Puncture robot tissue breaking state identification method based on human body parameter estimation

Abstract

The invention relates to a puncture robot breakthrough tissue state identification method based on human body parameter estimation, and the method comprises the following steps: S1, constructing a mathematical model of human body tissues of a far-end patient, and the mathematical model comprises a coupling relation between multiple layers of tissues and dynamic characteristics of the tissues; s2, according to the mathematical model constructed in the step S1, needle point data in the puncture process are collected in real time through a sensor, online parameter identification is conducted on model parameters, and the model parameters are updated in real time to reflect actual changes of tissue in the puncture process; s3, the mutation moment of the parameters in the puncture process is detected according to the model parameters updated in real time in the step S2, and the mutation moment is the moment when the deviation between the model parameters at a certain moment and a historical mean value exceeds a set threshold value, and it is judged that the moment is the moment when the needle tip breaks through the tissue layer. Compared with the prior art, the puncture robot has the advantages that the accuracy and safety of the puncture robot in the operation process are improved, and the like.

Description

Technical Field

[0001] The present invention relates to a master-slave remote control operation surgical system, and in particular to a method for identifying a tissue breakthrough state of a puncture robot based on human body parameter estimation. Background Art

[0002] With the development of robotics, puncture robots are increasingly being used in clinical medicine. However, due to the complex structure of human tissue and the significant differences in inter-layer properties, puncture robots struggle to accurately perceive the changes in the needle tip's state as it passes through different tissue layers, potentially leading to insufficient operational precision and safety. Existing technologies, most of which rely on single feedback signals such as force signals, struggle to respond to dynamic changes in human tissue in real time. Therefore, a real-time identification method based on tissue parameter estimation is needed to overcome the limitations of traditional puncture state recognition.

[0003] In practical applications, the literature [“Li Y, Li H. Penetration Event Identification Basedon Neural Network for Needle Tip Location in Robot Assisted Lumbar Puncture Surgery[J]. IEEE Robotics and Automation Letters, 2023, 9(1): 343-350.”] studied a neural network-based lumbar puncture event detection method to predict puncture events from real-time needle tip force signals. However, the disadvantage is that the neural network model needs to be retrained or calibrated according to the tissue characteristics of different patients. The literature [“Alamdar A, Patel N, Urias M, et al. Force and velocity based puncture detection in robotassisted retinal vein cannulation: in-vivo study[J]. IEEE Transactions on Biomedical Engineering, 2021, 69(3): 1123-1132.”] explored a force and velocity-based puncture detection method for retinal vein cannulation (RVC) surgery, aiming to identify the puncture moment and provide real-time feedback to avoid double vein puncture and retinal damage. The detection method combines force, velocity, and correlation analysis. The interaction force between the needle tip and tissue increases as the needle tip gradually penetrates the target tissue until the tissue elasticity reaches its limit, at which point the force suddenly drops, signaling puncture. The linear relationship between the needle-tissue interaction force and the tool-sclera interaction force is used to calculate a correlation coefficient to aid in the determination. However, the correlation metric introduces detection delay, making it unsuitable for high-speed puncture operations and potentially leading to time delays in actual applications.The paper [“Zhang E, Li J, Wang C, et al. A Safe Intervention Method for Robot-Assisted Ultrasonic Puncture Based on Interactive Force [C] / / 2024 IEEE 14th International Conference on CYBER Technology in Automation, Control, and Intelligent Systems (CYBER). IEEE, 2024: 376-381.)] proposes a method for robot-assisted ultrasonic puncture state identification based on interactive force. This method aims to achieve safe intervention by analyzing force changes during the puncture process, rather than relying on a preset force threshold. The puncture robot collects real-time force data from the needle tip and analyzes the changing characteristics of the force curve. When the puncture force suddenly drops, it determines that the needle tip has penetrated the target tissue. This method improves sensitivity through sliding average filtering and differentiation, but it carries the risk of false positives or negatives. Several of the methods involved all rely on a single source of force signals. However, due to the elasticity and structural complexity of tissue, the force signal may be affected by noise, increasing the possibility of misjudgment.

[0004] In addition, Chinese patent publication number CN118576284A discloses a puncture positioning compensation method, device, and puncture robot system. These methods utilize a dynamic error compensation mechanism. By providing real-time feedback on the actual position of each joint of the puncture robot, the position error relative to the command position is calculated, and a multi-cycle compensation strategy is used to dynamically correct the positioning command. When the error exceeds a threshold, the puncture position is updated using a compensation formula until the accuracy requirement or a stable state is achieved. Simultaneously, an inverse kinematics model is constructed, including a series-parallel hybrid digital structure model containing active / passive joints. The inverse kinematics relationship of the end effector is derived. The rotation matrix (Rn) accurately describes the puncture direction. Combined with the roll and pitch angles, the spatial position is converted into joint control commands. However, this existing patent is susceptible to interference from external factors such as changes in puncture speed and exhibits certain instability. Therefore, how to effectively filter out environmental noise and random interference from input and output to improve the robustness and reliability of recognition has become a technical problem that needs to be solved. Summary of the Invention

[0005] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and provide a method for identifying the state of a puncture robot penetrating tissue based on human body parameter estimation, so as to improve the accuracy and safety of the puncture robot during operation.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] According to one aspect of the present invention, a method for identifying a tissue penetration state of a puncture robot based on human body parameter estimation is provided, the method comprising the following steps:

[0008] Step S1, constructing a mathematical model of the distal patient's human tissue, wherein the mathematical model includes the coupling relationship between multiple layers of tissue and the dynamic characteristics of the tissue;

[0009] Step S2, based on the mathematical model constructed in step S1, and using sensors to collect real-time needle tip data during the puncture process, online parameter identification is performed on the model parameters, and the model parameters are updated in real time to reflect the actual changes in the tissue during the puncture process, wherein the needle tip data includes needle tip movement distance, force signal size, and velocity data;

[0010] In step S3, the model parameters are updated in real time according to step S2 to detect the moment of parameter mutation during the puncture process. The mutation moment is when the deviation of the model parameters from the historical mean at a certain moment exceeds the set threshold. This moment is judged to be the moment when the needle tip breaks through the tissue layer. This mutation moment serves as an accurate basis for judging the change of the puncture state and provides real-time feedback information to the puncture robot.

[0011] As a preferred technical solution, the mathematical model of the distal patient's human tissue is constructed in step S1 as follows:

[0012] According to the different requirements of human tissue conditions, a linear Kelvin-Voigt model of human tissue or a nonlinear Hunt-Crossley model of human tissue is established.

[0013] As a preferred technical solution, the linear Kelvin-Voigt model of human tissue is specifically:

[0014] The human body tissue is compared to a mechanical structure consisting of an ideal spring and an ideal damper in parallel. The total force F is the sum of the elastic force and the viscous force. The human body tissue model is expressed as follows:

[0015]

[0016] Among them, x(t), are the displacement and corresponding velocity values when the surgical instrument squeezes the human tissue, K and B are the elastic coefficient and viscosity coefficient of the tissue organ, respectively, and F(t) is the force applied by the surgical instrument on the human tissue.

[0017] As a preferred technical solution, the parameter identification process of the linear Kelvin-Voigt model of human tissue specifically includes:

[0018] For the linear model, the parameters that need to be identified are the stiffness parameter K and the viscosity parameter B. When the least squares method is used to identify the linear model, the update equation for the human tissue model parameter identification is as follows:

[0019]

[0020] Where t is the time bit in the discrete domain, is the unknown parameter that needs to be identified, which is the human tissue model parameter here, including the elastic coefficient K and viscosity coefficient of the tissue; Φ(t)∈R N is the known input quantity, specifically the joint displacement and joint velocity of the surgical instrument obtained through measurement; y(t) is the desired output quantity, specifically the contact force information obtained by the force detection element;

[0021] e(t) is the error value preset according to the parameter identification accuracy requirement; L(t) is the adaptive gain, P(t)∈R N is the covariance matrix, λ is the forgetting factor;

[0022] The parameters of the human tissue model can be calculated by combining equations (2)-(5).

[0023] As a preferred technical solution, the nonlinear Hunt-Crossley model of human tissue is specifically:

[0024] Considering that the human body is not an ideal viscoelastic model, soft tissue materials have energy dissipation and hysteresis when stress is unloaded, the nonlinear Hunt-Crossley model can be used to model human tissue to improve accuracy. The resulting human tissue model can be expressed as follows:

[0025]

[0026] in is the speed at which the surgical instrument squeezes the human tissue. The exponent n is a real number. In actual minimally invasive surgery, as the surgical instrument continues to squeeze the human tissue, the deformation x of the human tissue continues to increase. The exponential term n is used to represent the change in human tissue stiffness caused by the increase in the squeezing surface of the human tissue. k is the nonlinear elastic coefficient; b is the nonlinear viscosity coefficient, and F(t) is the force applied by the surgical instrument on the human tissue.

[0027] As a preferred technical solution, the parameter identification process of the nonlinear Hunt-Crossley model of human tissue specifically includes:

[0028] For nonlinear models, the parameters that need to be identified are the nonlinear elastic coefficient k, the nonlinear viscosity coefficient b, and the exponential term n. When the piecewise recursive Levenberg-Marquardt method is used for identification, the update equation for the human tissue model parameter identification is as follows:

[0029]

[0030] θ(t+1)=θ(t)+Δθ (8)

[0031] (J T J+λI)Δθ=J T r (9)

[0032] e j (θ)=F(t)-F HC (t) (10)

[0033]

[0034] Where t is the time bit in the discrete domain, is the unknown parameter that needs to be identified, specifically the human tissue model parameter, including the nonlinear elastic coefficient k, the nonlinear viscosity coefficient b and the exponential term n; F(t) is the known input force, specifically obtained by measuring the puncture needle force signal; F HC (t) is the estimated power of the Hunt-Crossley model, specifically the result of model fitting; e j (θ) is the residual vector, specifically the difference between the input force and the estimated force; J is the Jacobian matrix of the residual with respect to the parameter θ(t), specifically calculated by the partial derivative of the vector with respect to the parameter;

[0035] If the residual sum of squares decreases, that is, E(θ n+1 ) <E(θ n ), then reduce the damping factor v>1, where v is a constant greater than 1; if the residual sum of squares increases (i.e. E(θ n+1 )>E(θ n )), then increase the damping factor λ n+1 =νλ n ,ν>1; continue to iterate until the increment ||Δθ||<∈ or the number of iterations reaches the upper limit K max So far, where ∈ is the desired small limit value, the model parameters can be calculated by combining equations (7)-(11).

[0036] In the above model parameter identification process, no matter which model is used to establish the human tissue, as long as the three quantities in the model are known, the model parameters can be identified. These three quantities are: the deformation of the human tissue x, the speed of the surgical instrument squeezing the human tissue, and the speed of the surgical instrument squeezing the human tissue. and the squeezing force F borne by human tissue; the position quantity x can be directly obtained by a position sensor such as an encoder installed at the end of the surgical instrument actuator, and the speed quantity The speed can be directly obtained by differentiating the above-mentioned velocity; and the extrusion force can be detected by a force-torque sensor.

[0037] In general, if a system can be represented by constant parameters Known functions and It is constructed by linear combination, namely: Φ(t) T Θ=y(t), then The parameters contained in are called identifiable, where Φ(t)=[φ1(t),φ2(t),...,φ m (t)] is a known time function vector, Θ=[θ1(t),...,θ m (t)] is the unknown parameter that needs to be identified, and y(t) is a known time scalar function.

[0038] As a preferred technical solution, the parameters of the linear Kelvin-Voigt model of human tissue are updated as follows:

[0039] Based on the human tissue model parameters obtained by identification, a virtual human tissue model is established on the main controller of the doctor's operating table, and the following linear virtual model is obtained:

[0040]

[0041] in

[0042] The obtained linear virtual model is used to analyze human tissue, and the parameters of human tissue will change when breakthrough occurs.

[0043] As a preferred technical solution, the parameters of the nonlinear Hunt-Crossley model of human tissue are updated as follows:

[0044] Based on the human tissue model parameters obtained by identification, a virtual human tissue model is established on the main controller of the doctor's operating table to obtain a nonlinear virtual model:

[0045]

[0046] in

[0047] The human tissue is analyzed through the obtained nonlinear virtual model, and the parameters of the human tissue will change when breaking through.

[0048] As a preferred technical solution, step S3 is specifically as follows:

[0049] According to the changes of formula (12) or (13) in step S2, the breakthrough state of the puncture robot is identified. During the execution of the puncture robot, the historical mean value is calculated using the collected parameter data, and the fluctuation range of the parameter is determined by the standard deviation.

[0050] As a preferred technical solution, the identification of the tissue breakthrough state of the puncture robot is specifically as follows:

[0051] The dynamic changes of the parameters of the puncture robot during operation are monitored in real time. For the calculated k and b parameters, the following Z scores can be calculated respectively:

[0052]

[0053] Where X is the current value, specifically the current parameter k or b; μ is the average of the historical values. If X is calculated using the k parameter, this is the historical mean of k; if X is calculated using the b parameter, this is the historical mean of b; σ is the standard deviation of the historical data. If the first two values are calculated using the k parameter, this is the historical standard deviation of k; if the first two values are calculated using the b parameter, this is the historical standard deviation of b.

[0054] The Z score is the standard deviation distance between the current parameter and the historical mean, which is used to measure the degree of deviation of a new data from the mean. When it exceeds the threshold, it is considered that a significant change has occurred. When the real-time monitoring shows that the change amplitude of the parameter at a certain moment exceeds the preset threshold, it is judged that the puncture robot has broken through the current tissue layer and entered another tissue layer. The current moment and current state parameters are recorded, and combined with the parameter values, the change of tissue layer type in the puncture path is further determined.

[0055] Compared with the prior art, the present invention has the following advantages:

[0056] 1) The present invention combines multiple collected data to identify the dynamic parameters of the tissue in real time, directly reflecting the intrinsic mechanical properties of the tissue, rather than relying solely on external signals. This can effectively filter out environmental noise and random interference from input and output, thereby improving the robustness and reliability of identification;

[0057] 2) The present invention achieves dynamic tracking and immediate analysis of human tissue status by calculating model parameters at each sampling point in real time during the puncture operation, ensuring that the puncture robot can quickly respond to changes in tissue status;

[0058] 3) The present invention greatly improves the accuracy and safety of the puncture robot during operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 A simplified diagram of a master-slave remote teleoperation surgery system with time delay;

[0060] Figure 2 This is a block diagram of the control system of the master-slave remote teleoperation surgery system;

[0061] Figure 3 This is a block diagram of the tissue penetration state identification method of the puncture robot based on human body parameter estimation of the present invention. DETAILED DESCRIPTION

[0062] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0063] like Figure 1 As shown, the master-slave remote teleoperation surgery system includes an operator dynamic model 1, a master-end operating handle 2, a master-end joint drive motor 3, a network delay link 4, a slave-end controller 5, a slave-end drive motor 6, a slave-end mechanism actuator 7 and a remote environment model 8. The master-side operating handle 2 is fixedly connected to the master-side joint drive motor 3. The operator's dynamic model 1 moves the master-side operating handle 2, thereby driving the master-side joint drive motor 3 to rotate. A position sensor mounted on the motor shaft detects the displacement and transmits it to the slave-side controller 5 via a network delay link 4. This controller compares the current displacement of the slave mechanism with the displacement sent from the master, generating a corresponding drive torque signal for the slave drive motor 6, which drives the motor to rotate. This ultimately drives the slave mechanism actuator 7 mounted on the motor shaft into contact with the remote environment model 8. The contact force generated by the slave mechanism actuator 7 and the remote environment model 8 is fed back again via the network delay link 4 to the master-side joint drive motor 3, generating a corresponding drive force signal for the master-side joint drive motor 3, driving the master-side operating handle 2 in the opposite direction of the desired movement of the operator's dynamic model 1, thereby applying a counterforce to the operator. This completes the signal flow process of the teleoperation system.

[0064] Figure 2 The control system block diagram is shown in the form of Figure 1 The master-slave remote teleoperation system is described in detail. The operator dynamics model 1 is equivalent to Figure 2 The main end operator 9 in the operator dynamics model 1 exerts the action f h Feedback force f from the slave end sd After combining, the force f applied to the master robot 10 is obtained.m Here, the master robot 10 includes Figure 1 The main end robot 10 has two modules: the main end operating handle 2 and the main end joint driving motor 3. m Under the action of The displacement is converted into a signal after passing through the first network delay module 11.1. The displacement signal is sent to the slave end, and the slave end robot 13 (the slave end robot 13 includes Figure 1 The current displacement of the slave drive motor 6 and the slave actuator 7 in the The comparison value is transmitted to the slave controller module 12, and the slave controller module 12 calculates the force f s Split into two paths, one signal passes through the second network delay module 11.2 and becomes f sd Signal, transmitted to the master operator 9; force f s The other signal is connected to the environment model Z e The driving force f of the slave robot 13 is obtained by comparing the force signal sc , so that the slave robot 13 generates a motion signal

[0065] Figure 3 The figure shows a control block diagram of a method for identifying the state of a puncture robot breaking through tissue based on human body parameter estimation disclosed in the present invention. The parameter identification module 15 uses the displacement x obtained by measurement. s ,speed and contact force information f e Assume that the distal model 14 is represented by the linear Kelvin-Voigt model as follows:

[0066]

[0067] in, is the speed of the surgical instrument squeezing the human tissue; K and B are the elastic coefficient and viscosity coefficient of the tissue organ respectively; F(t) is the force exerted by the surgical instrument on the human tissue. The parameter identification module 15 uses the most commonly used recursive least squares method. Since the linear Kelvin-Voigt model is selected, the unknown parameter vector in the parameter identification equation (4) is for Recursive known variables The identification process of unknown parameters is obtained through the following recursive process: where y kThe force exerted by human tissue on surgical instruments is output after each cycle of recursive calculation, and the parameter identification results K and B are used to identify the model parameters. The obtained parameters of the remote model 14 are sent to the master end of the teleoperation system through the second network delay module 11.2, and the puncture state is identified based on the parameters to detect the breakthrough moment.

[0068] The puncture state recognition model 16 is used to determine whether the puncture is complete. It uses the Z score to detect whether the parameters have changed significantly. When the parameter change exceeds the threshold, it means that the tissue layer has been broken through and the puncture is completed, so the needle should be stopped immediately. In this system, the Z score is specifically expressed as follows:

[0069]

[0070] Where X is the current value, which here is the current parameter k or b, depending on which parameter is used for calculation and needs to be consistent with subsequent data; μ is the average of the historical values. If X is calculated using the k parameter, here it is the historical mean of k; if X is calculated using the b parameter, here it is the historical mean of b; σ is the standard deviation of the historical data. If the first two values are calculated using the k parameter, here it is the historical standard deviation of k; if the first two values are calculated using the b parameter, here it is the historical standard deviation of b. The result of expression (14) is compared with the threshold. When it exceeds the threshold, it can be considered that a significant change has occurred, and it is determined that the puncture robot has broken through the current tissue layer.

[0071] The switch module 17 is used to control whether the needle is inserted. If the puncture is not completed, the force signal f sd Feedback is sent to the master robot 10. If the needle is inserted, the system should stop immediately to avoid over-puncture. At this time, the switch is disconnected and the system stops moving. Through the above process, at each sampling moment, the system can determine whether the puncture is completed. If the puncture is not completed, the precise feedback force signal is transmitted to the master operator 9. If the puncture is completed, the needle is stopped.

[0072] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and such modifications or substitutions are intended to be within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be subject to the scope of protection of the claims.

Claims

1. A method for identifying tissue penetration status of a puncture robot based on human body parameter estimation, characterized in that: The method comprises the following steps: Step S1, constructing a mathematical model of the distal patient's human tissue, wherein the mathematical model includes the coupling relationship between multiple layers of tissue and the dynamic characteristics of the tissue; Step S2: Based on the mathematical model constructed in step S1, the needle tip data during the puncture process is collected in real time by a sensor, and the model parameters are identified online, and the model parameters are updated in real time to reflect the actual changes in the tissue during the puncture process; In step S3, the model parameters are updated in real time according to step S2 to detect the moment of parameter mutation during the puncture process. The mutation moment is when the deviation of the model parameters from the historical mean at a certain moment exceeds the set threshold, and it is judged that this moment is the moment when the needle tip breaks through the tissue layer.

2. The method for identifying tissue penetration status of a puncture robot based on human body parameter estimation according to claim 1, characterized in that: The mathematical model of the distal patient's human tissue is constructed in step S1 as follows: According to the different requirements of human tissue conditions, a linear Kelvin-Voigt model of human tissue or a nonlinear Hunt-Crossley model of human tissue is established.

3. The method for identifying tissue penetration status of a puncture robot based on human body parameter estimation according to claim 2, characterized in that: The linear Kelvin-Voigt model of human tissue is specifically: The human body tissue is compared to a mechanical structure consisting of an ideal spring and an ideal damper in parallel. The total force F is the sum of the elastic force and the viscous force. The human body tissue model is expressed as follows: Among them, x(t), are the displacement and corresponding velocity values when the surgical instrument squeezes the human tissue, K and B are the elastic coefficient and viscosity coefficient of the tissue organ, respectively, and F(t) is the force applied by the surgical instrument on the human tissue.

4. The method for identifying tissue penetration status of a puncture robot based on human body parameter estimation according to claim 3, characterized in that: The parameter identification process of the linear Kelvin-Voigt model of human tissue specifically includes: For the linear model, the parameters that need to be identified are the stiffness parameter K and the viscosity parameter B. When the least squares method is used to identify the linear model, the update equation for the human tissue model parameter identification is as follows: Where t is the time bit in the discrete domain, is the unknown parameter that needs to be identified, which is the human tissue model parameter here, including the elastic coefficient K and viscosity coefficient of the tissue; Φ(t)∈R N is the known input quantity, specifically the joint displacement and joint velocity of the surgical instrument obtained through measurement; y(t) is the desired output quantity, specifically the contact force information obtained by the force detection element; e(t) is the error value preset according to the parameter identification accuracy requirement; L(t) is the adaptive gain, P(t)∈R N is the covariance matrix, λ is the forgetting factor; The parameters of the human tissue model can be calculated by combining equations (2)-(5).

5. The method for identifying tissue penetration status of a puncture robot based on human body parameter estimation according to claim 2, characterized in that: The nonlinear Hunt-Crossley model of human tissue is specifically: Considering that the human body is not an ideal viscoelastic model, soft tissue materials have energy dissipation and hysteresis when stress is unloaded, the nonlinear Hunt-Crossley model can be used to model human tissue to improve accuracy. The resulting human tissue model can be expressed as follows: in, is the speed at which the surgical instrument squeezes the human tissue, the exponential term n is used to represent the change in human tissue stiffness caused by the increase in the squeezing surface of the human tissue; k is the nonlinear elastic coefficient; b is the nonlinear viscosity coefficient, and F(t) is the force applied by the surgical instrument on the human tissue.

6. The method for identifying tissue penetration status of a puncture robot based on human body parameter estimation according to claim 5, characterized in that: The parameter identification process of the nonlinear Hunt-Crossley model of human tissue specifically includes: For nonlinear models, the parameters that need to be identified are the nonlinear elastic coefficient k, the nonlinear viscosity coefficient b, and the exponential term n. When the piecewise recursive Levenberg-Marquardt method is used for identification, the update equation for the human tissue model parameter identification is as follows: θ(t+1)=θ(t)+Δθ (8) (J T J+λI)Δθ=J T r (9) e j (θ)=F(t)-F HC (t) (10) Where t is the time position in the discrete domain, θ(t)∈R N is the unknown parameter that needs to be identified, specifically the human tissue model parameter, including the nonlinear elastic coefficient k, the nonlinear viscosity coefficient b and the exponential term n; F(t) is the known input force, specifically obtained by measuring the puncture needle force signal; F HC (t) is the estimated power of the Hunt-Crossley model, specifically the result of model fitting; e j (θ) is the residual vector, specifically the difference between the input force and the estimated force; J is the Jacobian matrix of the residual with respect to the parameter θ(t), specifically calculated by the partial derivative of the vector with respect to the parameter; If the residual sum of squares decreases, that is, E(θ n+1 ) <E(θ n ), then reduce the damping factor Where v is a constant greater than 1; if the residual sum of squares increases (i.e. E(θ n+1 )>E(θ n )), then increase the damping factor λ n+1 =vλ n ,v>1; continue to iterate until the increment ||Δθ||<∈ or the number of iterations reaches the upper limit K max So far, where ∈ is the desired small limit value, the model parameters can be calculated by combining equations (7)-(11).

7. The method for identifying tissue penetration status of a puncture robot based on human body parameter estimation according to claim 4, characterized in that: The parameters of the linear Kelvin-Voigt model of human tissue are updated as follows: Based on the human tissue model parameters obtained by identification, a virtual human tissue model is established on the main controller of the doctor's operating table, and the following linear virtual model is obtained: in The obtained linear virtual model is used to analyze human tissue, and the parameters of human tissue will change when breakthrough occurs.

8. The method for identifying tissue penetration status of a puncture robot based on human body parameter estimation according to claim 6, characterized in that: The parameters of the nonlinear Hunt-Crossley model of human tissue are updated as follows: Based on the human tissue model parameters obtained by identification, a virtual human tissue model is established on the main controller of the doctor's operating table to obtain a nonlinear virtual model: in The human tissue is analyzed through the obtained nonlinear virtual model, and the parameters of the human tissue will change when breaking through.

9. A method for identifying tissue penetration status of a puncture robot based on human body parameter estimation according to claim 7 or 8, characterized in that: The step S3 is specifically as follows: According to the changes of formula (12) or (13) in step S2, the breakthrough state of the puncture robot is identified. During the execution of the puncture robot, the historical mean value is calculated using the collected parameter data, and the fluctuation range of the parameter is determined by the standard deviation.

10. The method for identifying tissue penetration status of a puncture robot based on human body parameter estimation according to claim 9, characterized in that: The identification of the tissue breakthrough status of the puncture robot is specifically as follows: The dynamic changes of the parameters of the puncture robot during operation are monitored in real time. For the calculated k and b parameters, the following Z scores can be calculated respectively: Where X is the current value, specifically the current parameter k or b; μ is the average of the historical values. If X is calculated using the k parameter, this is the historical mean of k; if X is calculated using the b parameter, this is the historical mean of b; σ is the standard deviation of the historical data. If the first two values are calculated using the k parameter, this is the historical standard deviation of k; if the first two values are calculated using the b parameter, this is the historical standard deviation of b. The Z score is the standard deviation distance between the current parameter and the historical mean, which is used to measure the degree of deviation of a new data from the mean. When it exceeds the threshold, it is considered that a significant change has occurred. When the real-time monitoring shows that the change amplitude of the parameter at a certain moment exceeds the preset threshold, it is judged that the puncture robot has broken through the current tissue layer and entered another tissue layer. The current moment and current state parameters are recorded, and combined with the parameter values, the change of tissue layer type in the puncture path is further determined.

Images