A data-driven handheld puncture robot system and its control method

Abstract

The present invention discloses a data-driven handheld puncture robot system and its control method, belonging to the technical field of auxiliary puncture robot control. This system includes a handheld puncture robot body, a control host computer module, and a navigation module; the handheld puncture robot body is used to perform puncture surgery under the operation or guidance of a doctor; the control host computer module is used to collect, monitor, and control the state information of the puncture robot and the patient, and provide an interaction interface; the navigation module is used to provide real-time surgical navigation information for the doctor, track the positional relationship between the puncture needle and the patient in real time, and automatically plan the optimal puncture path. The present invention dynamically adjusts parameters and learns the nonlinear characteristics of the system through a tuning algorithm, improving the puncture accuracy and stability of the handheld puncture robot in complex situations.

Description

Technical Field

[0001] The present invention belongs to the technical field of auxiliary puncture robot control, and particularly relates to a data-driven handheld puncture robot system and a control method thereof. Background Art

[0002] In the modern medical field, prostate cancer, as a disease that seriously threatens human health, its high incidence and high mortality rate bring a heavy burden to patients. If early prostate cancer can be detected and treated in time, the cure rate is relatively high, but late-stage prostate cancer often has a poor prognosis and a high mortality rate. Therefore, accurate early diagnosis has become the key to improving the survival rate of prostate cancer patients. As the "gold standard" for prostate cancer diagnosis, puncture biopsy currently mainly relies on doctors' manual operations to determine the puncture position and obtain tissue samples through the puncture needle reaching the target point, which has the advantages of less trauma and quick recovery. However, due to the different physical conditions of patients, the technical levels and surgical states of medical technicians are different, and deep small organ puncture surgeries similar to the prostate still face many challenges and risks.

[0003] Traditional puncture surgeries require doctors to hold a biopsy gun and perform punctures according to the imaging system and experience. However, due to the relatively long operation time, especially when performing deep small organ punctures multiple times, the puncture travel distance is long, and the operator's attention is highly concentrated, making it easy to get fatigued, thus resulting in a decrease in puncture accuracy.

[0004] Handheld puncture robots have obvious advantages in prostate cancer puncture biopsy, but there are also problems. Compared with traditional manual punctures, they rely on high-precision mechanical structures and advanced sensors to avoid operation errors caused by factors such as doctors' experience, fatigue, and emotions, achieving accurate positioning and stable operation, and improving puncture accuracy and repeatability. Their flexible and portable design facilitates doctors to adjust the puncture angle and depth in a narrow surgical space, increasing the operation freedom and improving the surgical success rate. However, the non-linear characteristics of human tissues, the time-varying system parameters caused by tissue density changes, and the disturbances generated by physiological activities such as breathing and heartbeat pose challenges to the control of handheld puncture robots. Traditional control algorithms are difficult to adapt to these complex situations, resulting in poor puncture accuracy and stability.

[0005] The tuning algorithm brings hope for solving the control problem of the handheld puncture robot. This algorithm can dynamically adjust control parameters according to the real-time operating state and historical data of the system, adapting to system changes and uncertainties. Compared with traditional fixed-parameter control algorithms, the tuning algorithm has prominent advantages in dealing with the complex situations faced by the handheld puncture robot. In the face of non-linear problems, it learns the non-linear characteristics of the system by monitoring input and output data and adjusts parameters to better track the puncture trajectory, while traditional algorithms are prone to control deviations. In dealing with the problem of time-varying system parameters, the tuning algorithm can adjust strategies in a timely manner according to parameter changes to ensure stable puncture performance, while traditional algorithms are difficult to handle. For the disturbances of physiological activities, the tuning algorithm has strong robustness and can quickly adjust control signals to suppress the influence of disturbances. At the same time, this algorithm can also better adapt to the characteristics of the repeated operation of the puncture robot and meet the safety control requirements under actual physical constraints, providing a new solution for improving the overall performance of the puncture robot and having important practical significance for promoting the development of prostate cancer puncture biopsy technology. Summary of the Invention

[0006] Aiming at the deficiencies of the prior art, the purpose of the present invention is to provide a data-driven handheld puncture robot system and its control method to improve the puncture accuracy and stability of the handheld puncture robot in complex situations.

[0007] The purpose of the present invention is achieved through the following technical solutions: A data-driven handheld puncture robot system, comprising:

[0008] The handheld puncture robot body, which is used to accurately perform puncture surgery under the operation or guidance of a doctor. Its design is light and flexible, facilitating the doctor to operate flexibly during the surgery, while ensuring the stability and safety of the surgery;

[0009] The control host computer module, which is used to collect, monitor and control the status information of the puncture robot and the patient, and provide an interactive interface for analyzing the condition and formulating a surgical strategy;

[0010] The navigation module, which is used to provide real-time surgical navigation information for the doctor. By constructing a three-dimensional anatomical structure model of the patient, it can track the positional relationship between the puncture needle and the patient in real time and automatically plan the best puncture path to help the doctor perform precise operations and avoid damaging important tissues and organs;

[0011] The handheld puncture robot body includes:

[0012] The human-computer interaction module, which is used to provide an interactive interface between the user and the handheld puncture robot body, including forward and backward, up and down, left and right movement instructions, power switch, and the automatic trigger function of the biopsy gun;

[0013] The execution module is obtained by integrating electrical components and structural components; the electrical components include three DC motors, a linear push rod, a linear guide rail, and drive control components for actuating the biopsy gun; the structural components include a sensor array for capturing the motion parameters of each module.

[0014] The control module is used to process signals and integrates some drive circuits to reduce the volume of the device.

[0015] The system support module is used to provide the basic guarantee for the puncture robot, and the basic guarantee includes power management, circuit protection, status monitoring, and device disinfection to ensure the stable operation and hygienic safety of the handheld puncture robot body.

[0016] Furthermore, the control module includes a communication interface, which is connected to the host computer and is used to monitor and analyze the overall operating state of the device.

[0017] Furthermore, the control host computer module includes:

[0018] The communication interface module is used to realize the information flow between the host computer, the robot system, and the navigation component to achieve signal interaction at the physical level.

[0019] The core processing unit is used to provide tools for controlling the robot and the navigation component, including adjusting device parameters, formulating treatment plans, providing an interactive control approach, and presenting the working conditions of each subsystem of the puncture robot.

[0020] The industrial control computer is used to execute the core algorithm for automatic control of the puncture robot, collect data of each module through the communication link, and build a bridge for the interaction between the control host computer module and the hardware level.

[0021] Furthermore, the navigation module includes:

[0022] The sensing component integrates an electromagnetic positioning subsystem and an ultrasonic detection subsystem. By fusing the information of the electromagnetic positioning subsystem and the ultrasonic detection subsystem, the positioning accuracy of the system is improved. It is used to collect patient-related data and realize the description of the patient's puncture environment and the state of the puncture instrument.

[0023] The graphics processing workstation is used to receive, process, and forward the patient-related data collected by the sensing component, execute the fusion algorithm to refine the target information, and upload the processing result to the control host computer module; it provides a real-time positioning visualization interface for the user to facilitate the user to adjust the treatment plan according to actual needs.

[0024] The communication interface module is used to communicate with other modules and transmit the status of the patient and the device during the operation.

[0025] The present invention also provides a control method for a data-driven handheld puncture robot system, including the following steps:

[0026] (1) Define the parameters of the handheld puncture robot system. Let the time-discrete sampling instants of the system be k = 0, 1, 2, …;

[0027] Input vector: Define the three-input vector u(k) = [u1(k), u2(k), u3(k)] T , where u1(k) and u2(k) respectively correspond to the control inputs of two rotating joints, affecting the joint angles θ1 and θ2, and u3(k) represents the puncture needle extension length l;

[0028] Output vector: The three-output vector y(k) = [y1(k), y2(k), y3(k)] T , representing the coordinates of the robot end in three-dimensional space, y1(k) = x, y2(k) = y, y3(k) = z;

[0029] Reference input vector: r(k) = [r1(k), r2(k), r3(k)] T , representing the desired coordinates of the robot end;

[0030] Error vector: e(k) = r(k) - y(k) = [e1(k), e2(k), e3(k)] T ;

[0031] Controller parameter vector: ρ = [ρ1, ρ2, …, ρ n T , and the input-output relationship of the controller is u(k) = C(ρ)e(k), where C(ρ) represents the parameterized controller structure;

[0032] (2) Define relevant functions. Cross-correlation function: Define the cross-correlation function matrix R er (τ) of the output error vector e(k) and the reference input vector r(k), where τ is the time delay and N represents the total length of the time series;

[0033]

[0034] Use finite-time estimation:

[0035]

[0036] where M is the number of sampled data; is a 3×3 matrix, and its elements are

[0037] (3) Construct the objective function J(ρ):​

[0038] Among them, D represents the maximum time delay, and tr(·) represents the trace operation of a matrix;

[0039] (4) Update the controller parameters using the gradient descent method to minimize the objective function J(ρ);

[0040] Calculate the gradient of the objective function with respect to the controller parameter ρ According to the chain rule, we get:

[0041]

[0042] Among them, ρ m represents the m-th component of the controller parameter, represents the partial derivative of the cross-correlation function with respect to the controller parameter;

[0043] The update formula for the controller parameter is:

[0044] Among them, α represents the learning rate, which is used to control the step size of parameter update;

[0045] (5) Assume that the output of the system is y(k) = F(u(k), y(k - 1),...), where F is the dynamic function of the system; According to the chain rule, we get:

[0046]

[0047] By continuously updating the controller parameter ρ, the objective function J(ρ) is gradually reduced, and the correlation between the output error and the reference input is decreased to achieve accurate tracking of the end - effector coordinates of the robot with respect to the reference coordinates.

[0048] The present invention also provides an electronic device, including a memory and a processor, the memory is coupled to the processor; wherein, the memory is used to store program data, and the processor is used to execute the program data to implement the control method of the data - driven handheld puncture robot system as described above.

[0049] The present invention also provides a computer - readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the control method of the data - driven handheld puncture robot system as described above.

[0050] The beneficial effects of the present invention are as follows: The tuning algorithm dynamically adjusts parameters, improving the ability of the handheld puncture robot to adapt to unknown complex situations; By learning the nonlinear characteristics of the system, the tuning algorithm achieves more accurate puncture trajectory tracking; The tuning algorithm quickly responds to disturbances, shortening the time to recover stability and improving the puncture efficiency and accuracy. Description of the Drawings

[0051] Figure 1 Schematic diagram of the handheld puncture robot system of the present invention;

[0052] Figure 2 Composition diagram of the handheld puncture robot body of the present invention;

[0053] Figure 3 Schematic diagram of the structure of the handheld puncture robot of the present invention; wherein, (a) is the detailed structure diagram of the handheld puncture robot, and (b) is the simplified structure diagram of the handheld puncture robot. Specific embodiments

[0054] In order to enable those skilled in the art to better understand the technical solutions in the embodiments of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art shall fall within the protection scope of the embodiments of the present application.

[0055] As Figure 1 shown, the embodiment of the present invention provides a data-driven handheld puncture robot system, which is divided into three major parts, namely the handheld puncture robot body, the control host computer module and the navigation module. This system can interact with the device through manual operation. In addition, after starting the advanced assisted puncture function, the navigation module of the control module can be controlled for automatic puncture.

[0056] The main structure of the handheld puncture robot body is composed of four core components, specifically as Figure 2 shown, and the components include a human-computer interaction module, an execution module, a control module and a system support module.

[0057] The human-machine interaction module adopts a streamlined and efficient combination of buttons, enhancing the intuitiveness and comprehensibility of operations. The button combinations cover forward and backward, up and down, left and right movement commands, power on / off, and the automated trigger function of the biopsy gun. The execution module is integrated from electrical components and structural components. The electrical components are designed according to the system architecture, covering three DC motors, a linear push rod, a linear guide rail, and the drive control elements for activating the biopsy gun. The communication mechanism adopts a bus architecture to reduce the complexity of the wire harness. At the same time, the execution components are equipped with a sensor array for capturing the motion parameters of each module, which not only facilitates the real-time monitoring of the end state but also provides data support for the fault diagnosis of the entire puncture device. The control module not only undertakes the task of signal processing but also integrates some drive circuits, which effectively reduces the overall volume of the device. In addition, this module is equipped with a communication interface with the upper computer, facilitating doctors and developers to monitor and analyze the overall operating state of the device. The system support module provides comprehensive basic guarantees for the puncture robot system, and its functions cover multiple aspects such as power management, circuit protection, status monitoring, and device disinfection, ensuring the stable operation, hygiene, and safety of the robot body.

[0058] The control upper computer module is responsible for the collection, monitoring, and control of the puncture robot and patient status information, and at the same time provides an interactive interface for doctors to facilitate their analysis of the condition and formulation of surgical strategies. This module is subdivided into three major components: a core processing unit (i.e., the host), an industrial control computer (IPC), and a communication interface module. The communication interface module is responsible for realizing the information flow between the upper computer, the robot system, and the navigation components, ensuring the realization of signal interaction at the physical level. The host, as a key node of human-machine interaction, provides necessary tools for doctors and technicians to control the robot and navigation components, allowing not only the adjustment of device parameters but also the support for medical technicians to formulate treatment plans, providing an interactive control path for doctors and being able to immediately display the working conditions of each subsystem of the puncture robot. The IPC focuses on executing the core algorithms of robot automatic control, collecting data from each module through the communication link, and building a bridge for the interaction between the upper computer and the hardware level. Compared with the upper computer, the IPC shows higher stability and has the ability to directly communicate with a variety of hardware devices, which is the key interface for realizing hardware interaction.

[0059] The navigation module is subdivided into three core components: a sensing component, a graphics processing workstation, and a communication interface module. The sensing component integrates an electromagnetic positioning subsystem and an ultrasonic detection subsystem. Although the ultrasonic module can obtain image information of internal human tissues in real time, there are certain limitations in the positioning accuracy of the puncture needle. By fusing the information of the ultrasonic module and the electromagnetic positioning module, the positioning accuracy of the entire system can be significantly improved, thereby achieving a more accurate description of the patient's puncture environment and the state of the puncture instrument. The graphics processing workstation is mainly responsible for receiving, processing, and forwarding patient-related data collected by the sensing component, executing a fusion algorithm to extract target information, and uploading the processing results to the host computer system. In addition, the workstation also provides a real-time positioning visualization interface for the physician, enabling the physician to flexibly adjust the treatment plan according to actual needs. The communication interface module mainly communicates with other modules of the system to transmit the status of the patient and the device during the operation.

[0060] The embodiment of the present invention also provides a control method for a handheld puncture robot system based on data driving, including the following steps:

[0061] Step 1: Model establishment

[0062] Next, the basic principle of the handheld puncture robot will be introduced and a system model will be constructed:

[0063] The handheld puncture robot targeted by the present invention is as shown in Figures (a) and (b) of Figure 3 . A and B are the planes where the two branch chains are located. Driving θ1, θ2, and l1 can make the RCM located at point O. Among them, e1, e2, and e3 respectively represent the direction vectors of the x, y, and z axes, and a i , b i , w i respectively represent the direction vectors of the connections between the connecting rods. Since the planes A and B coincide with the puncture needle axis, the relationship between the input θ1, θ2 and the output ψ x , ψ y can be established.

[0064] The planes A and B intersect at the z axis, and their corresponding normal vectors can be expressed as:

[0065]

[0066] Among them, s· = sin(·), c· = cos(·).

[0067] Determine the direction vector w3 of the puncture needle axis as:

[0068]

[0069] Determine n A , nB Relationship between and w3:

[0070] n A ×n B = kw3;

[0071] Obtain:

[0072]

[0073] where k = ||n A ×n B ||.

[0074] Determine the analytical solution ψ of the forward kinematics of the puncture robot y , let ψ x = θ1:

[0075]

[0076] where t· = tan(·).

[0077] Determine the analytical solution θ2 of the inverse kinematics of the puncture robot, let ψ x = θ1, then we can obtain:

[0078]

[0079] Determine the rod lengths l1, l2 of the puncture robot. According to the geometric relationship of the robot structure, we can obtain:

[0080]

[0081] where a, b, l3, β are the design parameters of the robot system. b i , w i is perpendicular to each plane normal vector, and we obtain:

[0082]

[0083] Combined with this model, it can be seen that the output of the system is the coordinate P of the needle tip tip (x, y, z), and the three inputs are θ1, θ2 and the puncture needle extension length where the needle tip coordinates are:

[0084]

[0085] Step 2: Controller design based on the tuning algorithm

[0086] 2.1. Symbol definition

[0087] Let the time-discrete sampling instants of the system be k = 0, 1, 2, ….

[0088] Input vector: Define the three-input vector \(u(k)=[u_1(k),u_2(k),u_3(k)]\) T , where \(u_1(k)\) and \(u_2(k)\) correspond to the control inputs \(\theta_1\) and \(\theta_2\) of two rotating joints respectively, and \(u_3(k)\) is the insertion length \(l\) of the puncture needle.

[0089] Output vector: The three-output vector \(y(k)=[y_1(k),y_2(k),y_3(k)]\) T , representing the coordinates of the robot end-effector in three-dimensional space, \(y_1(k)=x\), \(y_2(k)=y\), \(y_3(k)=z\).

[0090] Reference input vector: \(r(k)=[r_1(k),r_2(k),r_3(k)]\) T , which is the desired coordinates of the robot end-effector.

[0091] Error vector: \(e(k)=r(k)-y(k)=[e_1(k),e_2(k),e_3(k)]\) T .

[0092] Controller parameter vector: \(\rho=[\rho_1,\rho_2,\ldots,\rho\) n [[]] T , and the input-output relationship of the controller is \(u(k)=C(\rho)e(k)\), where \(C(\rho)\) is a parameterized controller structure.

[0093] 2.2. Definition of related functions

[0094] Cross-correlation function: Define the cross-correlation function matrix \(R\) er (\tau)\) of the output error vector \(e(k)\) and the reference input vector \(r(k)\), where \(\tau\) is the time delay and \(N\) represents the total length of the time series.

[0095]

[0096] In practical applications, use the finite-time estimation:

[0097]

[0098] where \(M\) is the number of sampled data. is a 3×3 matrix, and its elements are

[0099] 2.3. Objective function for related tuning

[0100] The goal of related tuning is to minimize the correlation between the output error and the reference input, and construct the objective function \(J(\rho)\) as follows:

[0101]

[0102] Where \(D\) is the maximum time delay, and \(tr(\cdot)\) represents the trace operation of a matrix.

[0103] 2.4. Controller Parameter Update Formula

[0104] To minimize the objective function \(J(\rho)\), the gradient descent method is used to update the controller parameters. First, calculate the gradient of the objective function with respect to the controller parameter \(\rho\)

[0105] According to the chain rule, we get:

[0106]

[0107] where \(\rho\) m represents the \(m\)-th component of the controller parameter, represents the partial derivative of the cross-correlation function with respect to the controller parameter.

[0108] The update formula for the controller parameter is:

[0109]

[0110] where \(\alpha\) is the learning rate, which is used to control the step size of parameter update.

[0111] 2.5. Handling of Partial Derivatives in Specific Calculations

[0112] To calculate Since the elements in are related to the error vector \(e(k)\), and \(e(k)=r(k)-y(k)\), and \(y(k)\) is related to \(u(k)=C(\rho)e(k)\) through the dynamics of the system. Assume that the output of the system can be expressed as \(y(k)=F(u(k),y(k - 1),\cdots)\), where \(F\) is the dynamics function of the system. Through the chain rule, we can get:

[0113]

[0114] The above formulas constitute the core content of controlling a three-input three-output handheld puncture robot using the correlation tuning algorithm. By continuously updating the controller parameter \(\rho\), the objective function \(J(\rho)\) gradually decreases, that is, the correlation between the output error and the reference input decreases, thereby achieving precise tracking of the end coordinates of the robot to the reference coordinates.

[0115] This embodiment of the present invention also provides an electronic device, including a memory and a processor, the memory is coupled to the processor; wherein, the memory is used to store program data, and the processor is used to execute the program data to implement the control method of the data-driven handheld puncture robot system.

[0116] An embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the control method of the data-driven handheld puncture robot system described above.

[0117] The computer-readable storage medium may be an internal storage unit of any device with data processing capabilities described in any of the foregoing embodiments, such as a hard disk or memory. The computer-readable storage medium may also be any device with data processing capabilities, such as a plug-in hard disk, a Smart Media Card (SMC), an SD card, a Flash Card, etc. equipped on the device. Further, the computer-readable storage medium may also include both an internal storage unit of any device with data processing capabilities and an external storage device. The computer-readable storage medium is used to store the computer program and other programs and data required by any device with data processing capabilities, and may also be used to temporarily store data that has been output or will be output.

[0118] After considering the specification and practicing the content disclosed herein, those skilled in the art will readily think of other implementation manners of the present application. The present application is intended to cover any variations, uses, or adaptations of the present application, and these variations, uses, or adaptations follow the general principles of the present application and include common general knowledge or conventional technical means in the technical field not disclosed in the present application. The specification and the embodiments are only regarded as exemplary.

[0119] It should be understood that the present application is not limited to the exact structure described above and shown in the drawings, and various modifications and changes can be made without departing from its scope.

Claims

1. An electronic device, comprising a memory and a processor, characterized in that, The memory is coupled to the processor; wherein, the memory is used to store program data, and the processor is used to execute the program data to implement a control method for a data-driven handheld puncture robot system, and the method includes the following steps: (1) Define the parameters of the handheld puncture robot system, and set the time-discrete sampling instants of the system as k = 0, 1, 2, …; Input vector: Define a three-input vector \(u(k)=[u_1(k), u_2(k), u_3(k)]\) T , where \(u_1(k)\) and \(u_2(k)\) respectively correspond to the control inputs of two rotating joints, affecting the joint angles \(\theta_1\) and \(\theta_2\), and \(u_3(k)\) represents the penetration needle extension length \(l\); Output vector: The three-output vector y(k) = [y1(k), y2(k), y3(k)] T , representing the coordinates of the robot end-effector in three-dimensional space, where y1(k) = x, y2(k) = y, and y3(k) = z; Reference input vector: r(k) = [r1(k), r2(k), r3(k)] T , representing the desired end - effector coordinates of the robot Error vector: e(k) = r(k) - y(k) = [e1(k), e2(k), e3(k)] T ; Controller parameter vector: ρ = [ρ1, ρ2, …, ρ n T , the input-output relationship of the controller is u(k) = C(ρ)e(k), where C(ρ) represents the parameterized controller structure;​ (2) Define relevant functions; Cross-correlation function: Define the cross-correlation function matrix \(R_{e,r}(\tau)\) of the output error vector \(e(k)\) and the reference input vector \(r(k)\), where \(\tau\) is the time delay and \(N\) represents the total length of the time series; er (\(\tau\), where \(\tau\) is the time delay and \(N\) represents the total length of the time series; Use finite-time estimation: where M is the number of sampled data; is a 3×3 matrix, and its elements are (3) Construct the objective function J(ρ): where D represents the maximum time delay, and tr(·) represents the trace operation of a matrix; (4) Update the controller parameters by using the gradient descent method to minimize the objective function J(ρ); Calculate the gradient of the objective function with respect to the controller parameter ρ Obtained according to the chain rule: where ρ m represents the m-th component of the controller parameter, represents the partial derivative of the cross-correlation function with respect to the controller parameter; The update formula for the controller parameters is as follows: where α represents the learning rate, which is used to control the step size of parameter update; (5) Assume that the output of the system is y(k) = F(u(k), y(k - 1), …), where F is the dynamic function of the system; according to the chain rule, we get: By continuously updating the controller parameter ρ, the objective function J(ρ) is gradually reduced to achieve the tracking of the end-effector coordinates to the reference coordinates.

Images