WO2021128755A1 - 骨注册方法、手术机器人及可读存储介质 - Google Patents

骨注册方法、手术机器人及可读存储介质 Download PDF

Info

Publication number
WO2021128755A1
WO2021128755A1 PCT/CN2020/095666 CN2020095666W WO2021128755A1 WO 2021128755 A1 WO2021128755 A1 WO 2021128755A1 CN 2020095666 W CN2020095666 W CN 2020095666W WO 2021128755 A1 WO2021128755 A1 WO 2021128755A1
Authority
WO
WIPO (PCT)
Prior art keywords
robotic arm
force
tracking device
interest
arm
Prior art date
Application number
PCT/CN2020/095666
Other languages
English (en)
French (fr)
Inventor
彭维礼
江洲
孙峰
何超
李涛
Original Assignee
苏州微创畅行机器人有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 苏州微创畅行机器人有限公司 filed Critical 苏州微创畅行机器人有限公司
Priority to EP20905068.1A priority Critical patent/EP4082470A4/en
Priority to JP2022539307A priority patent/JP2023508452A/ja
Priority to AU2020414777A priority patent/AU2020414777B2/en
Publication of WO2021128755A1 publication Critical patent/WO2021128755A1/zh

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
    • A61B34/30—Surgical robots
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
    • A61B34/20—Surgical navigation systems; Devices for tracking or guiding surgical instruments, e.g. for frameless stereotaxis
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
    • A61B34/10—Computer-aided planning, simulation or modelling of surgical operations
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
    • A61B34/25—User interfaces for surgical systems
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
    • A61B34/30—Surgical robots
    • A61B34/37—Master-slave robots
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
    • A61B34/70—Manipulators specially adapted for use in surgery
    • A61B34/76—Manipulators having means for providing feel, e.g. force or tactile feedback
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
    • G06T7/33—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods
    • G06T7/344—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods involving models
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B17/00—Surgical instruments, devices or methods, e.g. tourniquets
    • A61B2017/00681—Aspects not otherwise provided for
    • A61B2017/00725—Calibration or performance testing
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
    • A61B34/10—Computer-aided planning, simulation or modelling of surgical operations
    • A61B2034/101—Computer-aided simulation of surgical operations
    • A61B2034/105—Modelling of the patient, e.g. for ligaments or bones
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
    • A61B34/10—Computer-aided planning, simulation or modelling of surgical operations
    • A61B2034/107—Visualisation of planned trajectories or target regions
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
    • A61B34/20—Surgical navigation systems; Devices for tracking or guiding surgical instruments, e.g. for frameless stereotaxis
    • A61B2034/2046—Tracking techniques
    • A61B2034/2055—Optical tracking systems
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
    • A61B34/20—Surgical navigation systems; Devices for tracking or guiding surgical instruments, e.g. for frameless stereotaxis
    • A61B2034/2046—Tracking techniques
    • A61B2034/2059—Mechanical position encoders
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
    • A61B34/20—Surgical navigation systems; Devices for tracking or guiding surgical instruments, e.g. for frameless stereotaxis
    • A61B2034/2046—Tracking techniques
    • A61B2034/2061—Tracking techniques using shape-sensors, e.g. fiber shape sensors with Bragg gratings
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
    • A61B34/20—Surgical navigation systems; Devices for tracking or guiding surgical instruments, e.g. for frameless stereotaxis
    • A61B2034/2068—Surgical navigation systems; Devices for tracking or guiding surgical instruments, e.g. for frameless stereotaxis using pointers, e.g. pointers having reference marks for determining coordinates of body points
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
    • A61B34/20—Surgical navigation systems; Devices for tracking or guiding surgical instruments, e.g. for frameless stereotaxis
    • A61B2034/2068—Surgical navigation systems; Devices for tracking or guiding surgical instruments, e.g. for frameless stereotaxis using pointers, e.g. pointers having reference marks for determining coordinates of body points
    • A61B2034/207—Divots for calibration
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B90/00—Instruments, implements or accessories specially adapted for surgery or diagnosis and not covered by any of the groups A61B1/00 - A61B50/00, e.g. for luxation treatment or for protecting wound edges
    • A61B90/06—Measuring instruments not otherwise provided for
    • A61B2090/064—Measuring instruments not otherwise provided for for measuring force, pressure or mechanical tension
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B90/00—Instruments, implements or accessories specially adapted for surgery or diagnosis and not covered by any of the groups A61B1/00 - A61B50/00, e.g. for luxation treatment or for protecting wound edges
    • A61B90/39—Markers, e.g. radio-opaque or breast lesions markers
    • A61B2090/3983—Reference marker arrangements for use with image guided surgery
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20084—Artificial neural networks [ANN]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G06T2207/30008—Bone

Definitions

  • the present invention relates to the technical field of medical devices, in particular to a bone registration method, a surgical robot and a readable storage medium.
  • Computer-aided navigation system has become an important part of modern surgical technology, helping operators to perform a variety of complex operations more accurately and safely. It has many irreplaceable advantages in all aspects of orthopedics, because of its accuracy, safety, and low cost. Radiation and other characteristics make it gradually widely used in clinical practice.
  • the computer-assisted navigation system can be used to guide the operator to perform surgical training, make surgical plans, navigate surgical instruments in real time, and reduce patient trauma.
  • computer-assisted osteotomy is more accurate, prosthesis installation is more accurate, and the postoperative force line recovery is closer to the physiological state.
  • surgical robots need to register bones in advance to establish the correspondence between solid bones and virtual images, and provide three-dimensional information for the operator's operations such as osteotomy, bone surface treatment, and prosthesis placement.
  • the traditional bone registration method is that the operator holds a target pen and selects the registration point on the patient's bone to complete the bone registration.
  • the present invention provides a method for performing bone registration by a surgical robot, a surgical robot, and a readable storage medium, which are used to automatically complete or assist the operator in completing the bone registration process, improve the efficiency of bone registration, and help the operator to reduce the workload of the operator.
  • it can also identify the force status of the end of the robotic arm, so as to ensure that the registration points on the bones are accurately selected, which improves the accuracy and safety of bone registration.
  • a bone registration method for surgical robots which includes the following steps:
  • the robotic arm system receives the information of the interested registration point on the predetermined object selected by the navigation system;
  • the tracking device According to the force received by the end of the robotic arm, it is determined whether the tracking device has moved to the registered point of interest on the predetermined object.
  • the method further includes the following steps:
  • the tracking device According to the current location information of the tracking device, it is determined whether the tracking device has moved to a registered point of interest on a predetermined object.
  • the priority of judging whether the tracking device has moved to the registered point of interest on a predetermined object based on the force received by the end of the robotic arm is higher than that based on the tracking device To determine whether the tracking device has moved to the registered point of interest on the predetermined object.
  • the determination is made according to the force received by the end of the robotic arm to determine whether the tracking device has moved to a registered point of interest on a predetermined object.
  • the steps include:
  • the method further includes:
  • the method further includes the following steps:
  • control the mechanical arm to adjust the posture of the tracking device until the navigation system can normally collect the position information of the tracking device.
  • the robotic arm system plans a motion trajectory, and the robotic arm drives the tracking device to move to the registration point of interest along the motion trajectory.
  • the method before the step of judging whether the tracking device has moved to the registration point of interest on the predetermined object, the method further includes the following steps:
  • the force recognition system is updated.
  • the step of obtaining the force received by each mechanical arm joint by the force recognition system includes:
  • the force received by each of the robot arm joints is detected.
  • the step of obtaining the force received by the end of the robotic arm by the force recognition system includes:
  • the force received by the end of the robotic arm is detected.
  • a neural network calculation method is used to calculate the force received by the end of the robotic arm, and before outputting the force received by the end of the robotic arm , It also includes the step of performing neural network training on the force recognition system, and maintaining the network structure and weight files after neural network training.
  • a surgical robot which includes a control system, a force recognition system, a robotic arm system, and a navigation system;
  • the robotic arm system includes a robotic arm; the robotic arm end of the robotic arm is used for Connect a tracking device;
  • the control system is respectively communicatively connected to the force recognition system, the robotic arm system, and the navigation system, and the navigation system is communicatively connected to the robotic arm system;
  • the navigation system is configured to send information about the registration point of interest on a predetermined object to the robotic arm system; the robotic arm system is configured to control the robot arm system according to the information about the registration point of interest on the predetermined object.
  • the robotic arm moves to drive the tracking device to move to a registered point of interest on a predetermined object;
  • the force recognition system is provided on the mechanical arm, and the force recognition system is used to detect the force received by the end of the mechanical arm and send it to the control system; the control system is configured to be based on the mechanical arm The force received by the end of the arm determines whether the tracking device has moved to the registered point of interest on the predetermined object.
  • the navigation system is configured to collect current position information of the tracking device and send it to the control system; the control system is configured to determine according to the current position information of the tracking device Whether the tracking device has moved to a registered point of interest on a predetermined object.
  • control system is configured to determine whether the tracking device has moved to a predetermined object with a higher priority than the registration point of interest according to the force received by the end of the robotic arm. According to the current position information of the tracking device, it is a priority to determine whether the tracking device has moved to a registered point of interest on a predetermined object.
  • control system is configured to determine whether the force recognition system is working normally; if so, control the mechanical arm to drive the tracking device to move to the registration point.
  • the control system is configured to compare the force received by the end of the robotic arm with a calibration value pre-stored in the control system; if If it is equal to or greater than the calibration value, it is determined that the tracking device has moved to the registration point of interest on the predetermined object; if it is less than the calibration value, continue to control the robotic arm to drive the tracking device to the registration point Move until the force received by the end of the mechanical arm is equal to or greater than the calibration value.
  • control system is further configured to determine whether the navigation system can normally collect the position information of the tracking device If yes, control the robotic arm to drive the tracking device to move to the next registration point of interest; if not, control the robotic arm to adjust the posture of the tracking device until the navigation system can normally collect the tracking Location information of the device.
  • the mechanical arm includes a plurality of mechanical arm joints, and the force recognition system includes a first force recognition device and a second force recognition device;
  • the first force recognition device includes a plurality of first sensors, and at least one first sensor is provided at each of the robot arm joints; the first sensor is used to detect the force received by the corresponding robot arm joint;
  • the second force recognition device includes at least one second sensor arranged at the end of the mechanical arm, and the second sensor is used to detect the force received by the end of the mechanical arm;
  • the control system is configured to:
  • the force recognition system is updated.
  • the first sensor includes at least one of a dual encoder, a torque sensor, and a distributed grating sensor
  • the second sensor includes a six-axis sensor.
  • a dual encoder is provided at each joint of the robotic arm, and the dual encoder includes an absolute encoder and an incremental encoder;
  • the robotic arm also includes a A driver and a reducer for an arm joint, the output end of the driver is connected to the input end of the reducer, and the output end of the reducer is connected to the mechanical arm joint; wherein, the absolute encoder and the incremental encoder One of the reducers is arranged at the input end of the reducer, and the other is arranged at the output end of the reducer.
  • K is the stiffness model of the reducer
  • ⁇ is the deformation of the manipulator joint
  • T1 is the absolute position of the manipulator joint fed back by the absolute encoder
  • T2 is the increment The relative position of the robotic arm joint fed back by the encoder
  • c1 is the resolution of the absolute encoder
  • c2 is the resolution of the incremental encoder
  • S0 is the incremental encoder The start-up initialization position.
  • the dual encoder is communicatively connected with the control system; the control system is used to map the data deviation of the dual encoder and the force received by the mechanical arm joint, Obtain the force received by the robotic arm joint.
  • control system is configured to use a neural network calculation method to calculate the force received by the end of the robotic arm, and before outputting the force received by the end of the robotic arm, the The force recognition system conducts neural network training, and maintains the network structure and weight files after neural network training.
  • the surgical robot includes an automatic operation mode and an auxiliary operation mode
  • the robotic arm system plans a movement trajectory according to the received information of the registration point of interest, and controls the robotic arm to drive the tracking device along the movement The trajectory moves to the registered point of interest on the predetermined object;
  • the mechanical arm is driven by an external force and the tracking device is driven to move to a registered point of interest on a predetermined object.
  • a readable storage medium having a program stored thereon, and the above-mentioned bone registration method for a surgical robot is executed when the program is executed by a processor.
  • the use of surgical robots for bone registration is convenient to improve the efficiency of bone registration and help operators reduce the workload.
  • the recognition system can identify the force status of the end of the robotic arm, which is convenient for judging the cartilage penetration of the tracking device according to the force status of the robotic arm, thereby controlling the depth of cartilage insertion, ensuring that the registration point on the bone is correctly selected, and improving bone registration Accuracy and safety.
  • multiple sensors in the force recognition system are used to obtain the force information received by the end of the manipulator, so that the force information can be verified with each other to ensure the accuracy and reliability of the force information detection.
  • the bone registration process it is further judged whether the force recognition system is working normally. If it is not normal, the bone registration can be performed manually, which further ensures the accuracy and reliability of the force detection. Furthermore, the force information received by the end of the manipulator can be obtained according to the calculation method of the neural network, which further improves the accuracy of force detection.
  • Fig. 1 is an overall schematic diagram of a surgical robot in an embodiment of the present invention performing a surgical procedure
  • FIG. 2 is a schematic diagram of the structure of the robotic arm in the embodiment of the present invention.
  • Fig. 3 is a general flow chart of bone registration performed by the surgical robot in an embodiment of the present invention.
  • FIG. 5 is a flowchart of bone registration when the surgical robot in an embodiment of the present invention is in an automatic operation mode
  • Fig. 6a is a schematic diagram of the target pen outside the field of view of the navigation system in the embodiment of the present invention.
  • 6b is a schematic diagram of the target pen in the field of view of the navigation system in the embodiment of the present invention.
  • FIG. 7 is a schematic diagram of a joint module in an embodiment of the present invention.
  • Fig. 8 is a schematic diagram of measuring joint torque using dual encoders in an embodiment of the present invention.
  • Figure 9 is a flowchart of protecting the joint module in an embodiment of the present invention.
  • FIG. 11 is a flowchart of online model parameter update using a six-axis sensor in an embodiment of the present invention.
  • Figures 12 to 13 are respectively schematic diagrams of using neural networks to update parameters in an embodiment of the present invention.
  • the singular forms “a”, “an” and “the” include plural items unless the content clearly dictates otherwise.
  • the term “or” is usually used in the meaning including “and/or” unless the content clearly indicates otherwise.
  • the term “several” is usually used to include “at least one” unless the content clearly indicates otherwise.
  • the term “at least two” is generally used in the meaning including “two or more” unless the content clearly indicates otherwise.
  • the terms “first”, “second”, and “third” are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Thus, the features defined with “first”, “second”, and “third” may explicitly or implicitly include one or at least two of these features.
  • One of the main purposes of the present invention is to provide a surgical robot, which aims to complete the bone registration process through collaborative operation of the surgical robot, and with the aid of a force recognition system, can obtain the force information (including force and torque) received by the end of the robotic arm. Information) to ensure that the target pen/tracking device can accurately select the registration point on the bone, and ensure the accuracy and safety of the bone registration.
  • the sensor used in the present invention will not be damaged under the condition of excessive external force, which ensures the reliability of the sensor.
  • the sensors can also verify each other to further ensure the accuracy and reliability of bone registration.
  • FIG. 1 is an overall schematic diagram of the surgical robot 10 in this embodiment when performing knee joint surgery.
  • this embodiment provides a surgical robot 10, which includes a robotic arm system/mechanical arm subsystem 100 and a navigation system/navigation subsystem 200, and the two systems communicate with each other.
  • the robotic arm system 100 has an automatic operation mode and an auxiliary operation mode.
  • the surgical robot 10 automatically completes the bone registration process; when the robotic arm system 100 is in the auxiliary operation mode, the surgical robot 10 guides the operator to complete the bone registration process.
  • the navigation system 200 is used for preoperative planning and intraoperative navigation.
  • the navigation system 200 is the visual recognition part of the surgical robot 10.
  • FIG. 2 is a schematic diagram of the structure of the robot arm 110 in this embodiment.
  • the robotic arm system 100 includes a robotic arm 110, which includes a number of robotic arm joints connected in sequence and a robotic arm end.
  • the robotic arm joints may be revolving joints for controlling the robotic arm.
  • the mechanical arm 110 includes at least five revolute joints connected in sequence to achieve at least five degrees of freedom. Further, any two adjacent revolute joints are connected by a connecting arm.
  • the robotic arm 110 has five degrees of freedom and includes a first revolute joint 111, a second revolute joint 112, a rotation joint 113, a third revolute joint 114, and The fourth revolute joint 115.
  • the rotation axis of the first revolute joint 111, the rotation axis of the second revolute joint 112, and the rotation axis of the third revolute joint 114 are parallel to each other; the rotation axes of the third revolute joint 114 and the fourth revolute joint 115 are perpendicular to each other.
  • any two adjacent revolute joints are connected by a connecting arm 116.
  • the first revolute joint 111 is connected to a connecting arm 116, one end of the connecting arm 116 is connected to the first revolving joint 111, and the other end is fixedly connected to a base.
  • the base can be fixed on the hospital bed 20 or the operating trolley 120.
  • the fourth revolute joint 115 is connected to a surgical tool or a target pen.
  • the fourth revolute joint 115 is connected to the target pen, so that the target pen is connected through the robot arm 110, and the target pen is driven by the robot arm, and the registration point of interest is selected on the patient's bone to complete the bone registration. process.
  • the surgical robot 10 further includes a force recognition system, which is arranged on the mechanical arm 110 and is used to detect the force received by the end of the mechanical arm.
  • the force received by the end of the mechanical arm is the force received when the target pen punctures the cartilage or performs bone registration.
  • the surgical robot 10 further includes a control system, which is respectively communicatively connected with the force recognition system, the robotic arm system 100 and the navigation system 200.
  • the force recognition system feeds back information about the force received by the end of the robotic arm that it detects to the control system, and the control system determines whether the target pen has been activated based on the information about the force received by the end of the robotic arm. Pierce to the bone/hard bone surface, that is, determine whether the target pen moves to the registered point of interest on the bone.
  • a calibration value of the force received by the end of the mechanical arm is stored in the control system, and the calibration value can be calibrated through experiments.
  • the control system compares the force received by the detected end of the robotic arm with the calibration value.
  • the actual detected force matches the calibration value (refers to the actual detected force reaches the calibration value, that is, greater than or equal to the calibration value )
  • the actually detected force does not reach (that is, less than) the calibration value, continue to control the robotic arm drive station
  • the target pen moves to the registration point (such as further deepening), and repeatedly compares the force received by the end of the robotic arm with the calibration value, until the force received by the end of the robotic arm reaches the calibration value .
  • the force recognition system includes a first force recognition device and/or a second force recognition device.
  • the first force recognition device includes a plurality of first sensors, and at least one first sensor is arranged on each mechanical arm joint, and these first sensors are used to detect the force received by the revolving joint.
  • the control system calculates the force received by the end of the manipulator based on the information of the force of all the manipulator joints fed back by the first force identification device.
  • the control system can calculate the force received by the end of the robot arm according to a method known in the art.
  • the control system can use any existing PLC controller, single-chip microcomputer, microprocessor, FPGA, and those skilled in the art can know how to choose based on the disclosure of this application in combination with common knowledge in the field.
  • the second force recognition device includes at least one second sensor.
  • the second sensor is arranged on the end of the robot arm, for example, on the fourth revolute joint 115 of the robot arm, so as to directly detect the force received by the end of the robot arm.
  • the second force recognition device can directly obtain the force information of the end of the mechanical arm without calculation, which is simpler and more convenient to use, and more accurate.
  • the force received by the end of the robotic arm includes force and moment information.
  • the first force recognition device includes at least one of a position sensor device, a pressure sensor device, and a grating sensor device.
  • the position sensing device includes an absolute position sensing device and a relative position sensor device.
  • Absolute position sensor devices include absolute encoders, such as optical encoders, absolute magnetic encoders, absolute rotary transformer encoders or rotary potentiometers.
  • the relative position sensing device includes an incremental encoder.
  • the pressure sensing device includes a pressure sensor.
  • the grating sensing device includes a distributed grating sensor. That is, the first force recognition device includes one or more first sensors.
  • the type of the first sensor arranged at each revolute joint on the robot arm 110 may be the same or different.
  • a pressure sensor is provided on some revolving joints
  • a dual encoder is provided on other revolving joints.
  • Both dual encoders and pressure sensors are arranged on the same revolute joint.
  • one or more combinations of dual codes, pressure sensors, and grating sensors are set on the same revolute joint.
  • the second force recognition device preferably includes a six-axis sensor, and the number may be one or more.
  • FIG. 3 is a general flowchart of bone registration performed by the surgical robot 10 in this embodiment. As shown in FIG. 3, the process of performing bone registration by the surgical robot 10 includes the following steps: first enter the bone registration mode and start the bone registration S00; then, perform the following steps.
  • Step S01 Select the operating mode of bone registration, for example, select the mode in which the surgical robot automatically completes the bone registration, or select the mode in which the surgical robot assists in completing the bone registration.
  • Step S02 After specifying the operating mode, the navigation system 200 sends the information of the registration point of interest to the robotic arm system 100. Specifically, during preoperative planning, the navigation system 200 selects the registered points of interest based on the image information collected before the operation, patient information, prosthesis information, or feature points selected by the operator, and selects the registered points of interest. The information of the registration point is sent to the robotic arm system. More specifically, the navigation system 200 reconstructs the three-dimensional bone model according to the CT registration result, where the algorithm used can be a multi-slice reconstruction or a maximum density projection reconstruction method, and selects near the characteristic points according to the characteristic points selected by the operator Points of interest.
  • the algorithm used can be a multi-slice reconstruction or a maximum density projection reconstruction method
  • the navigation system 200 After selecting the registration point of interest, the navigation system 200 sends the position information of the registration point of interest in space relative to the base coordinates of the surgical robot to the robotic arm system 100, that is, the navigation system 200 sends the position of the registration point of interest to the robot arm system 100. The information is sent to the robotic arm system 100.
  • Step S03 After receiving the position information of the registration point of interest, the robotic arm system 100 plans a movement trajectory so that the end of the robotic arm moves to the registration point of interest along the preset movement trajectory.
  • the robot arm 110 is connected to the target pen and performs trajectory planning according to the position of the registration point of interest.
  • the robot arm system 100 preferably uses the robot's forward and inverse kinematics equations and dynamic equations to complete the trajectory planning.
  • Step S04 After planning the movement trajectory, according to the selected bone registration operation mode, the robot arm 110 drives the target pen to move to the registration point of interest, and selects the registration point on the bone.
  • the control system controls the robotic arm 110 to drive the target pen to move to the registration point of interest.
  • the operator operates the robotic arm 110 and drives the target pen to move to the registration point of interest.
  • step S05 is executed: the navigation system 20 collects the current position information of the target pen, and the control system determines whether the target pen has moved to the point of interest according to the current position information of the target pen The registration point, if yes, it is determined that the registration of the registration point has been completed, and the registration of the next registration point S06 can be performed, and the steps S03 to S05 are repeated until the bone registration of all the registration points of interest is completed.
  • the control system when judging whether the target pen has moved to the registration point of interest, the control system also judges whether the target pen has passed through the cartilage to the surface of the bone, that is, whether it has reached the surface of the hard bone, based on the information of the force received by the end of the robotic arm. . If you have passed through the cartilage to the hard bone surface, you can confirm that the target pen has moved to the registration point of interest and correctly selected the registration point on the bone. In some embodiments, the control system simultaneously judges whether the target pen has moved to the registration point of interest based on the current position information of the target pen and the information of the force received by the end of the robot arm.
  • control system preferentially judges whether the target pen has moved to the registration point of interest based on the information of the force received by the end of the robotic arm. Specifically, although the target pen has not moved to the target position, if the force received by the end of the robotic arm has reached the calibrated value at this time, it is considered that the target pen has moved to the registration point of interest. Further, if the target pen moves to the target position, but at this time the force received by the end of the mechanical arm has not reached the calibrated value, it is preferable that the operator determine whether to continue to puncture the bone surface.
  • the step of judging whether the target pen has moved to the registration point of interest has a higher priority than judging whether the target pen has moved to the registration point of interest according to the current position information of the target pen The priority of the point.
  • Fig. 4 is a flowchart of bone registration in the auxiliary operation mode in the preferred embodiment. As shown in Fig. 4, if the auxiliary operation mode is selected in step S01, the surgical robot 10 performs the following steps:
  • Step S101 the navigation system 200 sends the location information of the registration point of interest to the robotic arm system 100.
  • Step S102 After receiving the location information of the registration point of interest, under the guidance of the robotic arm 110, the operator operates the robotic arm 110 to drive the target pen to move to the registration point of interest, that is, the operator controls the robotic arm to drive The target pen pierces the cartilage until it reaches the surface of the bone, and selects a registration point on the bone.
  • the robotic arm system 100 can guide the operator to operate the robotic arm 110 through a method known in the art, such as an impedance control method.
  • step S1031 the control system judges whether the target pen moves in place according to the current position information of the target pen collected by the navigation system 200:
  • step S104 the operator intervenes and further determines whether to continue to puncture the bone surface
  • step S102 If not, continue to perform step S102.
  • step S1031 is performed, step S1032 is also preferably performed: the control system judges whether the cartilage piercing force reaches the calibrated value according to the information of the force received by the end of the manipulator:
  • step S105 If yes, it can be confirmed that the target pen has moved to the registration point of interest, that is, the result of step S105 is obtained;
  • step S102 If not, continue to perform step S102.
  • step S104 when performing step S104, if the operator thinks that the puncture to the bone surface is to be continued, it is preferable to perform step S1032; otherwise, if the operator believes that the puncture to the bone surface does not need to be continued, the result of step S105 can be directly obtained. This is because there may be slight differences between the actual patient’s bones and the results obtained by CT reconstruction, resulting in the fact that the actual bone surface reached and the surface of the three-dimensional model reconstructed by CT may be inconsistent. If only the force recognition system determines whether the bone surface has reached the bone surface, there is A certain risk, therefore, it is necessary to rely on the human judgment of the operator to reduce the risk of inaccurate selection of registration points, and to further ensure the accuracy of bone registration.
  • step S106 determine whether the navigation system 200 can normally collect the position information of the target pen:
  • step 200 drive the target pen to move, and register the next registration point
  • step S107 under the guidance of the robotic arm, the operator operates the robotic arm to adjust the posture of the target pen until the navigation system 200 can observe the position of the target pen.
  • Fig. 5 is a flowchart of bone registration in the automatic operation mode in the preferred embodiment. As shown in FIG. 5, if the automatic operation mode is selected in the above step S01, the surgical robot 10 performs the following steps:
  • Step S201 the navigation system 200 sends the location information of the registration point of interest to the robotic arm system 100.
  • Step S202 After receiving the location information of the registration point of interest, the robotic arm system 100 plans a movement trajectory.
  • Step S203 After planning the movement trajectory, the control system controls the robotic arm 110 to drive the target pen to move to the registered point of interest along the movement trajectory, that is, the target pen punctures the cartilage under the drive of the robotic arm until it reaches the surface of the bone, and selects on the bone Registration point.
  • step S2041 the control system judges whether the target pen moves in place according to the current position information of the target pen collected by the navigation system 200:
  • step S205 the operator intervenes and further determines whether to continue to puncture the bone surface
  • step S203 If not, continue to perform step S203.
  • step S2042 the control system determines whether the cartilage piercing force reaches the calibrated value according to the information of the force received by the end of the robotic arm:
  • step S206 If yes, it can be directly confirmed that the target pen has moved to the registration point of interest, that is, the result of step S206 is obtained;
  • step S205 if the operator thinks that the puncture to the bone surface is to be continued, step S2042 is preferably performed; on the contrary, if the operator believes that the puncture to the bone surface is unnecessary, the result of step S206 can be obtained.
  • step S206 the robotic arm system 100 sends a completion signal to the navigation system 200, and the navigation system 200 collects the position information of the target pen according to the completion signal, and executes step S207: judging whether the navigation system 200 can collect the target normally Pen location information:
  • step S200 drive the target pen to move, and register the next registration point
  • step S208 the control system controls the movement of the robotic arm to adjust the posture of the target pen until the navigation system 200 can observe the position of the target pen.
  • the robotic arm 110 drives the target pen to perform a rotational movement with the same spatial position, that is, the spatial position of the target pen remains unchanged and only the posture is changed.
  • the robot arm part of the rotating joint remains stationary, the other part of the rotating joint drives the target pen to rotate around an active stationary point, until the navigation system 200 can normally collect the position information of the target pen and then the robot arm stops moving. This is convenient for smooth operation. Perform bone registration at a registration point locally.
  • the robotic arm can guide the operator to hold the target pen to rotate around the active immobile point, and adjust the posture of the target pen.
  • the navigation system 200 includes a camera/optical sensing system 210 and an optical target 220.
  • the camera 210 is, for example, a binocular camera.
  • An optical target 220 is mounted on a target pen, and another optical target 220 Installed on the patient's bones and immobilized.
  • the posture of the target pen can be changed to make the optical target 220 rotate around the active immobile point to enter the field of view of the camera 210, namely As shown in Figure 6b.
  • the position sensing device includes two encoders (or called dual encoders), an absolute encoder 301 and an incremental encoder 302, respectively.
  • two encoders or called dual encoders
  • an absolute encoder 301 and an incremental encoder 302, respectively.
  • the torque detection of a revolute joint Take the torque detection of a revolute joint as an example.
  • a driver 303 and a reducer 304 are provided at the revolute joint.
  • the output end of the reducer 304 is connected to the revolving joint, and the input end of the reducer 304 is connected to the driver 303.
  • the driver 303 drives the revolute joint to rotate through the speed reducer 304.
  • the driver 303 includes a drive motor.
  • One of the absolute encoder 301 and the incremental encoder 302 is installed at the output end of the drive motor, and the other encoder is installed at the output end of the reducer 304 (such as the output shaft of a rotating joint) .
  • the absolute encoder 301 collects the absolute state parameter T1 of the revolute joint in real time
  • the incremental encoder 302 collects the relative state parameter T2 of the revolute joint in real time. Both the absolute state parameter T1 and the relative state parameter T2 are sent to the control system.
  • the output torque ⁇ of the drive motor should be equal to the external force torque ⁇ 'received by the revolute joint, and the output torque ⁇ of the drive motor is related to the deformation of the revolute joint:
  • K is the stiffness model of the reducer 304, such as the stiffness model of the harmonic reducer
  • ⁇ is the deformation of the rotating joint.
  • the deformation of the revolute joint has a corresponding relationship with the absolute state parameter T1 and the relative state parameter T2, and the corresponding relationship can be calibrated through experiments in advance. Or, calculated by the following formula:
  • T1 is the absolute position of the revolute joint fed back by the absolute encoder, the absolute position is, for example, the absolute rotation angle;
  • T2 is the relative position of the revolute joint fed back by the incremental encoder, and the relative position is, for example, the relative rotation angle ;
  • C1 is the resolution of the absolute encoder;
  • c2 is the resolution of the incremental encoder;
  • S0 is the startup initialization position of the incremental encoder.
  • control system can obtain the output torque ⁇ of the drive motor according to the deformation of the revolute joint and the stiffness model of the reducer, and can determine the external torque ⁇ 'of the revolute joint according to the output torque of the drive motor.
  • two encoders are respectively installed at the input end of the reducer 304 and the output end of the reducer.
  • the reducer and other transmission elements can be equivalent to an elastic body 306.
  • the equivalent elastic body 306 between the two encoders will be deformed, resulting in a deviation in the data collected between the two encoders, and this deviation is related to the external force. Therefore, through test calibration, the mapping relationship between the reading deviation of the two encoders and the external force can be obtained. Then use the driver to obtain the readings of the two encoders, and then upload the readings to the control system.
  • the control system obtains the deviation value of the two encoders according to the reading, and can obtain the external force on the joint according to the deviation value.
  • the joint module may also include at least one torque sensor 305.
  • 305 may be an optical mechanical deformation type, an electromagnetic induction type, or a resistance strain type, etc.
  • the torque information received by the revolute joint is directly detected by the torque sensor 305.
  • the position sensing device and the torque sensor can be mutually verified to monitor the working status of the two, so as to further improve the accuracy of the torque recognition of the revolute joint.
  • the torque sensor 305 can be installed on the output end of the reducer, such as a joint output shaft. Further, the output current of the driving motor can also be collected, so that the control system can obtain the external force information received by the rotating joint according to the output current of the driving motor, so as to realize redundant detection.
  • the torque sensor 305 is a pressure sensor.
  • the pressure sensor Before leaving the factory, the pressure sensor can be calibrated to the zero point. Specifically, the external force is applied to the rotating joint of the robotic arm and the output data of the pressure sensor is read, so that the pressure sensor can be zero-calibrated according to the measured output data. Further in use, when the rotating joint of the mechanical arm is subjected to an external force, the deformation of the pressure sensor is converted into an electric signal and sent to the control system, and the control system can calculate the external force received by the rotating joint according to the electric signal of the pressure sensor.
  • the torque sensor 305 may also be a distributed grating sensor.
  • the distributed grating sensor is arranged at the revolving joint and can be connected to the housing of the mechanical arm.
  • the grating sensor includes a fiber grating strain gauge, which constitutes a strain bridge.
  • the bridge layout method can refer to the existing resistance strain bridge layout method.
  • fiber grating strain gauges are installed at each rotating joint, and the fiber grating strain gauges of each joint module share one fiber channel. More specifically, the broadband light source placed at the back end of the robotic arm (such as the base) emits modulated light, passes through the fiber grating strain gauges of each joint module, and is reflected back. The control system demodulates and calculates the wavelength of the reflected light. .
  • the wavelength of the reflected light will change.
  • the amount of wavelength change is proportional to the strain of the distributed grating sensor. From this, the strain of each distributed grating sensor can be measured. The strain amount of the distributed grating sensor is calculated to obtain the external force experienced by the joint module.
  • the force recognition system can further verify the accuracy of the torque information. That is, the force recognition system preferably includes a first force recognition device and a second force recognition device, wherein the first force recognition device may include a variety of first sensors to detect torque information in different ways, such as the first force recognition device at the end of the robotic arm.
  • the second force recognition device directly senses the torque information at the end of the robotic arm through the six-axis sensor, while the first force recognition device can sense the torque received by the rotating joint through one or more of dual encoders, pressure sensors, and grating sensors. Information, the first force recognition device can also obtain torque information received by the joint by detecting the output current of the drive motor.
  • the force information at the end of the robotic arm can be obtained through multiple detection methods during the bone registration process, and the detection information can be verified to determine whether the torque information used in the calibration calculation is accurate.
  • the detection information received by the rotating joint it can also protect the joint module.
  • the robot arm has an active adjustment mode and/or a passive adjustment mode.
  • the power mechanism on the robot arm 110 is such as The working mode of driving the motor to drive the movement of the mechanical arm is the active adjustment of the mechanical arm.
  • the working mode of driving the movement of the mechanical arm by the Cartesian force is the passive adjustment of the mechanical arm.
  • the control system obtains the torque of each revolute joint on the robotic arm, and then controls all the drive motors on the revolute joints whose torque is greater than the preset calibration value (that is, the drive torque is greater than the preset calibration value).
  • the drive motor of the joint stop output.
  • the calibration value is, for example, zero. In this way, the problem of mechanical posture adjustment caused by human misoperation during the active adjustment mode operation of the mechanical arm can be avoided, thereby increasing the safety and reliability of the mechanical arm during use.
  • the control system obtains the torque of each joint on the manipulator, it then controls all the drive motors on the revolute joints whose torque is greater than the preset calibration value (the calibration value is zero, for example).
  • the drive motor on the revolute joint whose torque on the robotic arm is greater than the preset calibration value can cooperate with the operator to drive the revolute joint movement of the robotic arm, so that the operator can overcome the resistance of the revolute joint on the robotic arm, so that The control system can assist the operator to smoothly move the mechanical arm to a predetermined position. In this way, the comfort and convenience of the passive adjustment operation of the mechanical arm are increased.
  • the drive motors on the revolute joints whose torque on the robotic arm is greater than the preset calibration value (the calibration value is for example zero) output power, so that the operator It is more labor-saving and simple to adjust the mechanical arm passively.
  • each force recognition device transmits the torque information obtained to the control system, and the control system compares the calculations based on the dual encoders
  • the force received by the end of the robotic arm is the same as the force received by the end of the robotic arm directly detected by the six-axis sensor. That is, the control system verifies the received torque information and determines whether to update the calibration based on the verification result. According to the verification results, the control system can also take corresponding safety protection measures. For example, using the six-axis sensor as the calibration reference, update the stiffness coefficient K of the reducer in the dual encoder, or update the relevant parameters in the pressure sensor and the grating sensor.
  • control system is also used to determine whether the torque sensor exceeds the range of the torque sensor according to the received torque information, protect the torque sensor, and further ensure the accuracy of the joint torque information during the bone registration process, thereby further determining the cartilage puncture force information To ensure the accuracy of bone registration.
  • Fig. 9 is a flowchart of bone registration in a preferred embodiment.
  • step S301 is executed when the target pen moves to the registration point of interest
  • step S302 is also executed: the control system determines whether the torque sensors of each joint are working properly; if not, then Stop the automatic operation mode or auxiliary operation mode, and change to manual registration S303.
  • the operator manually drives the target pen to perform bone registration; if so, execute step S304: the control system continues to control the robotic arm to drive the target pen to the Register point movement, and judge whether the cartilage puncture force reaches the calibration value, that is, whether the torque value fed back by the six-axis sensor reaches the calibration value.
  • step S306 continue to drive the target pen to penetrate cartilage and other tissues. During this process, return to step S304 at any time , Until the puncture cartilage force reaches the calibration value.
  • two empirical thresholds can be set in the control system.
  • the first threshold is used for mutual calibration of sensors, and the second threshold is used to determine whether the torque sensor exceeds its range.
  • the control system determines whether the sensors on each joint module work normally according to the two thresholds. Specifically, the control system compares the force value received by the end of the robot arm measured according to the detection model of the dual encoder with the force value received by the end of the robot arm measured by the six-axis sensor. The deviation of the latter two exceeds the range defined by the first threshold, it indicates that the sensor is abnormal, and the relevant parameters of the sensor need to be updated. If the deviation of the two is within the range of the first threshold, it indicates that the sensor is working normally.
  • the data detected by the force recognition system can be used to control the depth of insertion of the target pen into the cartilage.
  • the present invention does not limit the specific size of the two thresholds, and can be determined through experiments according to the performance, detection accuracy and other requirements of the sensor.
  • the surgical robot 10 of this embodiment also provides a function of maintaining the force recognition system before surgery.
  • Fig. 10 is a flowchart of maintaining the force recognition system in a preferred embodiment.
  • step S401 before leaving the factory, the deviation of the encoder when each mechanical arm rotating joint receives an external force can be recorded and stored in the control system.
  • step S402 before leaving the factory, a verification program can be started to verify the reliability of the encoder. After the verification, in step S403, the surgical robot is automatically run and the verification parameters are updated to ensure that the encoder works normally.
  • step S404 it is also necessary to calibrate the data information collected by other torque sensors and ensure the accuracy of the collected data information, that is, to update the parameters of each sensor on the joint module according to the six-axis sensor. And to ensure the accuracy of these sensors.
  • step S405 can be performed before the operation to compare the data collected by each sensor with the factory data to ensure the rigidity of the reducer.
  • the present invention mainly checks the main mechanical components in the surgical robot, that is, the harmonic reducer, to ensure that the harmonic reducer operates normally.
  • the database will record the empirical value range of the different stiffness of each robot product. According to the time and empirical value range of the product, as well as the original mechanical and hardware design parameters, confirm whether the torque sensor is still used for a certain period of time. Can work normally.
  • a third threshold can be set in the control system. If the detection result of the torque sensor exceeds the third threshold, it is considered that the torque sensor cannot reach the required accuracy, and the torque sensor needs to be re-maintained and calibrated or replaced.
  • the third threshold is based on the empirical value range of different stiffness of each robot product, the time the product is used, and the original mechanical and hardware design parameters, such as the resolution and accuracy of dual encoders, torque sensors and six-axis torque sensors The resolution, zero drift, temperature drift, stiffness of the harmonic reducer, reduction ratio, reduction ratio error, and the stiffness model of the connector are determined.
  • control system of the surgical robot 10 adopts a neural network calculation method to update the sensor model parameters.
  • a neural network calculation method to update the sensor model parameters.
  • online neural network learning is added to update the sensor model parameters.
  • the neural network calculation method can provide online model parameter updates to ensure that the force of each revolute joint is more reliable, and the reliability of the force recognition system is further improved.
  • the flow chart of online model parameter update using six-axis sensors is shown in Figure 11.
  • the coefficients of the torque sensor model of each joint and the stiffness model of the reducer are used as the neural network input parameters, and in process S509, the input parameters are used for online neural network learning to update the parameters of the known model .
  • Including the stiffness model of the reducer, the joint torque sensor model, and the Cartesian force at the end of the new manipulator is calculated according to the learning results in the process S510, and then the output result of the neural network (the calculated Compare the Cartesian force at the end of the new robotic arm with the data collected by the actual six-axis sensor (the actual Cartesian force at the end of the robotic arm) to confirm whether to update the model parameters of the sensor.
  • Fig. 12 is a schematic diagram of a process of obtaining a torque sensor model based on a neural network model provided by a preferred embodiment.
  • the neural network has a three-layer structure of input layer, hidden layer and output layer.
  • the input layer has three nodes, which are torque sensor parameters 1, 2, and 3; the hidden layer has two nodes; and the output layer has one node, which is used to output a single joint torque.
  • the input of the three nodes of the input layer corresponds to the stiffness of the harmonic reducer, the angular velocity of the joint and the angular position of the joint; if it is a torque sensor, the input parameters are the torque conversion coefficient, Torque sensor zero point and torque sensor temperature.
  • the node output of the output layer is the actual torque of the torque sensor.
  • a neural network model after online training that is, the weights and thresholds of all nodes are first obtained through training samples when leaving the factory.
  • the model parameters in the neural network are continuously updated.
  • the specific training method may adopt the gradient descent method or other existing technologies, which is not limited in the present invention.
  • this embodiment also provides a model update method that is closer to the actual physical meaning, and the flow can refer to FIG. 13.
  • the neural network has a three-layer structure of input layer, hidden layer and output layer. And the input layer has three nodes, the hidden layer has two nodes, and the output layer has one node. Among them, the input of the three nodes of the input layer corresponds to the angular velocity of the joint, the angular position of the joint, and the output torque of the joint motor. The node output of the output layer is the actual torque of the torque sensor.
  • the least square method is used to update the parameter coefficients of the dual-encoder torque model and the torque sensor based on the obtained torque results, and use the updated coefficients as calculation confirmation.
  • control system of this embodiment includes a memory and a processor, and a program is stored in the memory, so that the processor can perform the above functions by running the program.
  • the surgical robot utilizes the force recognition system to achieve the purpose of automatically completing or assisting in completing the bone registration by the surgical robot, greatly improving the efficiency of bone registration, and at the same time improving the bone registration. Registration accuracy and success rate. Furthermore, the use of multiple force recognition devices ensures the accuracy of the cartilage puncture force during the operation, and improves the accuracy of the operation. Furthermore, the surgical robot also provides maintenance methods after leaving the factory, which can effectively protect the core components of the surgical robot, and confirm the operating status and component status of the robot to ensure that there will be no problems during the operation. Furthermore, for the six-axis sensor, the surgical robot also provides a method for online neural network learning and updating model parameters. This method can provide online model parameter updates to ensure that the force of each joint is more reliable, and the force recognition system is more reliable. Reliability has been further improved.
  • the surgical robot also has an online verification function, which can perform verification on the machine after the factory is used, to ensure that the state of the robot before the operation is intact and to ensure the smooth progress of the operation.

Landscapes

  • Health & Medical Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Surgery (AREA)
  • Robotics (AREA)
  • Animal Behavior & Ethology (AREA)
  • Public Health (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Medical Informatics (AREA)
  • Molecular Biology (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • General Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Veterinary Medicine (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Human Computer Interaction (AREA)
  • Manipulator (AREA)
  • Surgical Instruments (AREA)

Abstract

一种可读存储介质以及手术机器人（10），手术机器人（10）包括控制***、力识别***、机械臂***（100）和导航***（200）；机械臂***（100）包括机械臂（110），机械臂（110）的末端用于可拆卸地连接靶标笔；导航***（200）用于获取预定对象上感兴趣的注册点并发送给机械臂***（100）；机械臂***（100）用于根据预定对象上感兴趣的注册点，控制机械臂（110）运动，以使机械臂（110）驱动靶标笔运动，使靶标笔运动至预定对象上感兴趣的注册点；力识别***设置在机械臂（110）上，用于检测机械臂（110）末端所受到的作用力并发送给控制***；控制***用于根据机械臂（110）末端所受到的作用力，判断靶标笔是否已运动到预定对象上感兴趣的注册点。因此，可以提高骨注册的准确性。

Description

骨注册方法、手术机器人及可读存储介质 技术领域
本发明涉及医疗器械技术领域，特别涉及一种骨注册方法、手术机器人以及可读存储介质。
背景技术
计算机辅助导航***已成为现代外科技术的重要组成部分，帮助操作者更精确、更安全地进行多种复杂手术，在骨科各个方面具有许多不可替代的优越性，因其精确性、安全性、低辐射等特点使其在临床实践中逐渐得到广泛应用。计算机辅助导航***可用于引导操作人员进行手术训练、制定手术计划、实时导航手术器械和减少病人创伤。在人工关节置换领域，计算机辅助，使截骨更准确，假体安装更精确，术后力线恢复更接近生理状态。
一般的，手术机器人在使用时需要事先进行骨注册，建立实体骨骼与虚拟影像之间的对应关系，为操作者的截骨、骨面处理及假体放置等操作提供三维信息。以膝关节截骨操作为例，传统的骨注册方法为操作人员手持靶标笔，在患者骨头上选择注册点，完成骨注册。但是，操作人员操作靶标笔很难准确把握***软骨的深度，导致无法正确地选中骨头上的注册点，降低了骨注册的准确性，而且手动注册过程费时费力，效率低下，且容易产生注册偏差。
因此，有必要提供一种手术机器人，能够自动完成或辅助完成骨注册过程，提高骨注册的准确性以及注册效率。
发明内容
为实现上述目的，本发明提供一种手术机器人执行骨注册的方法、手术机器人以及可读存储介质，用于自动完成或辅助操作人员完成骨注册过程， 提高骨注册效率，帮助操作人员减轻工作负担，特别地，还能够识别机械臂末端的受力状况，以此保证准确地选中骨头上的注册点，提高了骨注册的准确性和安全性。
根据本发明的一个方面，提供了一种用于手术机器人的骨注册方法，包括以下步骤：
机械臂***接收导航***所选择的预定对象上感兴趣的注册点的信息；
控制机械臂***中的机械臂运动，以使所述机械臂控制跟踪装置运动至所述预定对象上感兴趣的注册点；
获取由力识别***检测的机械臂末端所受到的作用力；以及
根据所述机械臂末端所受到的作用力，判断所述跟踪装置是否已运动到所述预定对象上感兴趣的注册点。
可选的，在所述方法中，还包括以下步骤：
获取由导航***采集的所述跟踪装置的当前位置信息；
根据所述跟踪装置的当前位置信息，判断所述跟踪装置是否已运动到预定对象上感兴趣的注册点。
可选的，在所述方法中，根据所述机械臂末端所受到的作用力，以判断所述跟踪装置是否已运动到预定对象上感兴趣的注册点的优先级高于根据所述跟踪装置的当前位置信息，判断所述跟踪装置是否已运动到预定对象上感兴趣的注册点。
可选的，在所述方法中，当所述程序被运行时，所述根据所述机械臂末端所受到的作用力，判断所述跟踪装置是否已运动到预定对象上感兴趣的注册点的步骤包括：
判断力识别***是否工作正常；
若是，则控制所述机械臂驱动所述跟踪装置向所述注册点运动。
可选的，在所述方法中，若所述力识别***判断为工作正常，则在控制所述机械臂驱动所述跟踪装置向所述注册点运动的步骤之后，还包括：
将所述机械臂的末端所受到的作用力与标定值进行比较；
若等于或大于所述标定值，则确定所述跟踪装置已运动到预定对象上感兴趣的注册点；
若小于所述标定值，则继续控制所述机械臂驱动所述跟踪装置向所述注册点运动，直至所述机械臂末端所受到的作用力等于或大于所述标定值。
可选的，在所述方法中，在确定所述跟踪装置已运动到预定对象上感兴趣的注册点之后，还包括以下步骤：
判断所述导航***是否能够正常采集所述跟踪装置的位置信息；
若是，控制所述机械臂以驱动所述跟踪装置向下一个感兴趣的注册点运动；
若否，控制所述机械臂以调整所述跟踪装置的姿态，直至所述导航***能够正常采集所述跟踪装置的位置信息。
可选的，在所述方法中，在控制机械臂运动之前，还包括以下步骤：
根据所述导航***所选择的预定对象上感兴趣的注册点，所述机械臂***规划一运动轨迹，所述机械臂驱动所述跟踪装置沿所述运动轨迹运动至感兴趣的注册点。
可选的，在所述方法中，在判断所述跟踪装置是否已运动到预定对象上感兴趣的注册点的步骤之前，还包括以下步骤：
根据所述力识别***获取的每个机械臂关节所受到的作用力，计算得到所述机械臂的末端所受到的作用力；
获取所述力识别***所得到的所述机械臂末端所受到的作用力；
比较计算得到的所述机械臂的末端所受到的作用力，与利用力识别***所得到的所述机械臂末端所受到的作用力；
若两者的偏差超出预定范围，则更新所述力识别***。
可选的，在所述方法中，所述力识别***获取每个机械臂关节所受到的作用力的步骤包括：
根据安装在每个所述机械臂关节的第一传感器，检测每个所述机械臂关节所受到的作用力。
可选的，在所述方法中，所述力识别***得到所述机械臂末端所受到的作用力的步骤包括：
根据安装在所述机械臂末端的第二传感器，检测所述机械臂末端所受到的作用力。
可选的，在所述方法中，当所述程序被运行时，采用神经网络计算的方法计算所述机械臂末端所受到的作用力，且在输出所述机械臂末端所受到的作用力之前，还包括对所述力识别***进行神经网络训练的步骤，并保持神经网络训练后的网络结构以及权值文件。
根据本发明的另一个方面，还提供一种手术机器人，其包括控制***、力识别***、机械臂***和导航***；所述机械臂***包括机械臂；所述机械臂的机械臂末端用于连接跟踪装置；所述控制***分别与所述力识别***、所述机械臂***和所述导航***通信连接，所述导航***与所述机械臂***通信连接；
所述导航***被配置为将预定对象上感兴趣的注册点的信息发送给所述机械臂***；所述机械臂***被配置为根据所述预定对象上感兴趣的注册点 的信息，控制所述机械臂运动，以驱动所述跟踪装置运动至预定对象上感兴趣的注册点；
所述力识别***设置在所述机械臂上，所述力识别***用于检测所述机械臂末端所受到的作用力并发送给所述控制***；所述控制***被配置为根据所述机械臂末端所受到的作用力，判断所述跟踪装置是否已运动到预定对象上感兴趣的注册点。
在所述手术机器人中，所述导航***被配置为：采集所述跟踪装置的当前位置信息并发送给所述控制***；所述控制***被配置为根据所述跟踪装置的当前位置信息，判断所述跟踪装置是否已运动到预定对象上感兴趣的注册点。
在所述手术机器人中，所述控制***被配置为，根据所述机械臂末端所受到的作用力，以判断所述跟踪装置是否已运动到预定对象上感兴趣的注册点的优先级高于根据所述跟踪装置的当前位置信息，以判断所述跟踪装置是否已运动到预定对象上感兴趣的注册点的优先级。
在所述手术机器人中，所述控制***被配置为，判断力识别***是否工作正常；若是，则控制所述机械臂驱动所述跟踪装置向所述注册点运动。
在所述手术机器人中，若所述力识别***判断为工作正常，所述控制***被配置为将所述机械臂末端所受到的作用力与所述控制***内预存的标定值进行比较；若等于或大于所述标定值，则确定所述跟踪装置已运动到预定对象上感兴趣的注册点；若小于所述标定值，则继续控制所述机械臂驱动所述跟踪装置向所述注册点运动，直至所述机械臂末端所受到的作用力等于或大于所述标定值。
在所述手术机器人中，在确定所述跟踪装置已运动到预定对象上感兴趣 的注册点之后，所述控制***还被配置为判断所述导航***是否能够正常采集所述跟踪装置的位置信息；若是，控制所述机械臂驱动所述跟踪装置向下一个感兴趣的注册点运动；若否，控制所述机械臂调整所述跟踪装置的姿态，直至所述导航***能够正常采集所述跟踪装置的位置信息。
在所述手术机器人中，所述机械臂包括多个机械臂关节，所述力识别***包括第一力识别装置和第二力识别装置；
所述第一力识别装置包括多个第一传感器，每个所述机械臂关节处设置至少一个所述第一传感器；所述第一传感器用于检测对应的机械臂关节所受到的作用力；
所述第二力识别装置包括至少一个设置在所述机械臂末端的第二传感器，所述第二传感器用于检测所述机械臂末端所受到的作用力；
所述控制***被配置为：
根据多个所述第一传感器检测到的所述机械臂关节所受到的作用力，计算得到所述机械臂末端所受到的作用力；比较计算得到的所述机械臂末端所受到的作用力，与所述第二传感器所检测得到的所述机械臂末端所受到的作用力；
若两者的偏差超出预定范围，则更新所述力识别***。
在所述手术机器人中，所述第一传感器包括双编码器、力矩传感器以及分布式光栅传感器中的至少一种，所述第二传感器包括六轴传感器。
在所述手术机器人中，每个所述机械臂关节处设置有双编码器，所述双编码器包括绝对式编码器和增量式编码器；所述机械臂还包括用于驱动所述机械臂关节的驱动器和减速器，所述驱动器的输出端连接所述减速器的输入端，所述减速器的输出端连接所述机械臂关节；其中，所述绝对式编码器和 增量式编码器中的一个设置在所述减速器的输入端，另一个设置在所述减速器的输出端。
在所述手术机器人中，所述机械臂关节所受到的作用力τ的计算公式为：
τ＝Kδ
Figure PCTCN2020095666-appb-000001
其中，K为所述减速器的刚度模型；δ为所述机械臂关节的变形量；T1为所述绝对式编码器所反馈回的所述机械臂关节的绝对位置；T2为所述增量式编码器所反馈回的所述机械臂关节的相对位置；c1为所述绝对式编码器的分辨率；c2为所述增量式编码器的分辨率；S0为所述增量式编码器的启动初始化位置。
在所述手术机器人中，所述双编码器与所述控制***通信连接；所述控制***用于根据所述双编码器的数据偏差与所述机械臂关节所受到的作用力的映射关系，获取所述机械臂关节所受到的作用力。
在所述手术机器人中，所述控制***被配置为采用神经网络计算的方法计算所述机械臂末端所受到的作用力，且在输出所述机械臂末端所受到的作用力之前，对所述力识别***进行神经网络训练，并保持神经网络训练后的网络结构以及权值文件。
在所述手术机器人中，所述手术机器人包括自动运行模式和辅助运行模式；
当所述手术机器人处于所述自动运行模式时，所述机械臂***根据接收到的感兴趣的注册点的信息，规划运动轨迹，并控制所述机械臂驱动所述跟踪装置沿着所述运动轨迹运动至预定对象上感兴趣的注册点；
当所述手术机器人处于所述辅助运行模式时，由外力驱动所述机械臂运动并带动所述跟踪装置运动至预定对象上感兴趣的注册点。
根据本发明的另一个方面，还提供一种可读存储介质，其上存储有程序，当所述程序被处理器运行时执行如上所述的用于手术机器人的骨注册方法。
在本发明提供的用于手术机器人的骨注册方法以及手术机器人中，使用手术机器人进行骨注册，便于提高骨注册效率，帮助操作人员减轻工作负担，尤其地，在骨注册过程中，借助于力识别***可以识别机械臂末端的受力状况，便于根据机械臂的受力状况来判断跟踪装置穿刺软骨的情况，从而控制***软骨的深度，确保正确地选中骨头上的注册点，提高了骨注册的准确性和安全性。进一步，使用力识别***中的多种传感器来分别获取机械臂末端所受到的作用力信息，便于将这些作用力信息进行相互校验，确保受力信息检测的准确性和可靠性。而且在骨注册过程中，还进一步判断力识别***是否工作正常，若不正常可以手动进行骨注册，这样进一步确保了受力检测的准确性和可靠性。更进一步，还可根据神经网络的计算方法得到机械臂末端所受到的作用力信息，进一步提高了受力检测的精度。
附图说明
本发明的实施方法以及相关实施例的特征、性质和优势将通过结合下列附图进行描述，其中：
图1为本发明实施例中的手术机器人在执行手术过程的整体示意图；
图2为本发明实施例中的机械臂的结构示意图；
图3为本发明实施例中的手术机器人进行骨注册的总流程图；
图4为本发明实施例中的手术机器人处于辅助运行模式时进行骨注册的流程图；
图5为本发明实施例中的手术机器人处于自动运行模式时进行骨注册的流程图；
图6a为本发明实施例中的靶标笔在导航***视野外的示意图；
图6b为本发明实施例中的靶标笔在导航***视野内的示意图；
图7为本发明实施例中的关节模组的示意图；
图8为本发明实施例中的采用双编码器测量关节力矩的原理图；
图9为本发明实施例中对关节模组进行保护的流程图；
图10为本发明实施例中出厂前对力矩传感器进行维护标定的流程图；
图11为本发明实施例中利用六轴传感器进行在线模型参数更新的流程图；
图12～图13分别为本发明实施例中利用神经网络进行参数更新的原理图。
具体实施方式
以下将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
如在本发明中所使用的，单数形式“一”、“一个”以及“该”包括复数对象，除非内容另外明确指出外。如在本发明中所使用的，术语“或”通常是以包括“和/或”的含义而进行使用的，除非内容另外明确指出外。如在本发明中所使用的，术语“若干”通常是以包括“至少一个”的含义而进行使用的，除非内容另外明确指出外。如在本发明中所使用的，术语“至少两个”通常是以包括“两个或两个以上”的含义而进行使用的，除非内容另外明确指出外。此外，术语“第一”、“第二”、“第三”仅用于描述目的，而不能理解为指示或暗示相对重要性或者隐含指明所指示的技术特征的数量。由此，限定有“第一”、“第二”、“第三”的特征可以明示或者隐含地包括一个或者至少两个该特征。
本发明的主要目的之一是提供一种手术机器人，旨在通过手术机器人进行协同操作完成骨注册过程，并借助于力识别***，能够获取机械臂末端所受到的作用力信息(包括力和力矩信息)，确保靶标笔/跟踪装置能够准确地选中骨头上的注册点，保证骨注册的准确性和安全性。除此之外，本发明中使用的传感器受到过大外力的情况下而不会发生损坏，确保传感器使用的可靠性。尤其地，传感器之间还可以相互校验，进一步保证骨注册的准确性和可靠性。
图1为本实施例中的手术机器人10在操作实施膝关节手术时的整体示意图。如图1所示，本实施例提供一种手术机器人10，其包括机械臂***/机械臂子***100和导航***/导航子***200，该两个***相互通信。本实施例中，所述机械臂***100具有自动运行模式和辅助运行模式。当机械臂***100处于自动运行模式时，由手术机器人10全自动完成骨注册过程；当机械臂***100处于辅助运行模式时，由手术机器人10引导操作人员完成骨注册过程。而所述导航***200被用于术前制定手术计划和术中导航，可以在手术过程中跟踪机械臂末端的器械，并将器械的位置在病人术前和术中的影像实时更新显示出来，让操作人员随时知道器械的位置同目标对象的关系，使操作人员更精确的进行手术，因此，导航***200即是手术机器人10的视觉识别部分。
图2为本实施例中的机械臂110的结构示意图。如图2所示，所述机械臂***100包括机械臂110，所述机械臂110包括若干依次连接的机械臂关节以及一机械臂末端，机械臂关节如可为转动关节，用于控制机械臂末端所驱动的手术工具的空间位置和姿态。本实施例中，所述机械臂110至少包括依次连接的五个转动关节，以实现至少五个自由度。进一步，任意相邻两个转动关节通过连接臂连接。
以图2所示的机械臂110为例，所述机械臂110具有五个自由度，并包括依次连接的第一转动关节111、第二转动关节112、自转关节113、第三转动关节114和第四转动关节115。并且，第一转动关节111的旋转轴线、第二转动关节112的旋转轴线和第三转动关节114的旋转轴线相互平行；第三转动关节114和第四转动关节115的旋转轴线相互垂直。进一步，任意相邻两个转动关节之间通过连接臂116连接。其中，所述第一转动关节111连接一个连接臂116，该连接臂116的一端连接第一转动关节111，另一端固定连接底座，底座可以固定在病床20上，也可以固定在手术台车120上。第四转动关节115连接手术工具或者靶标笔。本实施例中，所述第四转动关节115连接靶标笔，从而通过机械臂110来连接靶标笔，并通过机械臂带动靶标笔运动，并在患者骨头上选择感兴趣的注册点，完成骨注册过程。
进一步，所述手术机器人10还包括力识别***，力识别***设置在机械臂110上，用于检测机械臂末端所受到的作用力。所述机械臂末端所受到的作用力也就是靶标笔穿刺软骨时或者进行骨注册时所受到的作用力。实际应用时，便于根据机械臂末端所受到的作用力，判断靶标笔是否已穿刺到硬骨表面，这样做，便于准确控制靶标笔***软骨的深度，从而正确地选中骨头上的注册点，提高骨注册的准确性和安全性。
本实施例中，所述手术机器人10还包括控制***，分别与力识别***、机械臂***100和导航***200通信连接。所述力识别***将其检测到的机械臂末端所受到的作用力的信息反馈至所述控制***，所述控制***根据所述机械臂末端所受到的作用力的信息，判断靶标笔是否已穿刺到骨头/硬骨表面，也即判断靶标笔是否运动到骨头上感兴趣的注册点。进一步，所述控制***内部存储有机械臂末端所受到的作用力的标定值，该标定值可通过实验完成标定。所述控制***将检测到的机械臂末端所受到的作用力与标定值进 行比较，若实际检测的作用力与标定值相符合(指实际检测的作用力达到标定值，即大于或等于标定值)，则可确定靶标笔已穿刺到骨头表面，并正确地选中了骨头上的注册点，反之，若实际检测的作用力未达到(即小于)标定值，则继续控制所述机械臂驱动所述靶标笔向所述注册点运动(如继续深入)，并重复将所述机械臂末端所受到的作用力与标定值进行比较，直至所述机械臂末端所受到的作用力达到所述标定值。
进一步，所述力识别***包括第一力识别装置和/或第二力识别装置。所述第一力识别装置包括多个第一传感器，且在每个机械臂关节上设置至少一个第一传感器，这些第一传感器均用于检测转动关节所受到的作用力。进而所述控制***根据第一力识别装置反馈回的所有机械臂关节的作用力的信息，计算得到机械臂末端所受到的作用力。此处，控制***可以根据本领域公知的方式计算得到机械臂末端所受到的作用力。此外，所述控制***可以采用现有任一PLC控制器、单片机、微处理器、FPGA，本领域技术人员可在本申请公开基础上结合本领域的公知常识能够知晓如何选择。所述第二力识别装置包括至少一个第二传感器，第二传感器设置在机械臂末端上，例如设置在机械臂的第四转动关节115上，以直接检测机械臂末端所受到的作用力。与第一力识别装置相比，通过第二力识别装置无需计算便可直接获取机械臂末端的受力信息，使用更为简单和方便，并且更加精确。本实施例中，所述机械臂末端所受到的作用力包括力和力矩信息。
进一步，所述第一力识别装置包括位置传感装置、压力传感装置以及光栅传感装置中的至少一种，当然也可以是两种或两种以上的传感装置组合使用。所述位置传感装置包括绝对位置传感装置和相对位置传感器装置。绝对位置传感器装置包括绝对式编码器，如光学编码器、绝对磁编码器、绝对旋转变压编码器或旋转电位器等。相应的，相对位置传感装置包括增量式编码 器。压力传感装置包括压力传感器。光栅传感装置包括分布式光栅传感器。也即，所述第一力识别装置包括一种或多种第一传感器。进一步，所述机械臂110上各个转动关节处所布置的第一传感器的类型可以相同或不相同，例如在一些转动关节上设置压力传感器，而在另一些转动关节上设置双编码器，优选，在同一个转动关节上既设置双编码器，又设置压力传感器。例如在同一个转动关节上设置双编码、压力传感器以及光栅传感器中的一种或多种组合。进一步，所述第二力识别装置优选包括六轴传感器，数量可以是一个或多个。
图3为本实施例中的手术机器人10进行骨注册的总流程图。如图3所示，所述手术机器人10执行骨注册的过程包括如下步骤：首先进入骨注册模式，开始骨注册S00；之后，执行以下步骤。
步骤S01：选择骨注册的运行模式，例如选择手术机器人自动完成骨注册的模式，或选择手术机器人辅助的方式完成骨注册的模式。
步骤S02：指定运行模式后，导航***200向机械臂***100发送感兴趣的注册点的信息。具体的，所述导航***200在进行术前规划时，根据术前采集的图像信息、病人信息、假体信息或操作者选择的特征点等，选择感兴趣的注册点，并将感兴趣的注册点的信息发送给机械臂***。更具体而言，导航***200根据CT配准结果重建三维骨头模型，其中，所采用的算法可以是多层面重建或最大密度投影重建方法，并根据操作者选择的特征点，在特征点附近选择感兴趣点。选择感兴趣的注册点后，导航***200将感兴趣的注册点在空间中相对于手术机器人的基坐标的位置信息发送给机械臂***100，也即导航***200将感兴趣的注册点的位置信息发送给机械臂***100。
步骤S03：在接收到感兴趣的注册点的位置信息后，机械臂***100规划运动轨迹，以使机械臂末端沿着预设的运动轨迹运动到感兴趣的注册点。在 规划运动轨迹时，机械臂110连接靶标笔并根据感兴趣的注册点的位置进行轨迹规划，规划过程中，机械臂***100优选运用机器人正逆运动学方程以及动力学方程完成轨迹规划。
步骤S04：规划运动轨迹后，根据所选择的骨注册运行模式，机械臂110驱动靶标笔运动至感兴趣的注册点，并选择骨头上的注册点。此处，如果是自动运行模式，则控制***控制机械臂110驱动靶标笔运动至感兴趣的注册点。如果是辅助运行模式，则在机械臂的引导下，操作人员操作机械臂110并带动靶标笔运动至感兴趣的注册点。
当靶标笔运动到感兴趣的注册点后，执行步骤S05：导航***20采集靶标笔的当前位置信息，且所述控制***根据靶标笔的当前位置信息，判断靶标笔是否已运动到感兴趣的注册点，若是，则确定已完成该注册点的注册，并可进行下一个注册点的注册S06，并返回重复执行步骤S03至S05，直至完成全部感兴趣的注册点的骨注册。
特别地，判断靶标笔是否运动到感兴趣的注册点时，所述控制***还根据机械臂末端所受到的作用力的信息，判断靶标笔是否已经穿过软骨到达骨头表面，即是否到达硬骨表面。若已经穿过软骨到达硬骨表面，则可确认靶标笔已运动到感兴趣的注册点并正确地选中了骨头上的注册点。在一些实施例中，所述控制***同时根据靶标笔的当前位置信息和机械臂末端所受到的作用力的信息，判断靶标笔是否已运动到感兴趣的注册点。进一步的，所述控制***优先根据机械臂末端所受到的作用力的信息判断靶标笔是否已运动到感兴趣的注册点。具体来说，虽然靶标笔未运动到目标位置，但如果此时机械臂末端所受到的作用力已达到标定值，则认为靶标笔已经运动到感兴趣的注册点。进一步，如果靶标笔运动到目标位置，但此时机械臂末端所受到的作用力还未达到标定值时，则优选由操作人员判断是否继续穿刺到骨表面。 因此，根据机械臂末端所受到的作用力，判断靶标笔是否已运动到感兴趣的注册点的步骤的优先级高于根据靶标笔的当前位置信息，判断靶标笔是否已运动到感兴趣的注册点的优先级。
图4为优选实施例中的辅助运行模式的骨注册流程图。如图4所示，如果在上述步骤S01中，选择辅助运行模式，则手术机器人10执行以下步骤：
步骤S101：导航***200向机械臂***100发送感兴趣的注册点的位置信息。
步骤S102：在接收到感兴趣的注册点的位置信息后，在机械臂110的引导下，操作人员操作机械臂110驱动靶标笔运动至感兴趣的注册点，也即由操作人员控制机械臂驱动靶标笔穿刺软骨直至到达骨头表面，并在骨头上选择注册点。该过程中，机械臂***100可通过本领域公知的方式如阻抗控制方法，引导操作人员操作机械臂110。
当靶标笔运动至感兴趣的注册点后，执行步骤S1031：控制***根据导航***200采集的靶标笔的当前位置信息，判断靶标笔是否运动到位：
若是，优选执行步骤S104：由操作人员介入并进一步判断是否继续穿刺到骨表面；
若否，继续执行步骤S102。
其中在执行步骤S1031的同时，还优选执行步骤S1032：控制***根据机械臂末端所受到的作用力的信息，判断穿刺软骨力是否达到标定值：
若是，即可确认靶标笔已运动到感兴趣的注册点，即获得步骤S105的结果；
若否，则继续执行步骤S102。
进一步地，在执行步骤S104时，若操作人员认为要继续穿刺到骨表面，则优选执行步骤S1032；反之，若操作人员认为不需要继续穿刺到骨表面，则 可直接获得步骤S105的结果。这是因为，实际的患者骨头与CT重建得到的结果可能会有细微的差别，致使实际到达的骨表面与CT重建的三维模型表面可能不一致，如果仅通过力识别***判断是否到达骨表面，存在一定的风险，因此，还需要依靠操作人员的人为判断去降低注册点选择不准确的风险，进一步保证骨注册的准确性。
进一步，在获得步骤S105的结果后，继续执行步骤S106：判定导航***200是否能够正常采集靶标笔的位置信息：
若是，则到步骤200：驱动靶标笔运动，并进行下一个注册点的注册；
若否，则执行步骤S107：在机械臂的引导下，由操作人员操作机械臂调整靶标笔的姿态，直到导航***200能够观察到靶标笔的位置。
图5为优选实施例中的自动运行模式的骨注册流程图。如图5所示，如果在上述步骤S01中，选择自动运行模式，则手术机器人10执行以下步骤：
步骤S201：导航***200向机械臂***100发送感兴趣的注册点的位置信息。
步骤S202：在接收到感兴趣的注册点的位置信息后，机械臂***100规划运动轨迹。
步骤S203：规划运动轨迹后，控制***控制机械臂110驱动靶标笔沿运动轨迹运动至感兴趣的注册点，也即靶标笔在机械臂的驱动下穿刺软骨直至到达骨头表面，并在骨头上选择注册点。
当靶标笔运动至感兴趣的注册点后，执行步骤S2041：控制***根据导航***200采集的靶标笔的当前位置信息，判断靶标笔是否运动到位：
若是，优选执行步骤S205：由操作人员介入并进一步判断是否继续穿刺到骨表面；
若否，继续执行步骤S203。
优选在执行步骤S2041的同时，还执行步骤S2042：控制***根据机械臂末端所受到的作用力的信息，判断穿刺软骨力是否达到标定值：
若是，则可直接确认靶标笔已运动到感兴趣的注册点，即获得步骤S206的结果；
若否，则继续执行步骤S203。
进一步地，在执行步骤S205时，若操作人员认为要继续穿刺到骨表面，则优选执行步骤S2042；反之，若操作人员认为不需要继续穿刺到骨表面，则可获得步骤S206的结果。
进一步，在获得步骤S206的结果后，机械臂***100将完成信号发送给导航***200，导航***200根据完成信号采集靶标笔的位置信息，并执行步骤S207：判断导航***200是否能够正常采集靶标笔的位置信息：
若是，则到步骤S200：驱动靶标笔运动，并进行下一个注册点的注册；
若否，则执行步骤S208：控制***控制机械臂运动以调整靶标笔的姿态，直到导航***200能够观察到靶标笔的位置。
具体来说，当靶标笔的位置不在导航***200的视野范围内，则机械臂110驱动靶标笔进行空间位置不变的旋转运动，即靶标笔的空间位置不变而仅做姿态的改变，此时，机械臂的一部分转动关节保持不动，另一部分转动关节驱动靶标笔围绕一主动不动点转动，直至导航***200可以正常采集到靶标笔的位置信息后机械臂停止运动，这样做，便于顺利地进行一个注册点的骨注册。而在辅助运行模式下，机械臂可引导操作人员握持靶标笔围绕主动不动点进行旋转，调整靶标笔的姿态即可。
如图6a和图6b所示，所述导航***200包括摄像头/光学感应***210和光学靶标220，摄像头210例如是双目摄像头，一个光学靶标220安装在靶标笔上，另有一个光学靶标220安装在患者骨头上并固定不动。例如图6a所 示，当靶标笔上的光学靶标220不在摄像头210的视野范围内时，可通过变换靶标笔的姿态，使光学靶标220围绕主动不动点旋转而进入摄像头210的视野范围内，即如图6b所示。
本实施例中，所述位置传感装置包括两个编码器(或称双编码器)，分别为绝对式编码器301和增量式编码器302。以一个转动关节的力矩检测作为示意。如图7和图8所示，在转动关节处，设置有驱动器303和减速器304，所述减速器304的输出端与转动关节相连接，所述减速器304的输入端与驱动器303相连接，所述驱动器303通过减速器304驱动转动关节转动。驱动器303包括驱动电机，在驱动电机的输出端安装了绝对式编码器301和增量式编码器302中的一个，另一个编码器则安装在减速器304的输出端(如转动关节输出轴)。绝对式编码器301实时采集转动关节的绝对状态参数T1，增量式编码器302实时采集转动关节的相对状态参数T2，绝对状态参数T1和相对状态参数T2均发送给控制***。
当转动关节受到外力作用时，根据力矩平衡原理，驱动电机的输出力矩τ应当与转动关节受到的外力力矩τ'大小相等，而驱动电机的输出力矩τ与转动关节的变形量有关：
τ＝Kδ
其中：K为减速器304的刚度模型，如谐波减速器的刚度模型；δ为转动关节的变形量。
本实施例中，转动关节的变形量与绝对状态参数T1和相对状态参数T2存在对应关系，可事先通过试验标定该对应关系。或者，通过如下公式计算得到：
Figure PCTCN2020095666-appb-000002
其中：T1为绝对式编码器所反馈回的转动关节的绝对位置，绝对位置例 如是绝对转动角度；T2为增量式编码器所反馈回的转动关节的相对位置，相对位置例如是相对转动角度；c1为绝对式编码器的分辨率；c2为增量式编码器的分辨率；S0为增量式编码器的启动初始化位置。
因此，控制***可以根据转动关节的变形量和减速器的刚度模型，获取驱动电机的输出力矩τ，便可根据驱动电机的输出力矩确定转动关节受到的外力力矩τ'。
在替代性实施例中，如图8所示，两个编码器分别安装在所述减速器304的输入端和减速器的输出端。理论上，减速器以及其他传动元件可以等效为一个弹性体306。当转动关节受外力作用时，两个编码器中间的等效弹性体306会发生形变，导致两个编码器互相之间所采集的数据产生偏差，且这一偏差与外力有关。因此，可通过试验标定，便可以得到两个编码器的读数偏差与外力之间的映射关系。进而利用驱动器获取两个编码器的读数，再将读数上传给控制***，控制***根据读数获取两个编码器的偏差值，并根据偏差值即可得到关节所受到的外力。
为了便于描述，本文中，将驱动器、减速器、转动关节和第一力识别装置一起统称为关节模组，关节模组除包括位置传感装置以外，还可包括至少一个力矩传感器305，力矩传感器305可以为光学机械变形类型、电磁感应类型或电阻应变类型等。如图7所示，通过力矩传感器305直接检测转动关节受到的力矩信息。这里，位置传感装置和力矩传感器可以相互校验，以对两者的工作状态进行监控，从而进一步提高转动关节力矩识别的准确性。其中力矩传感器305可以安装在减速器的输出端，如关节输出轴上。进一步地，还可以采集驱动电机的输出电流，使控制***根据驱动电机的输出电流获取转动关节受到的外力信息，以实现冗余检测。
进一步的，所述力矩传感器305为压力传感器。出厂前，可先对压力传 感器进行零点标定，具体的，将外力作用在机械臂的转动关节上，并读取压力传感器的输出数据，从而可根据所测输出数据对压力传感器进行零点标定。进而使用时，当机械臂转动关节受到外力作用时，所述压力传感器的形变量转化为电信号传送给控制***，控制***根据压力传感器的电信号，即可计算转动关节受到的外力。
在替代性实施例中，所述力矩传感器305还可以为分布式光栅传感器。分布式光栅传感器设置在转动关节处，可与机械臂外壳连接。光栅传感器包括光纤光栅应变计，构成应变桥，布桥方式可参考现有电阻应变桥的布桥方式。并且在各个转动关节处设置光纤光栅应变计，且各个关节模组的光纤光栅应变计共用一路光纤通道。更具体的，由放置在机械臂后端(如底座)的宽带光源发射调制光，经过各关节模组的光纤光栅应变计后，被反射回来，控制***进行解调之后解算出反射光的波长。如果转动关节受到外力作用，则反射光的波长将发生变化，波长变化量与分布式光栅传感器的应变量成正比，由此可测出各个分布式光栅传感器的应变量，机械臂控制***再根据分布式光栅传感器的应变量计算得到关节模组所受的外力。
应知晓，力识别***除了可以识别转动关节力矩和/或机械末端力矩以外，还可以进一步校验力矩信息的准确性。也即，力识别***优选包括第一力识别装置和第二力识别装置，其中第一力识别装置可以包括多种第一传感器，以此通过不同的方式检测力矩信息，例如机械臂末端的第二力识别装置通过六轴传感器直接感测机械臂末端的力矩信息，而第一力识别装置可通过双编码器、压力传感器、光栅传感器中的一种或多种感测转动关节所受到的力矩信息，第一力识别装置还可通过检测驱动电机的输出电流的方式获取关节所受到的力矩信息。因此，可以在骨注册过程中通过多种检测方式获取机械臂末端的受力信息，且对这些检测信息进行校验，校准计算使用的力矩信息是 否准确。此外，通过检测转动关节所受到的力矩信息，还可以起到对关节模组保护的作用，例如机械臂具有主动调整模式和/或被动调整模式，通常地，由机械臂110上的动力机构如驱动电机，驱动机械臂运动的工作方式即为机械臂的主动调整，与此相对的，由笛卡尔作用力驱动机械臂运动的工作方式即为机械臂的被动调整。
在主动调整模式下，所述控制***在获得机械臂上各个转动关节的力矩后，进而控制所有力矩大于预设的标定值的转动关节上的驱动电机(即驱动力矩大于预设的标定值的关节的驱动电机，下同)停止输出。所述标定值例如为零。这样可以避免在机械臂的主动调整模式操作中因发生人为误操作导致的机械位姿的调整的问题，由此增加机械臂使用过程中的安全性和可靠性。那么，在被动调整模式下，所述控制***在获得机械臂上各个关节的力矩后，进而控制所有力矩大于预设的标定值(所述标定值例如为零)的转动关节上的驱动电机进行输出；以此，使得机械臂上力矩大于预设的标定值的转动关节上的驱动电机能够配合操作人员驱动机械臂转动关节运动，以使得操作人员能够克服机械臂上转动关节的阻力，从而所述控制***能够协助操作人员使机械臂顺利运动至预定的位置，这样的方式，增加了机械臂被动调整操作的舒适度和便捷性。显然，自在机械臂上施加外部力的开始至结束，所述机械臂上力矩大于预设的标定值(所述标定值例如为零)的转动关节上的驱动电机均输出动力，从而使得操作人员在被动调整机械臂时更加省力和简便。
以在机械臂末端转动关节设置六轴传感器，以及在各个关节模组处设置双编码器作为示例，各力识别装置将获得的力矩信息传递给控制***，进而控制***比较根据双编码器计算得到的机械臂末端所受到的作用力，与六轴传感器直接检测到的机械臂末端所受到的作用力，也即，控制***对接收到的力矩信息进行校验，根据校验结果确定是否更新校验参数，同时根据校验 结果，控制***还可作出相应的安全保护措施。例如以六轴传感器作为校验基准，更新双编码器中的减速器的刚度系数K，或者更新压力传感器以及光栅传感器中的相关参数。
进一步，控制***还用于根据接收到的力矩信息，判断力矩传感器是否超过力矩传感器的量程，对力矩传感器进行保护，进一步保证骨注册过程中关节力矩信息的准确性，从而进一步判断穿刺软骨力信息的准确性，保证骨注册的准确性。
图9为优选实施例中的骨注册的流程图。如图9所示，骨注册过程中，在执行到步骤S301靶标笔运动至感兴趣的注册点的过程中，还执行步骤S302：控制***判断各个关节的力矩传感器是否工作正常；若否，则停止自动运行模式或辅助运行模式，改为手动注册S303，由操作人员手动操作驱动靶标笔进行骨注册；若是，则执行步骤S304：控制***继续控制所述机械臂驱动所述靶标笔向所述注册点运动，并判断穿刺软骨力是否达到标定值，也即六轴传感器反馈回的力矩值是否达到标定值，如果达到了标定值，则表明靶标笔已经穿过软骨到达骨表面并正确地选中了骨头上的注册点，即可进行下一个注册点的骨注册S305，如果未达到标定值，则执行步骤S306：继续驱动靶标笔穿透软骨以及其他组织，此过程中，随时返回执行步骤S304，直到穿刺软骨力达到标定值。
本实施例中，所述控制***中可设置两个经验阈值，第一个阈值用于传感器相互校验的阈值，第二个阈值用来判断力矩传感器是否超出其量程。使控制***根据该两个阈值判断各个关节模组上的传感器是否工作正常。具体来说，控制***将根据双编码器的检测模型所测量得到的机械臂末端所受到的作用力值，同六轴传感器所测量得到的机械臂末端所受到的作用力值进行比较，若比较后两者的偏差超过第一个阈值所限定的范围，则表明传感器出 现了异常，需要更新传感器的相关参数，若两者的偏差在第一个阈值的范围内，则表明传感器工作正常，后续可以使用力识别***所检测的数据控制靶标笔***软骨的深度。本发明对于两个阈值的具体大小不作限定，可根据传感器的性能、检测精度等要求通过实验确定。
为了进一步确保力识别***在出厂后实际使用过程中的可靠性，本实施例的手术机器人10还提供了术前对力识别***进行维护的功能。
图10为优选实施例中的对力识别***进行维护的流程图。如图10所示，在步骤S401中，出厂之前，可记录各个机械臂转动关节受外力时编码器的偏差并存储在控制***中。在步骤S402中，出厂之前，可启动校验程序，校验编码器的可靠性。校验之后，在步骤S403中，自动运行手术机器人，并更新校验参数，确保编码器工作正常。相应的，在出厂之前，在步骤S404中，也需要标定其他力矩传感器采集的数据信息，并确保采集的数据信息的准确性，也即根据六轴传感器来更新关节模组上各传感器的参数，而确保这些传感器的准确性。进而，出厂之后，在实际使用过程中，术前可执行步骤S405，将各个传感器采集的数据与出厂数据进行对比，以确保减速器的刚度。
需说明的是，本发明主要针对手术机器人中的主要机械部件，即谐波减速器进行检查，确保谐波减速器的运行状态正常。在进行刚度确认时，数据库将记录各个机器人产品不同的刚度的经验值范围，根据产品使用的时间以及经验值范围，以及原始的机械和硬件的设计参数，确认在力矩传感器使用一定时间后是否还能够正常工作。
进一步，可在控制***中设置第三个阈值，若力矩传感器的检测结果超过第三个阈值，则认为力矩传感器无法达到所需精度，需对力矩传感器进行重新维护标定或更换力矩传感器。第三个阈值根据各个机器人产品不同的刚度的经验值范围、产品使用的时间，以及原始的机械和硬件的设计参数，例 如，双编码器的分辨率、准确度，力矩传感器及六轴力矩传感器的分辨率，零漂，温漂，谐波减速器的刚度，减速比，减速比误差，以及连接件的刚度模型等确定。
通过以上确认，可以确保谐波减速器与双编码器组成的力矩传感器模型计算得到的力矩是可靠的，能够用在前述提到的保护程序与校验程序中。
在一些实施例中，手术机器人10的控制***采用神经网络计算方法进行传感器模型参数的更新，具体根据六轴传感器，增加在线神经网络学习，更新传感器模型参数。该神经网络计算方法可以提供在线模型参数的更新，保证各个转动关节的受力情况更加可靠，使力识别***的可靠性得到进一步的提升。利用六轴传感器进行在线模型参数更新的流程图如图11所示。
如图11所示，在神经网络学习中，需要在流程S501中事先采集六轴传感器反馈回的力矩数据，与此同时，还需要在流程S502以及S503中分别获取编码器的数据和力矩传感器的数据，进而在流程S504中根据编码器的数据获得关节力矩，并在流程S505中根据力矩传感器模型和零点标定得到力矩信息以实现关节力矩的相互校验，随后，在流程S506中根据各个关节的力矩传感器数据，计算得到机械臂末端的笛卡尔作用力，并在流程507中将该机械臂末端的笛卡尔作用力做为神经网络输出数据。更具体的，在流程S508中，以各个关节的力矩传感器模型和减速器刚度模型的系数作为神经网络输入参数，并在流程S509中，以输入参数进行在线神经网络学习，更新已知模型的参数，包括减速器刚度模型，关节力矩传感器模型，并在流程S510中根据学习后的结果计算得到新的机械臂末端的笛卡尔作用力，进而在流程S511中将神经网络的输出结果(计算得到的新的机械臂末端的笛卡尔作用力)与实际六轴传感器采集的数据(实际的机械臂末端的笛卡尔作用力)进行比较，以确认是否更新传感器的模型参数。
图12为优选实施例提供的基于神经网络模型获得力矩传感器模型的过程示意。如图12所示，神经网络具有输入层、隐藏层和输出层共三层结构。输入层具有三个节点，分别是力矩传感器参数1、2、3；隐藏层具有两个节点；而输出层则具有一个节点，用于输出单个关节力矩。其中，如果是双编码器，则输入层的三个节点的输入分别对应于谐波减速器的刚度、关节的角速度和关节的角位置；如果是力矩传感器，则输入参数分别为力矩转换系数、力矩传感器零点和力矩传感器温度。而输出层的节点输出为力矩传感器的实际力矩。
本实施例中，采用在线训练后的神经网络模型，即在出厂时首先通过训练样本获得所有节点的权重以及阈值。在使用手术机器人的过程中，不断的更新神经网络中的模型参数，具体的训练方法可以采用梯度下降法，也可以采用其他的现有技术，本发明对此不作限定。
除图12中提到的神经网络模型，本实施例同样提供一种更贴近实际物理意义的模型更新方法，流程可参考图13。
如图13所示，神经网络具有输入层、隐藏层和输出层共三层结构。且输入层具有三个节点，隐藏层具有两个节点，而输出层则具有一个节点。其中，输入层的三个节点的输入对应关节的角速度、关节的角位置和关节电机的输出力矩。输出层的节点输出为力矩传感器的实际力矩。当更新往模型后，根据得到的力矩结果，利用最小二乘法，更新双编码器力矩模型的参数系数，以及力矩传感器的参数系数，将更新后得到的系数用作计算确认。
进一步的，本实施例的控制***包括存储器和处理器，存储器上存储有程序，以使处理器通过运行所述程序完成以上的功能。
综上所述，根据本发明实施例提供的技术方案，手术机器人利用力识别***，实现了手术机器人自动完成或辅助完成骨注册的目的，较大地提高了 骨注册的效率，同时也提升了骨注册的精度以及成功率。进一步地，利用多种力识别装置保证了手术过程中的穿刺软骨力的准确性，提高了手术精度。更进一步地，手术机器人还提供出厂后的维护方法，能有效保护手术机器人的核心部件，并对机器人的运行状态及部件状态做确认，确保手术过程中不会出现问题。再进一步地，针对六轴传感器，手术机器人还提供一种在线神经网络学习更新模型参数的方法，该方法可以提供在线模型参数的更新，保证各个关节的受力情况更加可靠，使力识别***的可靠性得到进一步的提升。
尤其地，手术机器人还具有在线校验功能，能对出厂使用后的机器做校验，确保手术前机器人的状态完好，保证手术过程的顺利进行。
上述描述仅是对本发明较佳实施例的描述，并非对本发明范围的任何限定，本发明领域的普通技术人员根据上述揭示内容做的任何变更、修饰，均属于权利要求书的保护范围。

Claims (25)

  1. 一种用于手术机器人的骨注册方法，其特征在于，包括以下步骤：
    机械臂***接收导航***所选择的预定对象上感兴趣的注册点的信息；
    控制机械臂***中的机械臂的运动，以使所述机械臂控制跟踪装置运动至所述预定对象上感兴趣的注册点；
    获取由力识别***检测的机械臂末端所受到的作用力；以及
    根据所述机械臂末端所受到的作用力，判断所述跟踪装置是否已运动到所述预定对象上感兴趣的注册点。
  2. 根据权利要求1所述的方法，其特征在于，，还包括以下步骤：
    获取由导航***采集的所述跟踪装置的当前位置信息；
    根据所述跟踪装置的当前位置信息，判断所述跟踪装置是否已运动到预定对象上感兴趣的注册点。
  3. 根据权利要求2所述的方法，其特征在于，根据所述机械臂末端所受到的作用力，以判断所述跟踪装置是否已运动到预定对象上感兴趣的注册点的优先级高于根据所述跟踪装置的当前位置信息，以判断所述跟踪装置是否已运动到预定对象上感兴趣的注册点的优先级。
  4. 根据权利要求1所述的方法，其特征在于，所述根据所述机械臂末端所受到的作用力，判断所述跟踪装置是否已运动到预定对象上感兴趣的注册点的步骤包括：
    判断力识别***是否工作正常；
    若是，则控制所述机械臂驱动所述跟踪装置向所述注册点运动。
  5. 根据权利要求4所述的方法，其特征在于，若所述力识别***判断为 工作正常，则在控制所述机械臂驱动所述跟踪装置向所述注册点运动的步骤之后，还包括：
    将所述机械臂的末端所受到的作用力与标定值进行比较；
    若等于或大于所述标定值，则确定所述跟踪装置已运动到预定对象上感兴趣的注册点；
    若小于所述标定值，则继续控制所述机械臂驱动所述跟踪装置向所述注册点运动，直至所述机械臂末端所受到的作用力达到所述标定值。
  6. 根据权利要求1所述的方法，其特征在于，在确定所述跟踪装置已运动到预定对象上感兴趣的注册点之后，还包括以下步骤：
    判断所述导航***是否能够正常采集所述跟踪装置的位置信息；
    若是，控制所述机械臂以驱动所述跟踪装置向下一个感兴趣的注册点运动；
    若否，控制所述机械臂以调整所述跟踪装置的姿态，直至所述导航***能够正常采集所述跟踪装置的位置信息。
  7. 根据权利要求1所述的方法，其特征在于，在控制机械臂运动之前，还包括以下步骤：
    根据所述导航***所选择的预定对象上感兴趣的注册点，所述机械臂***规划一运动轨迹，所述机械臂驱动所述跟踪装置沿所述运动轨迹运动至感兴趣的注册点。
  8. 根据权利要求1所述的方法，其特征在于，在判断所述跟踪装置是否已运动到预定对象上感兴趣的注册点的步骤之前，还包括以下步骤：
    根据所述力识别***获取的每个机械臂关节所受到的作用力，计算得到所述机械臂的末端所受到的作用力；
    获取所述力识别***所得到的所述机械臂末端所受到的作用力；
    比较计算得到的所述机械臂的末端所受到的作用力，与利用力识别***所得到的所述机械臂末端所受到的作用力；
    若两者的偏差超出预定范围，则更新所述力识别***。
  9. 根据权利要求8所述的方法，其特征在于，所述力识别***获取每个机械臂关节所受到的作用力的步骤包括：
    根据安装在每个所述机械臂关节的第一传感器，检测每个所述机械臂关节所受到的作用力。
  10. 根据权利要求8所述的方法，其特征在于，所述力识别***得到所述机械臂末端所受到的作用力的步骤包括：
    根据安装在所述机械臂末端的第二传感器，检测所述机械臂末端所受到的作用力。
  11. 根据权利要求8-10中任一项所述的方法，其特征在于，采用神经网络计算的方法计算所述机械臂末端所受到的作用力，且在输出所述机械臂末端所受到的作用力之前，还包括对所述力识别***进行神经网络训练的步骤，并保持神经网络训练后的网络结构以及权值文件。
  12. 一种手术机器人，其特征在于，包括控制***、力识别***、机械臂***和导航***；所述机械臂***包括机械臂；所述机械臂的机械臂末端用于连接跟踪装置；所述控制***分别与所述力识别***、所述机械臂***和所述导航***通信连接，所述导航***与所述机械臂***通信连接；
    所述导航***被配置为将预定对象上感兴趣的注册点的信息发送给所述机械臂***；所述机械臂***被配置为根据所述预定对象上感兴趣的注册点的信息，控制所述机械臂运动，以驱动所述跟踪装置运动至预定对象上感兴 趣的注册点；
    所述力识别***设置在所述机械臂上，所述力识别***用于检测所述机械臂末端所受到的作用力并发送给所述控制***；所述控制***被配置为根据所述机械臂末端所受到的作用力，判断所述跟踪装置是否已运动到预定对象上感兴趣的注册点。
  13. 根据权利要求12所述的手术机器人，其特征在于，所述导航***被配置为：采集所述跟踪装置的当前位置信息并发送给所述控制***；所述控制***被配置为根据所述跟踪装置的当前位置信息，判断所述跟踪装置是否已运动到预定对象上感兴趣的注册点。
  14. 根据权利要求13所述的手术机器人，其特征在于，所述控制***被配置为，根据所述机械臂末端所受到的作用力，以判断所述跟踪装置是否已运动到预定对象上感兴趣的注册点的优先级高于根据所述跟踪装置的当前位置信息，以判断所述跟踪装置是否已运动到预定对象上感兴趣的注册点的优先级。
  15. 根据权利要求12所述的手术机器人，其特征在于，所述控制***被配置为，判断力识别***是否工作正常；若是，则控制所述机械臂驱动所述跟踪装置向所述注册点运动。
  16. 根据权利要求15所述的手术机器人，其特征在于，若所述力识别***判断为工作正常，所述控制***被配置为将所述机械臂末端所受到的作用力与所述控制***内预存的标定值进行比较；若等于或大于所述标定值，则确定所述跟踪装置已运动到预定对象上感兴趣的注册点；若小于所述标定值，则继续控制所述机械臂驱动所述跟踪装置向所述注册点运动，直至所述机械臂末端所受到的作用力等于或大于所述标定值。
  17. 根据权利要求12所述的手术机器人，其特征在于，在确定所述跟踪装置已运动到预定对象上感兴趣的注册点之后，所述控制***还被配置为判断所述导航***是否能够正常采集所述跟踪装置的位置信息；若是，控制所述机械臂驱动所述跟踪装置向下一个感兴趣的注册点运动；若否，控制所述机械臂调整所述跟踪装置的姿态，直至所述导航***能够正常采集所述跟踪装置的位置信息。
  18. 根据权利要求12所述的手术机器人，其特征在于，所述机械臂包括多个机械臂关节，所述力识别***包括第一力识别装置和第二力识别装置；
    所述第一力识别装置包括多个第一传感器，每个所述机械臂关节处设置至少一个所述第一传感器；所述第一传感器用于检测对应的机械臂关节所受到的作用力；
    所述第二力识别装置包括至少一个设置在所述机械臂末端的第二传感器，所述第二传感器用于检测所述机械臂末端所受到的作用力；
    所述控制***被配置为：
    根据多个所述第一传感器检测到的所述机械臂关节所受到的作用力，计算得到所述机械臂末端所受到的作用力；比较计算得到的所述机械臂末端所受到的作用力，与所述第二传感器所检测得到的所述机械臂末端所受到的作用力；
    若两者的偏差超出预定范围，则更新所述力识别***。
  19. 根据权利要求18所述的手术机器人，其特征在于，所述第一传感器包括双编码器、力矩传感器以及分布式光栅传感器中的至少一种，所述第二传感器包括六轴传感器。
  20. 根据权利要求19所述的手术机器人，其特征在于，每个所述机械臂 关节处设置有双编码器，所述双编码器包括绝对式编码器和增量式编码器；所述机械臂还包括用于驱动所述机械臂关节的驱动器和减速器，所述驱动器的输出端连接所述减速器的输入端，所述减速器的输出端连接所述机械臂关节；其中，所述绝对式编码器和增量式编码器中的一个设置在所述减速器的输入端，另一个设置在所述减速器的输出端。
  21. 根据权利要求20所述的手术机器人，其特征在于，所述机械臂关节所受到的作用力τ的计算公式为：
    τ＝Kδ
    Figure PCTCN2020095666-appb-100001
    其中，K为所述减速器的刚度模型；δ为所述机械臂关节的变形量；T1为所述绝对式编码器所反馈回的所述机械臂关节的绝对位置；T2为所述增量式编码器所反馈回的所述机械臂关节的相对位置；c1为所述绝对式编码器的分辨率；c2为所述增量式编码器的分辨率；S0为所述增量式编码器的启动初始化位置。
  22. 根据权利要求20所述的手术机器人，其特征在于，所述双编码器与所述控制***通信连接；所述控制***用于根据所述双编码器的数据偏差与所述机械臂关节所受到的作用力的映射关系，获取所述机械臂关节所受到的作用力。
  23. 根据权利要12所述的手术机器人，其特征在于，所述控制***被配置为采用神经网络计算的方法计算所述机械臂末端所受到的作用力，且在输出所述机械臂末端所受到的作用力之前，对所述力识别***进行神经网络训练，并保持神经网络训练后的网络结构以及权值文件。
  24. 根据权利要12所述的手术机器人，其特征在于，所述手术机器人包括自动运行模式和辅助运行模式；
    当所述手术机器人处于所述自动运行模式时，所述机械臂***根据接收到的感兴趣的注册点的信息，规划运动轨迹，并控制所述机械臂驱动所述跟踪装置沿着所述运动轨迹运动至预定对象上感兴趣的注册点；
    当所述手术机器人处于所述辅助运行模式时，由外力驱动所述机械臂运动并带动所述跟踪装置运动至预定对象上感兴趣的注册点。
  25. 一种可读存储介质，其上存储有程序，其特征在于，当所述程序被处理器运行时执行如权利要求1-11中任一项所述的用于手术机器人的骨注册方法。
PCT/CN2020/095666 2019-12-26 2020-06-11 骨注册方法、手术机器人及可读存储介质 WO2021128755A1 (zh)

Priority Applications (3)

Application Number Priority Date Filing Date Title
EP20905068.1A EP4082470A4 (en) 2019-12-26 2020-06-11 BONE REGISTRATION PROCEDURE, SURGICAL ROBOT AND READABLE STORAGE MEDIA
JP2022539307A JP2023508452A (ja) 2019-12-26 2020-06-11 骨登録方法、手術用ロボットおよび読み取り可能記憶媒体
AU2020414777A AU2020414777B2 (en) 2019-12-26 2020-06-11 Bone registration method, surgical robot, and readable storage medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201911368137.1A CN111035454B (zh) 2019-12-26 2019-12-26 可读存储介质以及手术机器人
CN201911368137.1 2019-12-26

Publications (1)

Publication Number Publication Date
WO2021128755A1 true WO2021128755A1 (zh) 2021-07-01

Family

ID=70240632

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/095666 WO2021128755A1 (zh) 2019-12-26 2020-06-11 骨注册方法、手术机器人及可读存储介质

Country Status (6)

Country Link
US (1) US11690681B2 (zh)
EP (1) EP4082470A4 (zh)
JP (1) JP2023508452A (zh)
CN (1) CN111035454B (zh)
AU (1) AU2020414777B2 (zh)
WO (1) WO2021128755A1 (zh)

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018059838A1 (en) * 2016-09-27 2018-04-05 Brainlab Ag Efficient positioning of a mechatronic arm
US11497566B2 (en) * 2018-10-26 2022-11-15 Biosense Webster (Israel) Ltd. Loose mode for robot
CN111035454B (zh) * 2019-12-26 2021-09-10 苏州微创畅行机器人有限公司 可读存储介质以及手术机器人
CN111843978B (zh) * 2020-05-29 2022-07-08 成都博恩思医学机器人有限公司 一种器械控制方法
WO2023040897A1 (zh) * 2021-09-14 2023-03-23 武汉联影智融医疗科技有限公司 一种手术机器人的空间注册位姿计算方法和***
CN114677429B (zh) * 2022-05-27 2022-08-30 深圳广成创新技术有限公司 一种机械手的定位方法、装置、计算机设备和存储介质
DE102022119730A1 (de) * 2022-08-05 2024-02-08 Schaeffler Technologies AG & Co. KG Verfahren zur präzisen Bestimmung eines Ausgangsdrehmoments und kollaborativer Roboter
CN115068113A (zh) * 2022-08-22 2022-09-20 科弛医疗科技(北京)有限公司 一种主从式遥操作骨科机器人***
CN116747026B (zh) * 2023-06-05 2024-06-25 北京长木谷医疗科技股份有限公司 基于深度强化学习的机器人智能截骨方法、装置及设备

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104626168A (zh) * 2014-12-16 2015-05-20 苏州大学 基于智能算法的机器人力位柔顺控制方法
US20170277285A1 (en) * 2016-03-28 2017-09-28 Pixart Imaging Inc. Pen-typed navigation device and related navigation module
CN109443207A (zh) * 2018-11-19 2019-03-08 华中科技大学 一种光笔式机器人原位测量***与方法
CN109500837A (zh) * 2018-12-18 2019-03-22 上海岭先机器人科技股份有限公司 一种基于双编码器的机器人关节力矩测量方法
CN110103217A (zh) * 2019-05-09 2019-08-09 电子科技大学 工业机器人手眼标定方法
CN110394801A (zh) * 2019-08-06 2019-11-01 前元运立(北京)机器人智能科技有限公司 一种机器人的关节控制***
CN110547873A (zh) * 2019-09-26 2019-12-10 北京爱康宜诚医疗器材有限公司 注册笔和电子装置
CN111035454A (zh) * 2019-12-26 2020-04-21 苏州微创畅行机器人有限公司 可读存储介质以及手术机器人

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10136952B2 (en) * 2016-06-16 2018-11-27 Zimmer, Inc. Soft tissue balancing in articular surgery
WO2018059838A1 (en) * 2016-09-27 2018-04-05 Brainlab Ag Efficient positioning of a mechatronic arm
US10244926B2 (en) * 2016-12-28 2019-04-02 Auris Health, Inc. Detecting endolumenal buckling of flexible instruments
US11071594B2 (en) * 2017-03-16 2021-07-27 KB Medical SA Robotic navigation of robotic surgical systems
US11033341B2 (en) * 2017-05-10 2021-06-15 Mako Surgical Corp. Robotic spine surgery system and methods
US11311342B2 (en) * 2017-10-30 2022-04-26 Cilag Gmbh International Method for communicating with surgical instrument systems
KR102636163B1 (ko) * 2018-07-23 2024-02-13 우니베르지타이트 겐트 뼈 절삭기 및 연조직 보호기
US11229493B2 (en) * 2019-01-18 2022-01-25 Nuvasive, Inc. Motion programming of a robotic device

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104626168A (zh) * 2014-12-16 2015-05-20 苏州大学 基于智能算法的机器人力位柔顺控制方法
US20170277285A1 (en) * 2016-03-28 2017-09-28 Pixart Imaging Inc. Pen-typed navigation device and related navigation module
CN109443207A (zh) * 2018-11-19 2019-03-08 华中科技大学 一种光笔式机器人原位测量***与方法
CN109500837A (zh) * 2018-12-18 2019-03-22 上海岭先机器人科技股份有限公司 一种基于双编码器的机器人关节力矩测量方法
CN110103217A (zh) * 2019-05-09 2019-08-09 电子科技大学 工业机器人手眼标定方法
CN110394801A (zh) * 2019-08-06 2019-11-01 前元运立(北京)机器人智能科技有限公司 一种机器人的关节控制***
CN110547873A (zh) * 2019-09-26 2019-12-10 北京爱康宜诚医疗器材有限公司 注册笔和电子装置
CN111035454A (zh) * 2019-12-26 2020-04-21 苏州微创畅行机器人有限公司 可读存储介质以及手术机器人

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP4082470A4 *

Also Published As

Publication number Publication date
AU2020414777A1 (en) 2022-07-14
CN111035454B (zh) 2021-09-10
US20210196397A1 (en) 2021-07-01
AU2020414777B2 (en) 2023-12-14
EP4082470A4 (en) 2023-06-21
JP2023508452A (ja) 2023-03-02
EP4082470A1 (en) 2022-11-02
CN111035454A (zh) 2020-04-21
US11690681B2 (en) 2023-07-04

Similar Documents

Publication Publication Date Title
WO2021128755A1 (zh) 骨注册方法、手术机器人及可读存储介质
US20210038325A1 (en) Drilling control system and drilling control method
US20240065779A1 (en) Systems and methods for fault reaction mechanisms for medical robotic systems
JP7168320B2 (ja) 動作に対する制約を特徴付けるロボットおよびロボットを制御するための方法
CN110811832B (zh) 截骨校验方法、校验设备、可读存储介质及骨科手术***
CN107405180B (zh) 用于外科手术机器人***的交互式引导和操纵检测安排、以及相关联方法
US20120059378A1 (en) Efficient Sculpting System
CN110811833B (zh) 截骨校验方法、校验工具、可读存储介质及骨科手术***
CN114311031A (zh) 手术机器人主从端延时测试方法、***、存储介质和设备
US20210393349A1 (en) Systems and methods for device verification and sensor calibration
JP2023511272A (ja) ナビゲーション支援手術中にオフセットをモニタするシステム及び方法
CN115300110A (zh) 内窥镜手术控制***
WO2021098177A1 (zh) 截骨校验方法、校验设备、可读存储介质及骨科手术***
KR101358668B1 (ko) 다자유도 수술도구의 힘 또는 토크를 로봇팔의 슬라이더에서 측정하는 장치 및 방법
US20220409307A1 (en) Systems and methods for detecting skiving in surgical instruments
CN115177366A (zh) 医疗器械控制方法、医疗器械控制***和手术***

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20905068

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2022539307

Country of ref document: JP

Kind code of ref document: A

ENP Entry into the national phase

Ref document number: 2020414777

Country of ref document: AU

Date of ref document: 20200611

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2020905068

Country of ref document: EP

Effective date: 20220726