each electrode assembly 2 has a wiring terminal and is electrically connected to the coding circuit board 3 through the wiring terminal. The coding circuit board 3 has a detection end to reflect different coding states. When in use, as the acupuncture needle penetrates the acupuncture depth sensor, the arranged electrode assemblies 2 are penetrated layer by layer, resulting in a short circuit between the electrode assemblies 2 penetrated by the needle. Since each electrode assembly 2 is electrically connected to the coding circuit board 3, at least one electrical parameter of the coding circuit board 3 changes accordingly, and this change is reflected through the detection end.

[0062] Different encoding circuit boards 3 have different internal circuit connections, different detection end settings, and different electrical parameters for reflecting the acupuncture depth.

[0063] For example, in some embodiments, a resistor is provided on the coding circuit board 3; a resistor is connected between two adjacent electrode assemblies 2 arranged at intervals, and two electrode assemblies 2 at both ends of the electrode assemblies 2 arranged at intervals each form a detection end; the resistance value between the two detection ends changes as the needle is inserted to different depths, causing different numbers of resistors to be short-circuited. In the initial state, that is, when there is no needle insertion, the resistance between all electrode assemblies 2 is independent, and the total resistance value of the detection end is composed of all series resistors. When the needle begins to penetrate, it first contacts the first electrode assembly 2 and forms an electrical connection with it. As the needle continues to penetrate, it will contact more electrode assemblies 2 in turn, and the originally independent resistors between these electrode assemblies 2 are short-circuited by the needle, forming a new electrical connection path, resulting in a change in the total resistance value. There is a corresponding relationship between this change in resistance value and the depth of needle insertion. The deeper the needle is inserted, the more electrode assemblies 2 the needle will contact, resulting in more resistors being short-circuited. Using this corresponding relationship, the depth of needle insertion can be obtained according to the resistance value between the two detection ends on the coding circuit board 3.

[0064] For example, in some embodiments, a capacitor is provided on the coding circuit board 3; capacitors are connected between two adjacent electrode assemblies 2 arranged at intervals, and two electrode assemblies 2 at both ends of the electrode assemblies 2 arranged at intervals each form a detection end; the capacitance value between the two detection ends changes as the needle is inserted to different depths, causing different numbers of capacitors to be short-circuited. In the initial state, that is, when there is no needle insertion, the capacitance between all electrode assemblies 2 is independent, and the total capacitance value of the detection end is composed of all parallel capacitors. When the needle begins to penetrate, it first contacts the first electrode assembly 2 and forms an electrical connection with it. As the needle continues to penetrate, it will contact more electrode assemblies 2 in turn, and the originally independent capacitances between these electrode assemblies 2 are short-circuited by the needle, forming a new electrical connection path, thereby changing the total capacitance value. There is a corresponding relationship between this change in capacitance value and the depth of needle insertion. The deeper the needle is inserted, the more electrode assemblies 2 the needle will contact, resulting in more capacitors being short-circuited. Using this corresponding relationship, the depth of needle insertion can be obtained according to the change in capacitance value between the two detection ends on the coding circuit board 3.

[0065] In some other embodiments, other electrical components may also be used, such as connecting an inductor between two adjacent electrode assemblies 2, and then encoding based on the inductor and other electrical parameters corresponding to the electrical components. The insertion of the needle causes the electrical parameters at the detection end of the encoding circuit board 3 to change to encode the acupuncture depth information, and then decode this information based on the corresponding relationship between the electrical parameters and the acupuncture depth to determine the actual acupuncture depth. This makes the measurement of the acupuncture depth more accurate and reliable, and also allows such measurement to be automated and programmed.

[0066] The second aspect is the acupuncture posture detection system.

[0067] The acupuncture posture detection system can provide real-time feedback on acupuncture depth and angle information. Based on this information, it can give an objective and accurate evaluation of whether the acupuncture operation skills meet the standards, which is of great help to professionals and students in learning and improving acupuncture skills.

[0068] The acupuncture posture detection system includes the acupuncture depth sensor of the first aspect and the acupuncture training needle 5. As described above, the acupuncture depth sensor can accurately feedback the acupuncture depth when the acupuncture needle is inserted. Fig.11As shown, an angle detection element is provided at the end, and an angle detection module 52 is built in the angle detection element to detect the acupuncture angle of the acupuncture training needle 5; the acupuncture depth sensor detects the acupuncture depth of the acupuncture training needle 5. Specifically, the acupuncture training needle 5 includes a needle body 51 and an angle detection module 52 arranged on the top of the needle body 51, and the angle detection module 52 includes a gyroscope sensor, a power supply, a wireless charging receiving unit and a wireless transmission unit; the gyroscope sensor monitors the orientation angle data, the power supply supplies power to the gyroscope sensor and the wireless transmission unit, the wireless charging receiving unit receives energy from the external wireless charging transmitting unit to charge the power supply, and the wireless transmission unit outputs the orientation angle data monitored by the gyroscope sensor.

[0069] Among them, the angle detection module 52 can effectively monitor and transmit angle data through the cooperation of gyroscope sensor, built-in power supply, wireless charging technology and wireless transmission unit, so as to provide instant feedback to the trainer. The gyroscope sensor is an existing device for measuring direction. It is used here to monitor the angle change during acupuncture, so as to evaluate the accuracy of the acupuncture angle. The power supply is used for the gyroscope and the wireless transmission unit. Considering the portability and ease of use, it is better to use a built-in power supply. The wireless charging receiving unit receives energy from the external wireless charging transmitting unit to charge the power supply, and receives energy from the external wireless charging transmitting unit through the principle of electromagnetic induction to charge the built-in power supply. This not only simplifies the charging process, but also improves the overall waterproof and dustproof capabilities and reduces the risk of damage to the physical interface.

[0070] The gyroscope sensor preferably uses a miniature six-axis gyroscope sensor, which integrates a three-degree-of-freedom gyroscope (used to measure angular velocity) and a three-degree-of-freedom accelerometer (used to measure linear acceleration), which can provide data in six dimensions: rotation (Pitch) in the X-axis direction, rotation (Roll) in the Y-axis direction, rotation (Yaw) in the Z-axis direction, acceleration in the X-axis direction, acceleration in the Y-axis direction, and acceleration in the Z-axis direction. This not only provides acupuncture angle information to help determine the accuracy of the acupuncture angle, but also provides needle acceleration information, which can be used to evaluate the force control during acupuncture. For example, when the acceleration of the acupuncture needle in a certain direction is too large, it indicates that the acupuncture force is too strong, which may affect the patient's comfort or even cause harm. In addition, acceleration data can also help identify whether the acupuncture movement is smooth. If the acceleration fluctuation exceeds the preset range, it indicates that the jitter is obvious and the acupuncture operation is unstable.

[0071] In some embodiments, the acupuncture posture detection system further comprises a signal processing unit, which amplifies and processes the sampled data measured by the gyroscope sensor through an unscented Kalman filter to obtain a more accurate and reliable result.

[0072] The dynamic data from the gyro sensor is processed using an unscented Kalman filter (UKF). First, the state vector and state covariance matrix are initialized to represent the best guess of the current state and its uncertainty. A set of weighted sigma points are selected based on the current state vector and covariance matrix. These points are distributed around the current state mean, and their positions are determined by the covariance matrix, and the weights reflect the importance of each point. Passing each sigma point to the next moment through a nonlinear dynamic model (state transfer equation) is a nonlinear way to process state transfer. The predicted state mean and prediction covariance are reconstructed using the predicted sigma points; the predicted sigma points are converted into points in the observation domain using the observation model; the observation prediction mean and observation prediction covariance are reconstructed; the cross covariance is calculated, which is the correlation between the state prediction and the observation prediction; and the Kalman gain is calculated to correct the state prediction in combination with the observation information. The state prediction is corrected using the Kalman gain to obtain the final state estimate. The state covariance matrix is ​​updated to reflect the uncertainty of the state estimate. Since UKF avoids the linearization step in EKF, it can handle nonlinear systems more accurately, thereby improving the accuracy of acupuncture posture data estimation. In addition, this method is more robust to nonlinear data such as acupuncture posture and can give a more accurate and reliable estimate.

[0073] The data of the acupuncture posture detection system will eventually be fed back to the host computer through the serial port. The interactive software of the host computer will display the posture of the acupuncture needle in real time based on the acupuncture depth data and acupuncture angle data. The trainer can use the interactive software to timely understand his or her acupuncture training status and make timely adjustments and improvements.

[0074] In other embodiments, a framework for optimizing the unscented Kalman filter is used for data processing to achieve effective use of the underlying physical laws of the system, automate the selection of the covariance matrix, and solve the performance problem of the UKF in the case of limited data.

[0075] Although the application of the unscented Kalman filter algorithm can achieve good results under normal circumstances, it does not mean that it can effectively track all states of the system, especially when sensor data is scarce, or when the model has potential states such as false touch and looseness, the performance will decrease significantly.

[0076] To address the above issues, a new framework is used to adjust the hyperparameters of the UKF by constructing a loss function that follows the governing equations of the system being analyzed, thereby establishing a physically aware optimization framework. The optimizer is used to force the estimation of states and parameters to satisfy the governing equations of the system. This approach is particularly effective when data is limited, especially when there is a deep understanding of the underlying physics of the system. In addition, the method can automatically select appropriate values ​​and structures for the Kalman covariance matrix during the filtering process, greatly alleviating the constraints on the covariance structure.

[0077] The underlying physical laws mainly solve data acquisition and noise problems, covering aspects such as acupuncture depth, needle deformation, and dynamic behavior of sensors. First of all, the acupuncture depth and tissue resistance show regular changes. The resistance is different when passing through the skin, muscle and other layers, which is the key basis for simulating the real environment. During the puncture process, the acupuncture needle will produce flexible deformation due to material, length and strength. This deformation will affect the posture and depth of acupuncture, and the system needs to provide accurate feedback accordingly. In addition, the posture sensor and depth sensor collect data in real time, and their dynamic sampling frequency and error model are also part of the physical laws of the system. To ensure the accuracy of the feedback, the system needs to deal with the noise and delay problems in the data. Overall, these physical laws and data acquisition links jointly construct the control model of the dynamic system.

[0078] The model can be abstracted as a noisy discrete-time state-space system. The state space is characterized by a state transition function M and a measurement function H. At time step k, the state-space model provides a state prediction x one step ahead k+1 and the predicted measurement y ’ k , the model is as follows:

[0079] x k+1 = M(x k , u k , θ) + v k , v k ~ N(0, Q)

[0080] y ’ k = H(x k , u k , θ) + μ k , μ k ~ N(0, R)

[0081] Among them, x k is the state vector at time step k, u k is an external input variable, the random variable v k is the process error vector, sampled from a central Gaussian distribution with covariance Q, and the random variable μk is the measurement error vector, sampled from a centered Gaussian distribution with covariance R, and the vector θ contains a set of dynamic parameters that characterize the physical system.

[0082] The goal of algorithmic reverse modeling is to use a set of observations {y k}, k∈K to reconstruct the state space as described above. To this end, it is necessary not only to restore the state {x k},k∈K It is also necessary to recover the parameters θ at the same time. The solution to this problem is to use existing state estimation techniques to expand the state into an enhanced state, that is, to process it by combining the state and the parameters in one vector. For time-invariant systems, the enhanced state transfer function is written as M^a=[M,I], where I is a unit vector. Since the dynamic parameters are implicit, the measurement function and the measurement error covariance remain unchanged.

[0083] In the process of testing and optimization, in order to solve the black box problem of complex objective functions, the NOMAD derivative-free optimization algorithm is used. In actual operation, it is found that the optimization space is extremely discontinuous and noisy, and NOMAD happens not to rely on derivatives. One of the main features of NOMAD is its adaptive grid refinement mechanism, which enables it to find a balance between extensive search and detailed local exploration. Since inverse problems are often ill-posed and UKF optimization may encounter multiple local optimal solutions, NOMAD's adaptive grid refinement mechanism can effectively solve this problem.

[0084] To adjust the hyperparameters of UKF, the loss function based on mean square error (MSE) and L2-norm is used for optimization as follows:

[0085] L data = ∑(from k1 to k2) ||y k -y ’ k ||2.

[0086] However, this criterion makes redundant use of the UKF and the data used to determine the best hyperparameters in the outer loop, so the state transfer function can be further used to construct the loss function, using the UKF to estimate the parameters and state. This imposes a strong regularization on the problem and focuses on estimates that satisfy the laws of physics. The resulting Physic-Aware (PA) loss function is expressed as:

[0087] L physics = ∑(from k1 to k2) ||x ’ k+1 -M(x ’ k , u k, θ)||2.

[0088] In the context of the Kalman filter, the PA loss function above quantifies the computational model so that it still satisfies the physical constraints to the vast majority of the extent possible. In addition, it takes into account not only the observed values, but also the underlying state of the system. When applied to the augmented state, the loss of the PA loss function not only forces the model to conform to the estimated values, but also enforces stability during parameter estimation. Depending on the choice of k1 and k2, the function can be used to optimize the accuracy of the estimate or to optimize the convergence speed.

[0089] When the hardware system was actually assembled, the control group that was not optimized was compared with the experimental group that used the optimization algorithm to confirm that the data accuracy, noise resistance and actual data fitting degree of the experimental group were significantly improved.

[0090] It can be seen that the use of optimized unscented Kalman filtering for data processing has at least the following three advantages compared to the unoptimized unscented Kalman filtering:

[0091] (1) Solving the performance problem of UKF in the case of limited data: The traditional unscented Kalman filter has poor optimization effect when sensor data is scarce or limited, and may not accurately track all states of the system. The optimization framework adjusts the hyperparameters of UKF by constructing a loss function that follows the system control equation, improving its applicability in the case of limited data.

[0092] (2) Effective use of the underlying physical laws of the system: In the acupuncture posture detection system, if the underlying physical laws of the system are deeply understood, the optimization framework can better utilize this physical information, thereby improving the effect of data optimization. This method performs particularly well in systems with complex models or with strong reliance on underlying physics.

[0093] (3) Automated covariance matrix selection: The optimization framework can automatically select appropriate values ​​and structures for the Kalman covariance matrix during the filtering process, avoiding the complexity and uncertainty of manual settings. This automated selection reduces the constraints on the covariance structure and improves the flexibility and adaptability of the system.

[0094] The third aspect is acupuncture teaching equipment.

[0095] The acupuncture teaching tool includes a simulated human body model 6 and an acupuncture posture detection system according to the second aspect. Fig.12 , Fig.13 and Fig.14 As shown, the surface of the simulated human body model 6 has a simulated skin layer 61, and the simulated skin layer 61 is elastic and penetrable; an acupuncture depth sensor is buried under the simulated skin layer 61, as shown in FIG. Fig.14 Shown in part A.

[0096] The simulated human body model 6 can be a complete human body model, including a simulated trunk, simulated buttocks, simulated upper limbs, simulated lower limbs, and simulated head, or a partial human body model. The simulated human body model 6 is provided with simulated acupuncture points corresponding to real acupuncture points of the human body. The simulated acupuncture points include a simulated skin layer 61 located on the surface of the simulated human body model 6, a simulated subcutaneous tissue 62 located under the simulated skin layer 61, and an acupuncture depth sensor embedded in the simulated subcutaneous tissue 62 and located under the simulated skin layer 61.

[0097] The simulated skin layer 61 may be made of a silicone layer, which has good flexibility and durability, and has a certain degree of self-closing property for the small holes formed by acupuncture, so it can better simulate the appearance and touch of human skin and provide acupuncture resistance similar to real skin. The simulated subcutaneous tissue 62 is elastic and may be filled with elastic foam, which can simulate the real feel of human subcutaneous tissue, so that the user can obtain a touch and elasticity closer to the human body when performing acupuncture, thereby increasing the realism of the operation.

[0098] This acupuncture teaching tool provides a practical platform for acupuncture learners by simulating a real acupuncture environment on site, allowing them to practice and improve their acupuncture techniques without involving real patients. When the trainer uses the acupuncture training needle 5 to practice on the simulated human body model 6, the acupuncture depth sensor will record the depth of the needle insertion. At the same time, the angle detection module 52 of the acupuncture training needle 5 will record the angle of acupuncture and send data through the wireless transmission unit. These data will be transmitted to the interactive software of the host computer, and the software will display the posture of the acupuncture needle in real time based on the received acupuncture depth and angle data. The trainer can understand the situation of his or her acupuncture operation by observing the interactive screen of the interactive software and querying the data, and adjust the technique and strength accordingly, and continuously improve the acupuncture operation technique.

[0099] As described above, the simulated human body model 6 is provided with simulated acupuncture points corresponding to the real acupuncture points of the human body. Fig.12 and Fig.13 As shown, simulated acupoints are provided on the simulated human body model 6 and correspond to the following acupoints of the human body: Yintang 601, Shuigou 602, Tanzhong 603, Qihai 604, Quze 605, Laogong 606, Zusanli 607, Taichong 608, Baihui 609, Dazhui 610, Huantiao 611, Quchi 612, Hegu 613, Weizhong 614, Kunlun 615.

[0100] The above embodiments are exemplary, and their purpose is to illustrate the technical concept and features of the present invention so that people familiar with the technology in this field can understand the content of the present invention and implement it accordingly, and they cannot be used to limit the protection scope of the present invention. Any equivalent changes or modifications made according to the spirit of the present invention should be included in the protection scope of the present invention.

Claims

1. A needle puncture depth sensor, characterized in that: The invention comprises a support body (1) and a plurality of layered electrode assemblies (2) arranged in the support body (1); the support body (1) has a storage space inside, and a plurality of the electrode assemblies (2) are arranged in an interval in the storage space; the electrode assemblies (2) are connected to a detection circuit, and each of the electrode assemblies (2) comprises a flexible conductive layer (21) and a conductive frame layer (22) sandwiched between two sides of the flexible conductive layer (21); at least one end of the support body (1) has an opening for a conductive needle to penetrate into the support body (1) and electrically contact the electrode assembly (2), thereby changing the electrical connection state of the electrode assembly (2) within the needle penetration depth range; The conductive frame layer (22) comprises a frame (221) and a contact network (222) fixed within the frame (221); the frame (221) and the contact network (222) are electrically connected, and the contact network (222) and the flexible conductive layer (21) are electrically connected; the conductive frame layer (22) provides support for the flexible conductive layer (21) on both sides to prevent the flexible conductive layer (21) from excessive deformation.

2. The acupuncture depth sensor according to claim 1, characterized in that: The flexible conductive layer (21) is a foamed composite material layer having elasticity, comprising a foamed material matrix (211) and a conductive embedded body (212) distributed in the foamed material matrix (211); the conductive embedded body (212) is selected from one or more of conductive fibers and conductive particles.

3. The acupuncture depth sensor according to claim 1, characterized in that: The flexible conductive layer (21) is a conductive textile, which is a layered material formed by connecting fibers through weaving or non-woven processes; the fibers forming the conductive textile are at least partially conductive.

4. The acupuncture depth sensor according to claim 1, characterized in that: On a side of the conductive frame layer (22) that is adjacent to the flexible conductive layer (21), the frame (221) protrudes from the contact network (222), thereby forming a recessed accommodation area within a range enclosed by the frame (221); the flexible conductive layer (21) is clamped in the accommodation areas on both sides.

5. The acupuncture depth sensor according to any one of claims 1 to 4, characterized in that: It also comprises a coding circuit board (3), and each of the electrode assemblies (2) is electrically connected to the coding circuit board (3); at least one electrical parameter between the detection ends of the coding circuit board (3) changes as the electrode assemblies (2) are punctured to different depths.

6. The acupuncture depth sensor according to claim 5, characterized in that: At least one resistor is provided on the encoding circuit board (3); resistors are connected between two adjacent ones of the plurality of electrode assemblies (2) arranged at intervals, and two of the plurality of electrode assemblies (2) arranged at both ends each form a detection end; the resistance value between the two detection ends changes as different numbers of resistor short circuits are caused by acupuncture to different depths.

7. The acupuncture depth sensor according to claim 5, characterized in that: At least one capacitor is provided on the encoding circuit board (3); capacitors are connected between two adjacent ones of the plurality of electrode assemblies (2) arranged at intervals, and two of the plurality of electrode assemblies (2) arranged at both ends each form a detection end; the capacitance value between the two detection ends changes as different numbers of capacitor short circuits are caused by needle puncture to different depths.

8. Acupuncture posture detection system, characterized by: The invention comprises the acupuncture depth sensor according to any one of claims 1 to 7 and an acupuncture training needle (5); an angle detection element is provided at the end of the acupuncture training needle (5), and an angle detection module (52) is built in the angle detection element to detect the acupuncture angle of the acupuncture training needle (5); and the acupuncture depth sensor detects the acupuncture depth of the acupuncture training needle (5).

9. The acupuncture posture detection system according to claim 8, characterized in that: The acupuncture training needle (5) comprises a needle body (51) and an angle detection module (52) arranged on the top of the needle body (51); the angle detection module (52) comprises a gyroscope sensor, a power supply, a wireless charging receiving unit and a wireless transmission unit; the gyroscope sensor monitors orientation angle data, the power supply supplies power to the gyroscope sensor and the wireless transmission unit, the wireless charging receiving unit receives energy from an external wireless charging transmitting unit to charge the power supply, and the wireless transmission unit outputs the orientation angle data monitored by the gyroscope sensor.

10. An acupuncture teaching tool, characterized in that: The invention comprises a simulated human body model (6) and the acupuncture posture detection system according to claim 9; the surface of the simulated human body model (6) has a simulated skin layer (61), and the simulated skin layer (61) has elasticity and penetrability; the acupuncture depth sensor is buried under the simulated skin layer (61).

11. The acupuncture teaching tool according to claim 10, characterized in that: The acupuncture posture detection system feeds back and outputs the data measured by the gyro sensor after amplification and unscented Kalman filtering.

12. The acupuncture teaching tool according to claim 10, characterized in that: The acupuncture posture detection system feeds back and outputs the data measured by the gyro sensor after amplification and optimized unscented Kalman filtering. The optimized unscented Kalman filter process adjusts hyperparameters by constructing a loss function that follows the control equations of the analyzed system, thereby establishing a physically aware optimization framework; using an optimizer to force the estimation of states and parameters to satisfy the control equations of the dynamic system; The governing equations of a dynamic system are represented as a noisy discrete-time state-space system. The state space contains the state transition function M and the measurement function H. At time step k, the state-space model provides a one-step-ahead state prediction x k+1 and the predicted measurement y ’ k , the model is as follows: x k+1 = M(x k , u k , θ) + v k , v k ~ N(0, Q) y ’ k = H(x k , u k , i) + m k , m k ~ N(0, R) Among them, x k is the state vector at time step k, u k is an external input variable, the random variable v k is the process error vector, sampled from a central Gaussian distribution with covariance Q, and the random variable μ k is the measurement error vector, sampled from a central Gaussian distribution with covariance R, and the vector θ contains a set of dynamic parameters that characterize the physical system; The enhanced state transition function is written as M^a=[M,I], where I is a unit vector; During the test optimization process, the NOMAD derivative-free optimization algorithm is used to solve the black box problem of the objective function; The optimization is performed using the loss function based on mean square error and L2-norm as follows: L data = ∑(from k1 to k2) ||y k -y ’ k ||2 Using the state transfer function to construct the loss function, the resulting physical perception loss function is expressed as: L physics = ∑(from k1 to k2) ||x ’ k+1 -M(x ’ k , u k , θ)||2.

Images