A robotic physiotherapy and rehabilitation system

Abstract

This invention discloses a robotic physiotherapy and rehabilitation system, comprising a robotic arm module, a depth camera module, and a control system. The robotic arm module includes multiple robotic arms, each with a force sensor and a replaceable pressure device installed at its end. The depth camera module is mounted on one of the robotic arms to acquire three-dimensional depth data of the patient's back. The control system includes a three-dimensional reconstruction module; an X-ray image import module; a correction point control parameter setting module; a correction point generation module; a robotic arm motion planning module; a force control feedback and closed-loop adjustment module; and a data recording module. This invention enables visualization and precise positioning of the correction position, allows multiple robotic arms to achieve high synchronization, enables coordinated force application at different correction points, and achieves high-precision setting and closed-loop control of pressure intensity, direction, and rhythm.

Description

Technical Field

[0001] This invention relates to a massage therapy robot, and more particularly to a robotic physiotherapy and rehabilitation system. Background Technology

[0002] Currently, massage and physiotherapy robots have the following problems: 1. They cannot accurately locate and correct based on the patient's true three-dimensional morphology. In traditional spinal rehabilitation, even with X-rays or other imaging data, doctors or therapists generally only observe with their naked eyes and manually locate the patient, estimating the approximate position on the patient's back skin based on experience before pressing or correcting. This has the following shortcomings: Doctors "fill in the gaps" between two-dimensional X-ray images and the patient's three-dimensional body surface, which introduces errors; it is impossible to obtain the patient's actual posture while lying on the treatment bed and the three-dimensional morphology of the soft tissues around the spine in real time, resulting in insufficient accuracy in the positioning of correction points; some existing rehabilitation / robotic systems only use simple position sensors or surface markers, which cannot achieve comprehensive three-dimensional reconstruction of the human body.

[0003] 2. Existing rehabilitation equipment and some medical robots are mostly single-acting head or single-arm structures, capable of pressing or traction on only one area at a time. They cannot simultaneously apply coordinated and linked corrective forces to multiple key vertebrae or soft tissue areas. Even multi-point devices are mostly simple multi-motor drives, lacking a unified control and synchronous coordination strategy. They are difficult to coordinate force and posture when pressing multiple points simultaneously, and cannot well simulate the action pattern of doctors applying comprehensive force at multiple points when correcting the spine.

[0004] 3. In current rehabilitation practices, the control of parameters such as pressure intensity, direction, duration of action, and frequency during treatment is relatively crude. Traditional techniques and simple electric devices cannot achieve real-time measurement and closed-loop control of pressure. It is difficult to precisely match individual patient differences and tolerance levels. The direction of pressure usually relies on experience and is difficult to accurately correspond to the direction of scoliosis and the degree of rotation. Even if the equipment provides simple pressure levels or fixed frequencies, it lacks the ability to continuously adjust and finely adjust, which is not conducive to developing differentiated rehabilitation plans. Summary of the Invention

[0005] To address the aforementioned existing technical problems, this invention provides a robotic physiotherapy and rehabilitation system that enables visualization and precise positioning of the correction position, allows multiple robotic arms to achieve high synchronization, can apply force collaboratively at different correction points, and achieves high-precision setting and closed-loop control of pressure intensity, direction, and rhythm.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: a robotic physiotherapy and rehabilitation system, comprising a robotic arm module, a depth camera module, and a control system; the robotic arm module includes multiple robotic arms, each with a force sensor and a replaceable pressing device installed at its end; the depth camera module is mounted on one of the robotic arms and is used to acquire three-dimensional depth data of the patient's back; the control system includes a three-dimensional reconstruction module: used to acquire the three-dimensional morphology of the patient's body surface on the massage bed and generate a three-dimensional human body model consistent with the robot's coordinate system; an X-ray image import module: used to import the patient's existing spinal X-ray images as visual reference information into the control system to assist the operator in manually judging and guiding the positioning of correction points; and correction point control parameters. The setting module configures corresponding execution control parameters for multiple preset correction points; the correction point generation module integrates the correction position selected by the operator with the corresponding control parameters to generate correction point information that the robotic arm can execute; the robotic arm motion planning module converts the correction points into the target pose of the robotic arm end effector based on the correction point information output by the correction point generation module; the force feedback and closed-loop adjustment module compares the real-time collected pressing force with the target pressing force and calculates the force deviation value; the closed-loop adjustment module dynamically adjusts the displacement or force output of the robotic arm end effector based on the deviation value to stabilize the actual pressing force within the target range; and the data recording module records and manages key operational data during the robotic spinal rehabilitation process.

[0007] Furthermore, the present invention includes three robotic arms, which are uniformly scheduled by the same control system to achieve multi-point coordinated force application.

[0008] Furthermore, the 3D reconstruction module of this invention includes a depth camera data acquisition unit, a point cloud preprocessing unit, a surface reconstruction unit, and a coordinate calibration and transformation unit. After the patient completes the positioning and fixation, the control system controls a robotic arm equipped with a depth camera to move to a photographing posture to acquire depth images of the patient's back area. The depth camera data acquisition unit acquires depth data containing distance information of the patient's back. The point cloud preprocessing unit processes the acquired depth data, converting the depth images into 3D point cloud data, and performs noise reduction, outlier removal, and region cropping on the point cloud data to retain the effective point cloud information corresponding to the patient's back. The surface reconstruction unit generates a continuous 3D surface model based on the point cloud data to represent the 3D shape structure of the patient's back. The 3D surface model reflects the patient's true body surface morphology in the current position. The coordinate calibration and transformation unit associates the generated 3D human body model with the massage bed coordinate system or with the robot coordinate system, establishing a unified spatial coordinate relationship between the 3D human body model and the robotic arm's motion space, thereby providing a spatial reference for subsequent correction point generation and robotic arm motion planning.

[0009] Furthermore, the X-ray image import module of this invention includes an image file import unit, an image display unit, an image overlay and transparency adjustment unit, and an image position and scale adjustment unit. The image file import unit loads a spinal X-ray image into the control system. The image display unit displays the X-ray image on the operating interface, simultaneously displaying it along with a color image acquired by a depth camera. The image overlay and transparency adjustment unit overlays the X-ray image onto the color image in a semi-transparent manner, allowing the operator to simultaneously observe the patient's physical appearance and the outline of the spinal image in the same two-dimensional view. The image position and scale adjustment unit allows the operator to manually translate, scale, and rotate the X-ray image, manually aligning the spinal position in the image with the corresponding area of ​​the patient's back in the color image, thus creating an intuitive visual overlap effect. The X-ray image generated in the above manner serves as a reference layer, providing visual guidance for the operator to determine the direction of scoliosis, select correction areas, and set correction points.

[0010] Furthermore, the correction point control parameter setting module of the present invention includes a spinal parameter input unit, a parameter verification unit, and a parameter storage and management unit. Based on the color image acquired by the depth camera and the superimposed X-ray image as a visual reference, the parameter input unit sets the angle parameter, pressing force parameter, and pressing frequency parameter corresponding to the force application direction for each correction point. The parameter verification unit verifies the input angle, pressing force, and frequency parameters to ensure that they are within the safe and executable range allowed by the control system. The verified parameters are stored by the parameter storage and management unit and transmitted to the correction point generation module to generate robotic arm control instructions corresponding to each correction point, guiding subsequent robotic arm motion planning and force control execution.

[0011] Furthermore, the correction point generation module of the present invention includes a correction point position determination unit, a correction point parameter association unit, and a correction point output unit. Based on the color image acquired by the depth camera and the superimposed X-ray image, a preset number of correction positions are manually selected in the operation interface. The correction point position determination unit converts the selected positions into spatial coordinates consistent with the human body model generated by the three-dimensional reconstruction module. The correction point parameter association unit associates the correction point positions with the angle parameters, pressure parameters, and pressure frequency parameters output by the correction point control parameter setting module to form complete correction point description information. The integrated correction point information is output by the correction point output unit to the robotic arm motion planning module to generate the corresponding robotic arm target pose and control commands.

[0012] Furthermore, the robotic arm motion planning module of this invention includes a target pose generation unit, a path planning unit, a multi-robotic arm coordination unit, and a collision detection unit. The target pose generation unit converts the correction points into the target pose of the robotic arm end effector based on the correction point information output by the correction point generation module. The path planning unit generates a motion trajectory from the initial pose to the target pose based on the structural parameters and current state of the robotic arm. In a multi-robotic arm working scenario, the multi-robotic arm coordination unit uniformly schedules the actions of the three robotic arms to ensure the coordination and consistency of each robotic arm in time and space. The collision detection unit detects potential collision risks between robotic arms, between robotic arms and patients, and between robotic arms and the bed in real time during the motion planning process to ensure the safety of the planned path.

[0013] Furthermore, the force control feedback and closed-loop adjustment module of this invention includes a force sensor data acquisition unit, a force feedback calculation unit, a closed-loop control unit, and a safety threshold protection unit. During the robotic arm compression process, the force sensor data acquisition unit acquires the actual compression force information at the end of the robotic arm in real time. The force feedback calculation unit compares the real-time acquired compression force with the target compression force and calculates the force deviation value. The closed-loop control unit dynamically adjusts the displacement or force output of the end of the robotic arm according to the deviation value to stabilize the actual compression force within the target range. When the compression force is detected to exceed the preset safety threshold or an abnormal change is detected, the safety threshold protection unit immediately triggers a force reduction or a stop action to ensure patient safety.

[0014] Furthermore, the data recording module of the present invention includes a data acquisition unit, a data storage unit, and a data retrieval unit. The data acquisition unit acquires treatment-related data during the rehabilitation process, including human three-dimensional model data, correction point location information, control parameters corresponding to each correction point, the motion trajectory of the robotic arm, and data on the change of pressure over time. The data storage unit stores the acquired data uniformly and associates it with the corresponding rehabilitation process. The data retrieval unit is used to retrieve historical data in subsequent rehabilitation processes to support the reproduction of the same or similar correction schemes, parameter comparison, and treatment process retrospection.

[0015] The beneficial effects of adopting the above technical solution are: 1. Based on the existing imaging diagnosis, the introduction of depth camera and three-dimensional reconstruction technology can obtain the patient's real body position and three-dimensional back shape on the massage bed, associate the scoliosis information in the X-ray image with the reconstructed human body model, establish correction points in three-dimensional space, and realize the visualization and precise positioning of the correction position.

[0016] 2. Enables multiple robotic arms with force sensors to achieve high synchronization under the same controller, allowing them to apply force collaboratively at different correction points, thereby more closely approaching or even surpassing the effect of manual multi-point comprehensive correction.

[0017] 3. By integrating a force sensor at the end of the robotic arm, the operator can directly input the target pressing force, offset angle, and pressing frequency through the control system, thereby achieving high-precision setting and closed-loop control of pressing force, direction, and rhythm, and thus meeting the personalized rehabilitation needs of different patients at different stages. Attached Figure Description

[0018] Figure 1 This is a block diagram of the system composition of the present invention.

[0019] Figure 2 This is a schematic diagram of an embodiment of the robotic arm of the present invention. Detailed Implementation

[0020] The present invention will now be described in further detail with reference to the accompanying drawings.

[0021] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0022] One embodiment of the present invention: as follows Figure 1 and Figure 2As shown, a robotic physiotherapy and rehabilitation system includes a robotic arm module, a depth camera module, and a control system. The robotic arm module includes multiple robotic arms, each with a force sensor and a replaceable pressure device installed at its end. The depth camera module is mounted on one of the robotic arms and is used to acquire three-dimensional depth data of the patient's back. The control system includes a three-dimensional reconstruction module for acquiring the three-dimensional morphology of the patient's body surface on the massage bed and generating a three-dimensional human body model consistent with the robot's coordinate system; an X-ray image import module for importing the patient's existing spinal X-ray images as visual reference information into the control system to assist the operator in manually judging and guiding the positioning of correction points; a correction point control parameter setting module for configuring corresponding execution control parameters for multiple preset correction points; and a correction point generation module for integrating the correction position selected by the operator with the corresponding control parameters to generate correction point information that can be executed by the robotic arm. The robotic arm motion planning module converts the correction points into the target pose of the robotic arm's end effector based on the correction point generation module's output information. The force feedback and closed-loop adjustment module compares the real-time collected pressing force with the target pressing force and calculates the force deviation. The closed-loop adjustment module dynamically adjusts the displacement or force output of the robotic arm's end effector based on the deviation value, ensuring the actual pressing force remains stable within the target range. The data recording module records and manages key operational data during the robotic spinal rehabilitation process.

[0023] Based on the above embodiments, as follows Figure 2 As shown, there are three robotic arms, which are uniformly scheduled by the same control system to achieve multi-point coordinated force application.

[0024] Based on the above embodiments, as follows Figure 1 As shown, the 3D reconstruction module includes a depth camera data acquisition unit, a point cloud preprocessing unit, a surface reconstruction unit, and a coordinate calibration and transformation unit. After the patient completes the positioning and fixation, the control system controls a robotic arm equipped with a depth camera to move to a photographing posture to acquire depth images of the patient's back area. The depth camera data acquisition unit obtains depth data containing distance information of the patient's back. The point cloud preprocessing unit processes the acquired depth data, converting the depth images into 3D point cloud data, and performs noise reduction, outlier removal, and region cropping on the point cloud data to retain the effective point cloud information corresponding to the patient's back. The surface reconstruction unit generates a continuous 3D surface model based on the point cloud data to represent the 3D shape structure of the patient's back. The 3D surface model reflects the patient's true body surface morphology in the current position. The coordinate calibration and transformation unit associates the generated 3D human body model with the massage bed coordinate system or with the robot coordinate system, establishing a unified spatial coordinate relationship between the 3D human body model and the robotic arm's motion space, thereby providing spatial reference for subsequent correction point generation and robotic arm motion planning.

[0025] Based on the above embodiments, as follows Figure 1 As shown, the X-ray image import module includes an image file import unit, an image display unit, an image overlay and transparency adjustment unit, and an image position and scale adjustment unit. The image file import unit loads a spinal X-ray image into the control system. The image display unit displays the X-ray image on the operating interface, simultaneously displaying it along with a color image acquired by a depth camera. The image overlay and transparency adjustment unit overlays the X-ray image onto the color image in a semi-transparent manner, allowing the operator to simultaneously observe the patient's physical appearance and the outline of the spinal image in the same two-dimensional view. The image position and scale adjustment unit allows the operator to manually translate, scale, and rotate the X-ray image, aligning the spinal position in the image with the corresponding area of ​​the patient's back in the color image, thus creating a visually intuitive overlap effect. The X-ray image generated in this way serves as a reference layer, providing visual guidance for the operator to determine the direction of scoliosis, select correction areas, and set correction points.

[0026] Based on the above embodiments, as follows Figure 1 As shown, the correction point control parameter setting module includes a spinal parameter input unit, a parameter verification unit, and a parameter storage and management unit. Based on color images acquired by a depth camera and superimposed X-ray images as visual references, the parameter input unit sets the angle parameter, pressure parameter, and pressure frequency parameter corresponding to the force application direction for each correction point. The parameter verification unit verifies the input angle, pressure, and frequency parameters to ensure they are within the safe and executable range allowed by the control system. The verified parameters are stored by the parameter storage and management unit and transmitted to the correction point generation module to generate robotic arm control instructions corresponding to each correction point, guiding subsequent robotic arm motion planning and force control execution.

[0027] Based on the above embodiments, as follows Figure 1 As shown, the correction point generation module includes a correction point location determination unit, a correction point parameter association unit, and a correction point output unit. Based on the color images acquired by the depth camera and the superimposed X-ray images, a preset number of correction positions are manually selected in the operation interface. The correction point location determination unit converts the selected positions into spatial coordinates consistent with the human body model generated by the 3D reconstruction module. The correction point parameter association unit associates the correction point positions with the angle parameters, pressure parameters, and pressure frequency parameters output by the correction point control parameter setting module to form complete correction point description information. The integrated correction point information is output by the correction point output unit to the robotic arm motion planning module to generate the corresponding robotic arm target pose and control commands.

[0028] Based on the above embodiments, as follows Figure 1 As shown, the robotic arm motion planning module includes a target pose generation unit, a path planning unit, a multi-robotic arm coordination unit, and a collision detection unit. The target pose generation unit converts the correction points into the target pose of the robotic arm end effector based on the correction point information output by the correction point generation module. The path planning unit generates the motion trajectory from the initial pose to the target pose based on the structural parameters and current state of the robotic arm. In a multi-robotic arm working scenario, the multi-robotic arm coordination unit uniformly schedules the actions of the three robotic arms to ensure the coordination and consistency of each robotic arm in time and space. The collision detection unit detects potential collision risks between robotic arms, between robotic arms and patients, and between robotic arms and the bed in real time during the motion planning process to ensure the safety of the planned path.

[0029] Based on the above embodiments, as follows Figure 1 As shown, the force control feedback and closed-loop adjustment module includes a force sensor data acquisition unit, a force feedback calculation unit, a closed-loop control unit, and a safety threshold protection unit. During the robotic arm compression process, the force sensor data acquisition unit acquires the actual compression force information at the end of the robotic arm in real time. The force feedback calculation unit compares the real-time collected compression force with the target compression force and calculates the force deviation value. The closed-loop control unit dynamically adjusts the displacement or force output of the end of the robotic arm according to the deviation value to stabilize the actual compression force within the target range. When the compression force is detected to exceed the preset safety threshold or an abnormal change is detected, the safety threshold protection unit immediately triggers force reduction or stops the action to ensure patient safety.

[0030] Based on the above embodiments, as follows Figure 1 As shown, the data recording module includes a data acquisition unit, a data storage unit, and a data retrieval unit. The data acquisition unit acquires treatment-related data during the rehabilitation process, including human 3D model data, correction point location information, control parameters corresponding to each correction point, the motion trajectory of the robotic arm, and data on the change of pressure over time. The data storage unit stores the acquired data uniformly and saves it in association with the corresponding rehabilitation process. The data retrieval unit is used to retrieve historical data in subsequent rehabilitation processes to support the reproduction of the same or similar correction schemes, parameter comparison, and treatment process retrospection.

[0031] Based on the above embodiments, the working steps of the present invention are as follows:

[0032] Step 1: Position and fix the patient in a fixed position.

[0033] (1) The patient lies prone on the massage bed.

[0034] (2) Use multiple fixed straps on the shoulders, waist and legs to restrain the patient's position on the bed in a set posture so that the patient will not shift position during treatment.

[0035] (3) The control system records the bed coordinates and the patient's initial posture as a spatial reference for subsequent 3D modeling and robot force application.

[0036] Step 2: Depth vision imaging and 3D human body reconstruction.

[0037] (1) The control system moves a robotic arm to the shooting position and controls the depth camera to obtain a depth image of the patient's back.

[0038] (2) Based on depth data, three-dimensional point cloud acquisition of the patient's back is performed, and a three-dimensional human body model of the patient's back is generated by filtering, registration and surface reconstruction algorithms.

[0039] (3) The three-dimensional model is bound to the patient coordinate system and the bed coordinate system for subsequent correction point positioning and motion planning.

[0040] Step 3: Medical image analysis and correction point calibration.

[0041] (1) Medical staff import the patient’s spinal X-ray (or other imaging data) into the control interface.

[0042] (2) Set the doctor's correction parameters according to the treatment plan after the doctor's diagnosis.

[0043] (3) Medical staff determine the location of the vertebra or soft tissue to be corrected based on the image information, and the system generates the corresponding correction point on the three-dimensional human body model.

[0044] (4) Each correction point includes: 1. Three-dimensional spatial coordinates; 2. Suggested pressing force; 3. Suggested pressing direction vector; 4. Suggested pressing frequency; 5. Whether multi-arm coordinated force application is required.

[0045] Step 4: Mapping the correction point to the robotic arm's motion.

[0046] The system uses spatial coordinate transformation and motion planning to map correction point information into the operating position and attitude control parameters of the robotic arm's end effector, specifically including:

[0047] 1. Generate the target pose of the robotic arm based on the correction points:

[0048] (1) Pressing position → End position of robotic arm;

[0049] (2) Pressing direction → End-effector posture (offset angle);

[0050] (3) According to the pressure target → force control system closed-loop parameters.

[0051] 2. Path planning:

[0052] To avoid collisions between the robotic arm and the patient's body or other robotic arms, the system performs collision-free planning for the movement of any robotic arm.

[0053] 3. Multi-arm coordinated scheduling:

[0054] When multiple robotic arms need to apply force simultaneously at the correction point, the controller synchronously schedules three robotic arms according to the mechanical model and treatment plan.

[0055] Step 5: Force control pressing and closed-loop adjustment.

[0056] (1) When the robotic arm applies force, the force sensor detects the magnitude of the pressing force in real time.

[0057] (2) The control system performs closed-loop adjustment based on the deviation between the input target pressure and the real-time measured pressure to ensure accurate and stable force application.

[0058] (3) By adjusting the angle of the pressing direction, the corrective effect can be improved by precisely applying pressure to the corresponding direction of the scoliosis.

[0059] (4) The pressing frequency, pulse period and number of repetitions can be set to realize dynamic correction or continuous correction mode.

[0060] Step 6: Multi-robotic arm collaborative correction.

[0061] Depending on the treatment plan, two or three robotic arms can be used to apply force simultaneously at different correction points, specifically including:

[0062] 1. Synchronous pressure mode: Multiple pressure points apply force simultaneously to achieve a balanced distribution of corrective force.

[0063] 2. Differential pressure mode: Apply different pressure and direction to different areas according to the degree of scoliosis.

[0064] 3. Sequential pressure mode; multi-point pressure is triggered in sequence to simulate the comprehensive techniques of a doctor.

[0065] In the above mode, the control system ensures the synchronization, safety, and non-interference of paths among the multiple arms.

[0066] Step 7: Data recording and treatment plans can be reused.

[0067] The system automatically records key data throughout the treatment process, including: a 3D human body model; the correlation between X-ray images and the 3D model; parameters for each correction point; the motion trajectory of the robotic arm; the actual pressure curve; and the pressure frequency.

[0068] The data can be used to replicate the same treatment plan or improve treatment parameters in the next treatment, thus realizing personalized and standardized robotic spinal rehabilitation programs.

[0069] In summary, compared with existing technologies, this invention has the following advantages: 1. It breaks through the limitations of traditional reliance on single images, fusing deep spinal structure data from CT scans with real-time 3D visual acquisition of body surface posture data to construct a full-dimensional digital human body model. Equipped with intelligent algorithms, it automatically identifies the type, degree, and distribution of surrounding tissues of spinal deformities, generating personalized rehabilitation pathways without manual intervention, thus solving the problems of traditional methods relying on doctors' experience and being highly subjective.

[0070] 2. A multi-arm linkage control module is designed. The bed automatically adapts to the patient's body shape through adjustable constraint components to achieve stable fixation, and simultaneously triggers multiple arms to work collaboratively according to preset division of labor. This avoids the force limitations of single-arm operation, achieves precise matching of the "fixation-positioning-pressing" action sequence, improves the stability and operational efficiency of the correction process, and solves the problem of disconnect between fixation and correction actions in existing technologies.

[0071] 3. The therapy head integrates a high-precision pressure sensor and an adjustable actuator, which can automatically switch the pressing mode according to the rehabilitation plan and provide real-time feedback of contact pressure data. Combined with a closed-loop control algorithm, it dynamically adjusts the pressing force, frequency, and depth to ensure that the pressure error is ≤±3N, avoiding the risk of secondary injury caused by the uncontrolled pressing force of traditional therapy heads.

[0072] 4. During rehabilitation therapy, continuously collect dynamic change data of CT images, 3D visual posture data, and force control feedback data to construct a multi-dimensional monitoring network and realize adaptive adjustment of the rehabilitation path with dynamic real-time feedback.

[0073] 5. It can pre-define the multi-arm working area based on the human digital model, and is equipped with a dynamic obstacle avoidance algorithm to calculate the motion trajectory of each robotic arm in real time, avoiding multi-arm interference and unintended contact with the patient's body. It automatically allocates the functions of each arm according to the complexity of the rehabilitation task, improving the efficiency of multi-arm collaboration and solving the problems of fixed task allocation and insufficient flexibility in existing multi-arm robots.

[0074] Note that the above description is merely a preferred embodiment of the present invention. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of the present invention, the scope of which is determined by the scope of the appended claims.

Claims

1. A robotic physiotherapy and rehabilitation system, characterized in that, Includes a robotic arm module, a depth camera module, and a control system; The robotic arm module includes multiple robotic arms, each with a force sensor and a replaceable pressing fixture installed at its end. The depth camera module is mounted on one of the robotic arms and is used to acquire three-dimensional depth data of the patient's back. The control system includes a three-dimensional reconstruction module: used to acquire the three-dimensional shape of the patient's body surface on the massage bed and generate a three-dimensional human body model that is consistent with the robot coordinate system. X-ray image import module: used to import the patient's existing spinal X-ray images as visual reference information into the control system to assist the operator in manually judging and locating the correction points. Correction point control parameter setting module: used to configure corresponding execution control parameters for multiple preset correction points; Correction point generation module: This module integrates the correction position selected by the operator with the corresponding control parameters to generate correction point information that can be executed by the robotic arm. Robotic arm motion planning module: Based on the correction point information output by the correction point generation module, convert the correction points into the target pose of the robotic arm end effector; Force feedback and closed-loop adjustment module: The force feedback module compares the real-time pressing force with the target pressing force and calculates the force deviation value; the closed-loop adjustment module dynamically adjusts the displacement or force output of the robotic arm end based on the deviation value to keep the actual pressing force stable within the target range; Data recording module: used to record and manage key operational data during the robotic spinal rehabilitation process; The X-ray image import module includes an image file import unit, an image display unit, an image overlay and transparency adjustment unit, and an image position and scale adjustment unit. An X-ray image of the spine is loaded into the control system via the image file import unit. The image display unit displays the X-ray image on the user interface, simultaneously displaying it along with a color image captured by a depth camera. The image overlay and transparency adjustment unit overlays the X-ray image onto the color image in a semi-transparent manner, allowing the operator to simultaneously observe the patient's appearance and the outline of the spine in the same two-dimensional view. The image position and scale adjustment unit allows the operator to manually translate, scale, and rotate the X-ray image, aligning the spine position in the image with the corresponding area of ​​the patient's back in the color image, thus creating a visually intuitive overlap effect. The X-ray image generated in the above manner serves as a reference layer, providing visual guidance for the operator to determine the direction of scoliosis, select the correction area, and set correction points. The correction point generation module includes a correction point location determination unit, a correction point parameter association unit, and a correction point output unit. Based on the color image acquired by the depth camera and the superimposed X-ray image, a preset number of correction points are manually selected in the operation interface. The correction point location determination unit converts the selected positions into spatial coordinates consistent with the human body model generated by the three-dimensional reconstruction module. The correction point parameter association unit associates the correction point position with the angle parameters, pressing force parameters and pressing frequency parameters output by the correction point control parameter setting module to form a complete correction point description information. The integrated correction point information is output by the correction point output unit to the robotic arm motion planning module to generate the corresponding robotic arm target pose and control commands.

2. The robotic physiotherapy and rehabilitation system according to claim 1, characterized in that, There are three robotic arms, which are uniformly scheduled by the same control system to achieve multi-point coordinated force application.

3. A robotic physiotherapy and rehabilitation system according to claim 1 or 2, characterized in that, The 3D reconstruction module includes a depth camera data acquisition unit, a point cloud preprocessing unit, a surface reconstruction unit, and a coordinate calibration and transformation unit. After the patient is positioned and fixed, the control system controls a robotic arm equipped with a depth camera to move to a photographing posture to acquire depth images of the patient's back area. The depth camera data acquisition unit acquires depth data containing the distance information of the patient's back. The point cloud preprocessing unit processes the acquired depth data, converts the depth image into three-dimensional point cloud data, and performs noise reduction, outlier removal, and region cropping on the point cloud data to preserve the effective point cloud information corresponding to the patient's back. The surface reconstruction unit generates a continuous three-dimensional surface model based on the point cloud data to represent the three-dimensional shape structure of the patient's back. The three-dimensional surface model is used to reflect the patient's true body surface morphology in the current position. The coordinate calibration and transformation unit associates the generated 3D human body model with the massage bed coordinate system or with the robot coordinate system, so as to establish a unified spatial coordinate relationship between the 3D human body model and the robotic arm motion space, thereby providing a spatial reference for subsequent correction point generation and robotic arm motion planning.

4. A robotic physiotherapy and rehabilitation system according to claim 1 or 2, characterized in that, The correction point control parameter setting module includes a spinal parameter input unit, a parameter verification unit, and a parameter storage and management unit; Using color images captured by a depth camera and superimposed X-ray images as visual references, the parameter input unit sets the angle, pressure, and frequency parameters corresponding to the force application direction for each correction point; the parameter verification unit verifies the input angle, pressure, and frequency parameters to ensure that they are within the safe and executable range allowed by the control system. The verified parameters are stored by the parameter storage and management unit and transmitted to the correction point generation module to generate robotic arm control instructions corresponding to each correction point, guiding subsequent robotic arm motion planning and force control execution.

5. A robotic physiotherapy and rehabilitation system according to claim 1 or 2, characterized in that, The robotic arm motion planning module includes a target pose generation unit, a path planning unit, a multi-robotic arm coordination unit, and a collision detection unit. The target pose generation unit converts the correction points into the target pose of the robotic arm end based on the correction point information output by the correction point generation module; the path planning unit generates the motion trajectory from the initial pose to the target pose based on the structural parameters and current state of the robotic arm; in the multi-robotic arm working scenario, the multi-robotic arm coordination unit performs unified scheduling of the actions of multiple robotic arms to ensure the coordination and consistency of each robotic arm in time and space. The collision detection unit detects potential collision risks between robotic arms, between robotic arms and patients, and between robotic arms and the bed in real time during the motion planning process, ensuring the safety of the planned path.

6. A robotic physiotherapy and rehabilitation system according to claim 1 or 2, characterized in that, The force control feedback and closed-loop adjustment module includes a force sensor data acquisition unit, a force feedback calculation unit, a closed-loop control unit, and a safety threshold protection unit. During the pressing process of the robotic arm, the force sensor data acquisition unit acquires the actual pressing force information at the end of the robotic arm in real time. The force feedback calculation unit compares the pressing force collected in real time with the target pressing force and calculates the force deviation value; the closed-loop control unit dynamically adjusts the displacement or force output of the robotic arm end based on the deviation value to keep the actual pressing force stable within the target range; When the pressure applied exceeds the preset safety threshold or changes abnormally, the safety threshold protection unit immediately triggers a reduction in pressure or stops the operation to ensure patient safety.

7. A robotic physiotherapy and rehabilitation system according to claim 1 or 2, characterized in that, The data recording module includes a data acquisition unit, a data storage unit, and a data retrieval unit; The data acquisition unit acquires treatment-related data during the rehabilitation process, including human 3D model data, correction point location information, control parameters corresponding to each correction point, movement trajectory of the robotic arm, and data on pressure changes over time. The data storage unit stores the acquired data uniformly and saves it in association with the corresponding rehabilitation process. The data retrieval unit is used to retrieve historical data in subsequent rehabilitation processes to support the reproduction of the same or similar correction schemes, parameter comparison, and treatment process retrospection.

Images