A pulse diagnosis dexterous hand, pulse diagnosis system and pulse collection method
By designing a clever hand and pulse diagnosis system for pulse diagnosis, using dynamic and static pressure sensing modules, robotic arms and deep learning models, the precise positioning and multi-dimensional acquisition of pulse wave data is achieved, and the data accuracy and accuracy in the existing technology are solved. The problem of standardization improves the quantitative ability and data reliability of pulse diagnosis.
Patent Information
- Application Number
- CN202211401236.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-11-09
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2042-11-09
AI Technical Summary
The existing pulse wave acquisition system is difficult to achieve accurate positioning and data standardization, resulting in insufficient accuracy and reliability of the collected data.
A smart pulse diagnosis hand is designed, including a handrail, finger motor, dynamic and static pressure sensing module, combined with a robotic arm and a four-degree of freedom moving device, to achieve precise positioning and Multi-dimensional pulse wave data acquisition.
It realizes accurate positioning and multi-point acquisition of the human wrist, ensures the accuracy and reliability of pulse wave data, supports the standardized accumulation of data, and improves the quantitative ability of pulse diagnosis.
Smart Images

Figure CN115778332B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of medical devices, and in particular relates to a pulse diagnosis dexterous hand, a pulse diagnosis system and a pulse collection method thereof. Background Art
[0002] Common diagnostic methods used in traditional Chinese medicine include observation, auscultation, questioning, and palpation, among which pulse diagnosis is one of the important diagnostic methods in traditional Chinese medicine. Pulse diagnosis by fingers in traditional Chinese medicine is to apply different pressures to the radial artery tube with the fingertips, i.e. floating, medium, and sinking pressures, and rely on tactile, pressure, and vibration receptors to detect pulse information. The accuracy of traditional Chinese medicine doctors' pulse diagnosis depends on the doctor's state. When the doctor's state is unstable, the accuracy of pulse diagnosis will also decrease. Pulse diagnosis in traditional Chinese medicine is difficult to quantify, and the pulse diagnosis techniques, pulse diagnosis feelings, and pulse diagnosis judgments of every Chinese medicine doctor are exactly the same, and there is a strong individuality. If pulse diagnosis is not quantified, it is difficult to have a basis for unified analysis.
[0003] At present, most pulse wave acquisition systems mainly rely on manual positioning methods to find the position of the wrist joint. The positioning process is completely dependent on a single manual operation without feedback information, which is prone to inaccurate positioning. In addition, it is difficult for existing pulse acquisition systems to form a stable and accurate closed loop of positioning, perception and data acquisition, which can neither guarantee the quality and reliability of the collected data nor achieve standardization of accumulated data. Summary of the invention
[0004] The purpose of the present invention is to overcome the defects of inaccurate positioning and data collection in the prior art, and to provide a pulse diagnosis dexterous hand, a pulse diagnosis system and a pulse collection method that can accurately locate, have reliable quality and achieve data standardization.
[0005] The technical solution adopted by the present invention to solve the technical problem is: a pulse diagnosis dexterous hand, comprising a hand frame, a mounting hole in the vertical direction provided at the end of the hand frame, and a finger assembly installed in the mounting hole; the finger assembly comprises a finger motor installed on the hand frame, a 7-shaped top column with one end connected to the output shaft of the finger motor and the other end installed in the mounting hole, a bushing sleeved in the mounting hole, and a flexible substrate connected to the bushing; a connecting tube is arranged between the bushing and the mounting hole, a pressure spring is arranged between the inner bottom surface of the bushing and the lower end surface of the 7-shaped top column, and a counter-pressure spring is arranged between the outer shaft shoulder of the bushing and the inner shaft shoulder of the connecting tube;
[0006] At least one static pressure sensing module is arranged at the connection between the flexible substrate and the bushing. The surface of the flexible substrate in contact with the human wrist is an arc surface, and a plurality of dynamic pressure sensing modules are arranged on the arc surface.
[0007] Furthermore, there is one static pressure sensing module; there are nine dynamic pressure sensing modules, which are arranged in two columns, one column with five and the other column with four, arranged on the curved surface of the flexible substrate.
[0008] Furthermore, the finger components are divided into three groups, corresponding to the Cun, Guan and Chi positions of the human wrist respectively.
[0009] A pulse diagnosis system, comprising a mechanical arm mounted on a pulse diagnosis platform, a four-degree-of-freedom mobile device mounted on the mechanical arm, a dexterous hand mounted on the output end of the four-degree-of-freedom mobile device, a main control module and a host computer; the dexterous hand is the pulse diagnosis dexterous hand according to any one of claims 1 to 3 above;
[0010] The end of the mechanical arm is equipped with a lighting module, an infrared image acquisition module for collecting photos of the human wrist, and a laser distance measurement module for measuring the distance between the dexterous hand and the human wrist; the lighting module, the infrared image acquisition module and the laser distance measurement module are arranged correspondingly to the dexterous hand;
[0011] The main control module is communicatively connected with the four-degree-of-freedom mobile device, the dexterous hand, the lighting module and the laser ranging module; the main control module and the infrared image acquisition module are both communicatively connected with the host computer.
[0012] Specifically, the lighting module is a combination of multiple LED light groups, and each LED light group is controlled individually.
[0013] A composite pulse wave acquisition method for a pulse diagnosis system, characterized in that it comprises the following steps:
[0014] S1, using a laser ranging module to measure the vertical distance from the end of the robotic arm to the wrist of the person being measured;
[0015] S2, obtaining multiple images of the subject's wrist under different lighting angles;
[0016] S3, the deep learning model calculates the wrist joint position and the three-dimensional structure of the wrist of the subject according to the image information obtained in S1;
[0017] S4, taking the position of the joint calculated in S3 as the target point, controlling the four-degree-of-freedom mobile device of the pulse diagnosis system to move, and using the output end of the four-degree-of-freedom mobile device to press the multiple dynamic pressure sensing modules on the dexterous hand of the subject's wrist perpendicular to the target point position;
[0018] S5, pre-collecting multi-dimensional pulse wave data from the wrist of the subject;
[0019] S6, transmitting the data signal pre-collected in step S5 to the host computer, and the host computer (9) compares the pre-collected pulse wave data with the typical feature library of the pulse wave to determine the accuracy of the joint position;
[0020] S7, if the pulse wave data pre-collected in step S6 contains typical features of the pulse wave, the center position and the position of the joint of the dynamic pressure sensor device corresponding to the joint are accurately located; the main control module makes micro-adjustments to the positions of the multiple dynamic pressure sensor devices on the finger assembly corresponding to the joint position according to the feedback signal output by the host computer;
[0021] S8, after determining the positions of the multiple dynamic pressure sensing devices on the finger assembly corresponding to the joint position and the precise positioning of the joint position, the pulse wave of the subject's wrist is finally collected in multiple dimensions, and the collected data is transmitted to the host computer.
[0022] Furthermore, in step S6, if the collected data does not contain the typical features of the pulse wave, it is determined that the positioning of the joint is inaccurate, and the operations of steps S1 to S4 are repeated to reposition the joint or perform manual assisted positioning; if the collected data contains the typical features of the pulse wave, it is determined that the positioning of the joint is accurate, and the operation of step S7 is performed.
[0023] Furthermore, in step S7, the host computer uses a deep learning model of time series to analyze the pre-collected multi-channel pulse wave data in real time, and fits the peak value of the real-time pulse wave data into the pulse width curve of the subject. If the maximum peak data of the pulse width curve is not at the center position, the host computer feeds back the analyzed signal to the main control module, and the main control module outputs a control signal to control the four-degree-of-freedom mobile device to fine-tune the center position of the dynamic pressure sensing module to the key position.
[0024] Furthermore, in step S8, the host computer analyzes the collected data to extract parameters and imports them into the prediction model in combination with the pulse width, pulse position information and the subject's physiological parameters to predict the results.
[0025] Further, in step S5, the main control module sets the collected pre-pressure value of each finger assembly in the dexterous hand and calculates the optimal pressure; the collected pre-pressure value is a plurality of gradient pressure values, and the optimal pressure collection time is greater than the collection time of other pressure gradient values; the collected pre-pressure values are ten pressure gradient values of 50 / 70 / 90 / 110 / 130 / 150 / 170 / 190 / 210 / 230gf respectively.
[0026] The beneficial effects of the pulse diagnosis dexterous hand, pulse diagnosis system and pulse collection method of the present invention are:
[0027] 1. The pulse diagnosis dexterous hand of the present invention is provided with three groups of finger components, which correspond to the cun, guan and chi positions of the human wrist respectively, and each group of finger components is provided with at least one static pressure sensing module and multiple dynamic pressure sensing modules, so that the pulse diagnosis dexterous hand can collect data from multiple points of the human wrist, and fit the data collected from multiple points to the pulse width curve of the subject to ensure the accuracy of the collected data. At the same time, the surface in contact with the human wrist is an arc surface, which conforms to the natural state of the wrist, and further ensures the accuracy of the collected data.
[0028] 2. The pulse diagnosis system of the present invention includes a robotic arm, a dexterous hand, an infrared image acquisition module, a main control module and a host computer, etc. The main control module is communicatively connected with the four-degree-of-freedom mobile device, the dexterous hand, the lighting module and the laser ranging module. The main control module and the infrared image acquisition module are both communicatively connected with the host computer to realize a closed-loop system of pulse wave positioning, collection and feedback, which can provide data reference for realizing the standardization of accumulated data, is conducive to the quantification of pulse diagnosis and can ensure the accuracy and reliability of quantification.
[0029] 3. The composite pulse wave acquisition method of the pulse diagnosis system of the present invention determines the joint position by obtaining multiple image information of the subject's wrist under different lighting angles, and pre-collects the wrist, and compares the pre-collected data with the typical feature library of the pulse wave. The accuracy of the joint position can be determined multiple times, avoiding the inaccuracy and unreliability of the prior art that requires a single medical staff to determine the joint position.
[0030] 4. The host computer in the acquisition method of the composite pulse wave of the pulse diagnosis system of the present invention extracts and analyzes and predicts the pre-acquired characteristic parameters, uses the deep learning model of time series to analyze the collected multi-channel pulse wave data in real time, and fits the peak value of the real-time pulse wave data into the pulse width curve of the subject. And the pulse width curve with the maximum peak value not in the center position is fed back to the main control module, and the main control module fine-tunes the position of the dexterous hand to ensure that the position of the joint and the corresponding dynamic pressure sensor device can be accurately positioned. The entire acquisition method forms a closed system of signal acquisition, calculation analysis, feedback, adjustment and acquisition output, which effectively ensures the quality and reliability of the collected data. BRIEF DESCRIPTION OF THE DRAWINGS
[0031] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.
[0032] Figure 1 is a three-dimensional diagram of the pulse diagnosis dexterous hand according to an embodiment of the present invention;
[0033] Figure 2 is a top view of the pulse diagnosis dexterous hand according to an embodiment of the present invention;
[0034] Figure 3 is a cross-sectional view of the pulse diagnosis dexterous hand according to an embodiment of the present invention;
[0035] Figure 4 is an installation diagram of a static pressure sensing module and a dynamic pressure module of the pulse diagnosis dexterous hand according to an embodiment of the present invention;
[0036] Figure 5 is a layout diagram of a dynamic pressure module on a flexible substrate according to an embodiment of the present invention;
[0037] Figure 6 is a schematic structural diagram of a pulse diagnosis system according to an embodiment of the present invention;
[0038] Figure 7 is a front view of a pulse diagnosis system according to an embodiment of the present invention;
[0039] Figure 8 is a control diagram of a pulse diagnosis system according to an embodiment of the present invention;
[0040] Fig. 9 is a flow chart of a composite pulse wave acquisition method according to an embodiment of the present invention;
[0041] Fig.10 is a partial image acquired by the infrared image acquisition module of an embodiment of the present invention;
[0042] Fig.11 It is a specific operation flow chart of the collection method according to an embodiment of the present invention.
[0043] In the figure: 1. Pulse diagnosis platform, 2. Robotic arm, 3. Four-degree-of-freedom mobile device, 4. Dexterous hand, 41. Hand frame, 43. Finger assembly, 431. Finger motor, 432. 7-shaped top column, 433. Bushing, 434. Flexible substrate, 435. Pressure spring, 436. Back pressure spring, 437. Static pressure sensing module, 438. Dynamic pressure sensing module, 439. Metal connector, 4310. Arc surface, 4311. Connecting tube, 4312. Fixing part, 4313. Circuit board, 5. Lighting module, 6. Infrared image acquisition module, 7. Laser ranging module, 8. Main control module, 9. Host computer, 10. Network interface. DETAILED DESCRIPTION
[0044] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.
[0045] like Figure 1-Figure 5The specific embodiment of a pulse diagnosis dexterous hand 4 of the present invention shown in the figure comprises a hand frame 41, a mounting hole in the vertical direction opened at the end of the hand frame 41, and a finger assembly 43 installed in the mounting hole; the finger assembly 43 comprises a finger motor 431 installed on the hand frame 41, a 7-shaped top column 432 with one end connected to the output shaft of the finger motor 431 and the other end installed in the mounting hole, a bushing 433 sleeved in the mounting hole, and a flexible substrate 434 connected to the bushing 433; a pressure spring 435 is arranged between the inner bottom surface of the bushing 433 and the lower end surface of the 7-shaped top column 432, and a counter-pressure spring 436 is arranged between the outer shoulder of the bushing 433 and the inner shoulder of the mounting hole; at least one static pressure sensing module 437 is arranged at the connection between the flexible substrate 434 and the bushing 433, and the surface of the flexible substrate 434 in contact with the human wrist is an arc surface 4310, and a plurality of dynamic pressure sensing modules 438 are arranged on the arc surface 4310.
[0046] The output shaft of the finger motor 431 is connected to the 7-shaped top column 432 through a screw, a connecting tube 4311 is arranged between the bushing 433 and the mounting hole, and the flexible substrate 434 is connected to the bushing 433 through a metal connector 439, wherein the static pressure sensor device is installed on the outside of the metal connector 439, and when the finger motor 431 shaft is recovered to apply downward pressure to the 7-shaped top column 432, the static pressure sensor device is pushed downward from the connecting tube 4311; the pressure spring 435 in direct contact with the 7-shaped top column 432 and the back pressure spring 436 arranged at the outer shoulder of the bushing 433 constitute the dynamic spring of the finger assembly 43, and when the 7-shaped top column 432 transmits pressure, a dynamic force balance system can be formed to reduce the damage of the pressing device and facilitate closed-loop force control. A fixing member 4312 is arranged above the three 7-shaped top columns 432, and a circuit board 4313 is installed on the hand frame for communication connection between the finger mechanism 43 and the main control module 8.
[0047] It should be noted that the dexterous hand 4 of the present invention adopts a combined dynamic and static pressure sensing module. The static pressure sensing module 437 adopts a static pressure sensor as a piezoresistive sensor, which is connected by screws at the end of the metal connector 439 and has a sensing accuracy of 0.02gf; the dynamic pressure sensing device is a dynamic pressure sensor. In this embodiment, there is one static pressure sensing module 437; there are 9 dynamic pressure sensing modules 438, which are arranged in two columns, one column with five and the other column with four, arranged on the curved surface 4310 of the flexible substrate 434. There are 9 dynamic pressure sensing modules 438, which are installed in a 5+4 flexible array on the curved surface 4310 of the flexible substrate 434. The dynamic pressure sensor is a capacitive sensor, the width of a single sensor channel is 0.6mm, and the interval between each sensor channel is 0.2mm. The thickness of the flexible substrate 434 is about 3mm, and it is attached to the metal connector 439 by bonding. The 9 dynamic pressure sensors arranged in a flexible array are attached to the flexible substrate 434, and the pulse wave information of a single part can be collected by attaching it to the wrist of the subject.
[0048] It should be noted that the finger assembly 43 is divided into three groups, corresponding to the Cun, Guan, and Chi positions of the human wrist. The pulse diagnosis dexterous hand 4 of the present invention is provided with three groups of finger assemblies 43, corresponding to the Cun, Guan, and Chi positions of the human wrist, and each group of finger assemblies 43 is provided with at least one static pressure sensing module 437 and a plurality of dynamic pressure sensing modules 438, so that the pulse diagnosis dexterous hand 4 can collect data from multiple points on the human wrist, and fit the data collected from multiple points to the pulse width curve of the subject to ensure the accuracy of the collected data. At the same time, the surface in contact with the human wrist is an arc surface 4310, which conforms to the natural state of the wrist, further ensuring the accuracy of the collected data.
[0049] A specific implementation of a pulse diagnosis system is as follows: Figure 6-8As shown, it includes a robotic arm 2 installed on a pulse diagnosis platform 1, a four-degree-of-freedom mobile device 3 installed on the robotic arm 2, a dexterous hand 4 installed at the output end of the four-degree-of-freedom mobile device 3, a main control module 8 and a host computer 9; the dexterous hand 4 is the pulse diagnosis dexterous hand 4 mentioned above; the end of the robotic arm 2 is equipped with a lighting module 5, an infrared image acquisition module 6 for collecting photos of the human wrist and a laser ranging module 7 for measuring the distance between the dexterous hand 4 and the human wrist; the lighting module 5, the infrared image acquisition module 6 and the laser ranging module 7 are arranged corresponding to the dexterous hand 4; the main control module 8 is communicatively connected with the four-degree-of-freedom mobile device 3, the dexterous hand 4, the lighting module 5 and the laser ranging module 7; the main control module 8 and the infrared image acquisition module 6 are both communicatively connected with the host computer 9. The four-degree-of-freedom mobile device 3 in this embodiment includes four degrees of freedom, namely, X-axis movement, Y-axis movement, Z-axis movement, and rotation along the X-axis axis. The degree of freedom of rotation along the X-axis axis can rotate the dexterous hand 4 by 180° according to the pulse wave collection of the left or right hand of the subject, so as to collect the pulse waves of the left and right hands. The collection system of the present invention can collect the pulse waves of the left and right hands of the subject, and has certain versatility. It should be noted that the four degrees of freedom of the four-degree-of-freedom mobile device 3 can be realized by any movement method that can be easily thought of by those skilled in the art, and the specific structure of the four-degree-of-freedom mobile device 3 will not be described in detail here.
[0050] Before the pulse diagnosis instrument starts working, the subject needs to sit down and place his wrist in the collection area of the pulse diagnosis system to ensure that the Cun, Guan and Chi parts are located between the light spots of the two sets of laser rangefinders. Then the pulse diagnosis system is started, and the pulse diagnosis system will obtain the position of the patient's Guan part through image recognition and three-dimensional reconstruction; then the main control module 8 drives and controls the movement of the three degrees of freedom of the X, Y and Z axes of the four-degree-of-freedom mobile device 3, and transports the dynamic pressure sensing module 438 to about 2-5mm above the subject's wrist, and then according to the inclination of the subject's Guan skin through the four-degree-of-freedom mobile device 3, the dynamic pressure sensing module 438 is moved to the position of the patient's Guan part by the four-degree-of-freedom mobile device 3. The rotational freedom of the degree of movement device 3 is used to adjust the fingertip inclination of the pulse diagnosis dexterous hand 4, so that the dynamic pressure sensing module 438 can be in vertical contact with the joints of the subject; then the finger motor 431 is controlled to retract the output shaft of the finger motor 431 to push out the static pressure sensing module 437 at the end of the three fingers to contact the skin of the subject. The static pressure sensing module 437 collects real-time pressing force for closed-loop force control, while the dynamic pressure sensing module 438 collects pulse waves and transmits the data back to the host computer 9. The X, Y, and Z directions in this embodiment are Figure 7 Bit reference.
[0051] In this embodiment, the lighting module 5 is a combination of multiple groups of LED light groups, each group of LED light groups is controlled separately, and the multiple groups of LED light groups are composed of 40 near-infrared lamp beads, which are used to obtain the wrist blood vessel morphology under the irradiation of near-infrared light source; the main control module 8 is customized and developed by the project team, which integrates the data transmission control of the lighting module 5, the laser ranging module 7, the static and dynamic pressure sensor module 438, and the four-degree-of-freedom mobile device 34 parts, and is used to interact with the host computer 9 system; the near-infrared image acquisition module 6 is a near-infrared camera of Hikvision, which takes wrist pictures for locating the pulse position; it should be noted that the pulse diagnosis system also includes a network interface 10 to connect the near-infrared image acquisition module 6 and the host computer 9 system, and the interface method is a network port, which can quickly transmit large-capacity image data.
[0052] The composite pulse wave acquisition method using the above-mentioned pulse diagnosis system comprises the following steps:
[0053] S1, using the laser ranging module 7 to measure the vertical distance from the end of the robot arm 2 to the wrist of the person being measured;
[0054] S2, obtaining multiple images of the subject's wrist under different lighting angles;
[0055] S3, the deep learning model calculates the wrist joint position and the three-dimensional structure of the wrist of the subject according to the image information obtained in S1;
[0056] S4, taking the position of the joint calculated in S3 as the target point, controlling the four-degree-of-freedom mobile device 3 of the pulse diagnosis system to move, and using the output end of the four-degree-of-freedom mobile device 3 to press the multiple dynamic pressure sensing modules 438 on the dexterous hand 4 of the subject's wrist perpendicularly to the target point position;
[0057] S5, pre-collecting multi-dimensional pulse wave data from the wrist of the subject;
[0058] S6, the data signal pre-collected in step S5 is transmitted to the upper computer 9, and the upper computer 9 compares the pre-collected pulse wave data with the typical feature library of the pulse wave to determine the accuracy of the position of the key part; in said step S6, if the collected data does not contain the typical features of the pulse wave, it is determined that the positioning of the key part is inaccurate, and the operations of said steps S1 to S4 are repeated to reposition the position of the key part or perform manual auxiliary positioning; if the collected data contains the typical features of the pulse wave, it is determined that the village entrance is accurately positioned, and the operation of said step S6 is performed.
[0059] S7, if the pulse wave data pre-collected in step S6 contains typical features of the pulse wave, the center position and the position of the joint of the dynamic pressure sensor device 438 corresponding to the joint are accurately located; the main control module 8 makes micro-adjustments to the positions of the multiple dynamic pressure sensor devices 438 on the finger assembly 43 corresponding to the joint position according to the feedback signal output by the host computer 9;
[0060] S8, after determining the positions of the multiple dynamic pressure sensing devices 438 on the finger assembly 43 corresponding to the joint position and accurately positioning the joint position, perform a final multi-dimensional acquisition of the pulse wave of the subject's wrist, and transmit the acquired data to the host computer 9.
[0061] Specific reference Fig. 9 In step S7, the host computer 9 uses a time series deep learning model to analyze the pre-collected multi-channel pulse wave data in real time, and fits the peak value of the real-time pulse wave data into the pulse width curve of the subject. If the maximum peak value data of the pulse width curve is not at the center position, the host computer 9 transmits the analyzed signal feedback to the main control module 8, and the main control module 8 outputs a control signal to control the four-degree-of-freedom mobile device 3 to fine-tune the center position of the dynamic pressure sensor module 438 to the key position. It should be noted that in actual applications, it is difficult to achieve no deviation or error, so here it is moved to the key position as much as possible so that the maximum amplitude of the fitted pulse width curve is near the center position of the curve.
[0062] In step S8, the host computer 9 analyzes the collected data, extracts parameters and imports them into the prediction model in combination with pulse width, pulse position information and the physiological parameters of the subject to predict the results. The prediction model in the present embodiment is obtained by combining expert data with neural network training, which is not specifically described in the present invention, and is intended to pave the way for pulse diagnosis quantification and standardization of accumulated data. The pulse diagnosis system not only provides an artificial intelligence algorithm to extract pulse characteristics, analyze pulse elements and predict the pulse of the subject, but also provides a telemedicine interface, which enables remote experts to see the subject's inquiry data, pulse wave data, pulse analysis prediction results, etc. in real time when necessary, and gives more targeted expert advice, thereby realizing a full closed loop of pulse wave positioning, collection and diagnosis.
[0063] In step S5, the main control module 8 sets the collection pre-pressure value of each finger assembly 43 in the dexterous hand 4 and calculates the optimal pressure; the collection pre-pressure value is a plurality of gradient pressure values, and the optimal pressure collection time is greater than the collection time of other pressure gradient values; the collection pre-pressure values are 50 / 70 / 90 / 110 / 130 / 150 / 170 / 190 / 210 / 230gf ten pressure gradient values. The acquisition method of the composite pulse wave of the pulse diagnosis system of the present invention obtains multiple picture information of the subject's wrist under different lighting angles, and the collected partial photos are as follows Fig.10 As shown, the position of the joint is determined, the wrist is pre-collected, and the pre-collected data is compared with the typical feature library of the pulse wave. The accuracy of the joint position can be determined multiple times, and the position of the dexterous hand can be fine-tuned, avoiding the inaccuracy and unreliability of the joint position determined by a single medical staff in the prior art. In this embodiment, the optimal gradient pressure collection time is 30 seconds, and the collection time of other gradient pressure values is 5 seconds. It should be further explained that the pre-pressure when the pulse wave of the subject's wrist is formally collected in step S8 is the same as or slightly different from the pre-pressure during the pre-collection in step 5.
[0064] Specifically, if Fig.11 As shown, the composite pulse wave acquisition method of the pulse diagnosis system is as follows: first, the laser distance measuring device measures the distance of the near-infrared camera relative to the wrist of the subject, 40 lamp beads are cycled, and the near-infrared camera takes 40 pictures of the wrist under different lighting angles, and the wrist joint position and three-dimensional structure are calculated by the deep learning model. The four-degree-of-freedom mobile device 3 moves to the joint position with this as the target point and ensures that the array-arranged sensor device at the end of the finger assembly 43 is perpendicular to the wrist joint position. After the positioning is completed, the pulse wave is pre-collected first. If it is found that the collected data does not contain the typical characteristics of the pulse wave, it is determined that the joint positioning is inaccurate, and manual auxiliary positioning or re-positioning of the joint position is required;
[0065] In order to accurately imitate the basic pulse collection fingering of "floating, sinking" of Chinese medicine practitioners and standardize it, the present invention detects the pressure applied to the joints based on a high-precision absolute pressure sensor, and uses this as a feedback signal to set the size of the pulse wave collection pre-pressure. At present, when collecting pulse waves in multiple gradients, 10 pressure gradients such as 50 / 70 / 90 / 110 / 130 / 150 / 170 / 190 / 210 / 230gf are set respectively, and the control accuracy of the tactile perception can reach ±5gf; when data collection starts, 9 dynamic pressure sensors cycle 9 times, and different gradients are collected for 5s. The deep learning model based on time series starts to analyze the collected 9 pulse wave data in real time, and fits the peak value of the real-time pulse wave data into the pulse width curve of the subject. If the peak data of the pulse width curve is not at the center, it may be due to the deviation of the collected Guan position. The system can accurately move the center position of multiple dynamic pressure sensing modules to the Guan position by controlling the mechanical arm 2 to make fine adjustments left and right / front and back, so that the maximum amplitude of the fitted pulse width curve is near the center of the curve. After the data collection is completed, the best pulse waves of the three parts of Cun, Guan and Chi are calculated and parameter values and information are extracted. Then, the results are displayed to the host computer 9 and the cloud server as needed, and the pulse data and other data are stored, and the whole collection process ends.
[0066] It should be understood that the specific embodiments described above are only used to explain the present invention, and are not used to limit the present invention. Obvious changes or modifications derived from the spirit of the present invention are still within the protection scope of the present invention.
Claims
1. A pulse diagnosis dexterous hand, characterized in that: The invention comprises a hand frame (41), a mounting hole in a vertical direction provided at the end of the hand frame (41), and a finger assembly (43) mounted in the mounting hole; the finger assembly (43) comprises a finger motor (431) mounted on the hand frame (41), a 7-shaped top column (432) having one end connected to an output shaft of the finger motor (431) and the other end mounted in the mounting hole, a bushing (433) sleeved in the mounting hole, and a flexible substrate (434) connected to the bushing (433); a connecting tube (4311) is arranged between the bushing (433) and the mounting hole, a pressure spring (435) is arranged between the inner bottom surface of the bushing (433) and the lower end surface of the 7-shaped top column (432), and a counter-pressure spring (436) is arranged between the outer shaft shoulder of the bushing (433) and the inner shaft shoulder of the connecting tube (4311); When the shaft of the finger motor (431) is retracted to apply downward pressure to the 7-shaped top column (432), the static pressure sensor device is pushed downward from the connecting tube (4311); the pressure spring (435) directly contacted by the 7-shaped top column (432) and the counter-pressure spring (436) arranged at the outer shoulder of the bushing (433) constitute a dynamic spring system of the finger assembly (43), and when the 7-shaped top column (432) transmits pressure, a dynamic force balance system is formed; At least one static pressure sensing module (437) is provided at the connection between the flexible substrate (434) and the bushing (433); the surface of the flexible substrate (434) in contact with the human wrist is an arc-shaped surface (4310); and a plurality of dynamic pressure sensing modules (438) are provided on the arc-shaped surface (4310).
2. The pulse diagnosis dexterous hand according to claim 1, characterized in that: There is one static pressure sensing module (437); there are nine dynamic pressure sensing modules (438), which are arranged in two columns, one column with five and the other column with four, arranged on the curved surface (4310) of the flexible substrate (434).
3. The pulse diagnosis dexterous hand according to claim 1, characterized in that: The finger components (43) are divided into three groups, corresponding to the cun, guan and chi positions of the human wrist respectively.
4. A pulse diagnosis system, characterized in that: The device comprises a mechanical arm (2) mounted on a pulse diagnosis platform (1), a four-degree-of-freedom mobile device (3) mounted on the mechanical arm (2), a dexterous hand (4) mounted on the output end of the four-degree-of-freedom mobile device (3), a main control module (8) and a host computer (9); the dexterous hand (4) is the pulse diagnosis dexterous hand according to any one of claims 1 to 3 above; The end of the mechanical arm (2) is equipped with a lighting module (5), an infrared image acquisition module (6) for acquiring a photograph of a human wrist, and a laser distance measurement module (7) for measuring the distance between the dexterous hand (4) and the human wrist; the lighting module (5), the infrared image acquisition module (6), and the laser distance measurement module (7) are arranged corresponding to the dexterous hand (4); The main control module (8) is communicatively connected to the four-degree-of-freedom mobile device (3), the dexterous hand (4), the lighting module (5) and the laser ranging module (7); the main control module (8) and the infrared image acquisition module (6) are both communicatively connected to the host computer (9).
5. The pulse diagnosis system according to claim 4, characterized in that: The lighting module (5) is a combination of multiple LED light groups, and each LED light group is controlled individually.
6. A composite pulse wave acquisition method using a pulse diagnosis system as claimed in claim 5, characterized in that: The following steps are involved: S1, using a laser distance measurement module (7) to measure the vertical distance from the end of the robotic arm (2) to the wrist of the person being measured; S2, obtaining multiple images of the subject's wrist under different lighting angles; S3, the deep learning model calculates the wrist joint position and the three-dimensional structure of the wrist of the subject according to the image information obtained in S1; S4, using the position of the joint calculated in S3 as the target point, controlling the four-degree-of-freedom moving device (3) of the pulse diagnosis system to move, and using the output end of the four-degree-of-freedom moving device (3) to press the multiple dynamic pressure sensing modules (438) on the dexterous hand (4) of the subject's wrist perpendicularly to the target point position; S5, pre-collecting multi-dimensional pulse wave data from the wrist of the subject; S6, transmitting the data signal pre-collected in step S5 to the host computer (9), and the host computer (9) compares the pre-collected pulse wave data with a typical feature library of the pulse wave to determine the accuracy of the joint position; S7, if the pulse wave data pre-collected in step S6 contains typical features of the pulse wave, the center position of the dynamic pressure sensing device (438) corresponding to the joint and the joint position are accurately located; the main control module (8) makes micro-adjustments to the positions of the multiple dynamic pressure sensing devices (438) on the finger assembly (43) corresponding to the joint position according to the feedback signal output by the host computer (9); In step S7, the host computer (9) uses a time series deep learning model to analyze the pre-collected multi-channel pulse wave data in real time, and fits the peak value of the real-time pulse wave data into the pulse width curve of the subject. If the maximum peak value data of the pulse width curve is not at the center position, the host computer (9) transmits the analyzed signal feedback to the main control module (8), and the main control module (8) outputs a control signal to control the four-degree-of-freedom mobile device (3) to fine-tune the center position of the dynamic pressure sensing module (438) to the key position; S8, after determining the positions of the multiple dynamic pressure sensing devices (438) on the finger assembly (43) corresponding to the joint position and accurately positioning the joint position, the pulse wave of the subject's wrist is finally collected in multiple dimensions, and the collected data is transmitted to the host computer (9).
7. The composite pulse wave acquisition method of the pulse diagnosis system according to claim 6, characterized in that: In step S6, if the collected data does not contain the typical characteristics of the pulse wave, it is determined that the positioning of the joint is inaccurate, and the operations of steps S1 to S4 are repeated to reposition the joint or perform manual assisted positioning; if the collected data contains the typical characteristics of the pulse wave, it is determined that the positioning of the joint is accurate, and the operation of step S7 is performed.
8. The composite pulse wave acquisition method of the pulse diagnosis system according to claim 6, characterized in that: In step S8, the host computer (9) analyzes the collected data, extracts parameters, and imports them into the prediction model in combination with the pulse width, pulse position information and the subject's physiological parameters to predict the results.
9. The composite pulse wave acquisition method of a pulse diagnosis system of claim 6, characterized in that: In step S5, the main control module (8) sets a collected pre-pressure value for each finger assembly (43) in the dexterous hand, and calculates an optimal pressure; the collected pre-pressure value is a plurality of pressure gradient values, and the optimal pressure collection time is greater than the collection time of other pressure gradient values; The collected pre-pressure values are ten pressure gradient values of 50 / 70 / 90 / 110 / 130 / 150 / 170 / 190 / 210 / 230gf respectively.
Citation Information
Patent Citations
Pulse sensor capable of calibrating static force and pulse collection device
CN106108865A
Pulse information acquisition system based on dexterous hand
CN108420412A
Portable intelligent TCM pulse diagnosis instrument with multi-probe
CN108968930A
Pulse feeling data acquisition method and device, electronic equipment and system
CN113892919A
Precise positioning method for wrist pulse mouth position based on near-infrared camera shooting and deep learning
CN114298978A
