An intravascular plaque laser ablation robot and control system

Abstract

The application discloses a kind of endovascular plaque laser ablation robot and control system, it is related to vascular intervention technical field, robot includes catheter, the inside of catheter is embedded with mechanical feedback guide wire;One end of catheter is provided with hemispherical head end by elastic connecting ring;Hemispherical head end is installed with laser confocal endoscope and laser ablation optical fiber, the tail end of laser ablation optical fiber is connected with laser ablation system, and laser confocal endoscope and laser ablation optical fiber are all arranged in catheter;The end of catheter shell away from hemispherical head end is installed with propeller, and controller and power device are arranged in the inside of propeller, and catheter is detachably installed on power device;Transmission assembly is arranged in the inside of catheter, and hemispherical head end is transmission matched with power device by transmission assembly, and mechanical feedback guide wire, propeller and power device are electrically connected with controller.The application can greatly reduce the risk of blood vessel perforation, and single ablation operation time can be shortened to about 20 minutes.

Description

Technical Field

[0001] This invention relates to the field of vascular interventional technology, and in particular to a robot and control system for laser ablation of plaques in blood vessels. Background Technology

[0002] Vascular diseases are among the most prevalent and deadliest diseases worldwide. Interventional therapy is currently the most common and effective treatment method, significantly reducing patient mortality. However, for high-resistance vascular lesions, calcified lesions, chronic occlusive lesions, and diffuse lesions, the surgeries are complex, with high complication rates, and often require procedures such as coronary rotational atherectomy, excimer laser, and shockwave balloon angioplasty. These procedures greatly improve the success rate of treating these diseases.

[0003] However, these techniques are relatively complex, have a learning curve, are suitable for experienced operators, are relatively difficult, require more consumables, and have a relatively high complication rate. Furthermore, these techniques share a common characteristic: a guidewire must first pass through the lesion to establish a track before subsequent procedures can be performed. Therefore, the main reasons for interventional treatment failure can be attributed to: ① the lesion is difficult to diagnose, making it impossible to use the instruments smoothly; ② complications requiring termination of the procedure occur; ③ track establishment fails, failing to meet the necessary conditions for instrument activation.

[0004] Currently, while there are devices that provide real-time OCT (Optical Coherence Tomography) imaging during surgery for CTO (chronic total occlusion), allowing surgeons to guide the cutting catheter within the vessel lumen and minimize arterial damage during passage through the occlusion, these devices have certain drawbacks: ① Blood can affect OCT imaging, necessitating continuous injection of perfusion fluid to drain blood during the procedure, which can cause iatrogenic myocardial ischemia; ② OCT can only scan the surrounding tissue behind the device tip, exhibiting hysteresis and increasing the risk of complications such as vessel rupture and dissection.

[0005] CN114469333A describes an ablation catheter, laser ablation system, and method for intravascular laser plaque ablation; CN116672075A describes an imaging catheter, system, and method. Both of these existing technologies use OCT as an ablation guidance tool, which suffers from the aforementioned drawbacks of OCT. Furthermore, these technologies can only perform imaging and cannot reliably guide the direction of laser ablation, nor can they automatically advance the laser; they rely solely on manual operation by the operator, lacking precision. Moreover, traditional lasers emit a single wavelength, offering no targeted energy and wavelength adjustment for different lesions.

[0006] In summary, current intravascular plaque laser ablation technology has the following problems: ① lack of effective imaging methods to guide the laser direction; ② lack of adjustable ablation laser energy; ③ poor precision. Therefore, existing technology suffers from a high complication rate in vascular laser ablation, severely limiting its application. Summary of the Invention

[0007] The purpose of this invention is to provide a robot and control system for laser ablation of plaques in blood vessels, so as to solve the problems existing in the prior art.

[0008] To achieve the above objectives, the present invention provides a laser ablation robot for intravascular plaques, comprising a catheter, wherein multiple mechanical feedback guidewires are embedded inside the catheter; one end of the catheter is provided with a hemispherical tip, and an elastic connecting ring is provided between the hemispherical tip and the catheter, wherein the hemispherical tip is fixedly mounted on the catheter through the elastic connecting ring; A laser confocal endoscope and a laser ablation fiber are mounted on the hemispherical tip. The laser ablation fiber is arranged at equal intervals along the circumference outside the laser confocal endoscope. The end of the laser ablation fiber is connected to a laser ablation system. Both the laser confocal endoscope and the laser ablation fiber are inserted into the catheter. The laser confocal endoscope is used to identify the type of plaque in the plaque area within the blood vessel, and the laser ablation fiber is used for laser transmission and ablation of the plaque. The outer side of the conduit is fitted with a conduit shell, and the conduit shell is provided with a coolant injection port for injecting coolant. A pusher is installed at the end of the conduit shell away from the hemispherical head end. The pusher is equipped with a controller and a power device. The conduit is detachably installed on the power device. A transmission assembly is installed inside the conduit. The hemispherical head end is connected to the power device through the transmission assembly. The mechanical feedback guide wire, the pusher, and the power device are all electrically connected to the controller.

[0009] Preferably, the transmission assembly includes an inner transmission shaft, one end of which is connected to the power device, and the other end of which is fixedly connected to an inner annular hinge; an outer transmission shaft is sleeved on the outside of the inner transmission shaft, one end of which is connected to the power device, and the other end of which is fixedly connected to an outer annular hinge, with the inner annular hinge rotatably engaging with the outer annular hinge; both the outer annular hinge and the outer transmission shaft are disposed within the guide tube; Both the inner and outer annular hinges have two first arc-shaped protrusions on the side near the hemispherical head end. The central angle between the two first arc-shaped protrusions on the same annular hinge is 135°, and at most one of the first arc-shaped protrusions on the inner annular hinge and the first arc-shaped protrusion on the outer annular hinge is opposite each other. Four second arc-shaped protrusions are equally spaced on the side of the hemispherical head end near the first arc-shaped protrusions. The inner and outer annular hinges adjust the orientation of the hemispherical head end through the first arc-shaped protrusions and the second arc-shaped protrusions.

[0010] Preferably, the end of the conduit shell away from the hemispherical head is provided with a shell buckle, and the head end of the pusher is provided with a shell groove. The conduit shell and the pusher are connected through the shell buckle and the shell groove. The head end of the power unit is provided with multiple conduit interfaces. The conduits are detachably installed on the power unit through the conduit interfaces, and the mechanical feedback guidewire is electrically connected to the controller through the conduit interfaces.

[0011] A robotic control system for intravascular plaque laser ablation, based on the aforementioned intravascular plaque laser ablation robot, includes: The sensing module is used to identify the plaque type in the plaque area within the blood vessel and adjust the laser wavelength according to the plaque type to ablate the plaque; A propulsion module is used to control the motion state of the sensing module according to the magnitude of the provided thrust. The laser control module is used to adjust the laser parameters in the sensing module according to the identified patch type and to determine the laser irradiation range. The dynamic ablation planning module is used to dynamically plan the laser ablation sequence and path of the patch area based on the guide wire tension, patch distribution and location. The thermal management and safety control module is used to monitor the temperature of the eroded area in real time and ensure that the temperature is always within a safe range.

[0012] Preferably, the method for the sensing module to identify patch types includes: S1.1 Feature extraction; High-resolution images of the vascular intima were obtained using laser confocal endoscopy, and the following structural indicators were extracted: Endothelial continuity index : Measures whether there is tearing or interruption of the endothelium, with a range of values. The endothelial fluorescence signal is continuously 0, and the discontinuous endothelial fluorescence signal is 1, indicating tearing or interruption. Surface roughness index : Calculated from image texture entropy; Reflectivity distribution characteristics By analyzing the reflectance changes in images obtained by laser confocal endoscopy, the lipid core and calcified particles can be identified. Morphological statistical characteristics This includes parameters such as patch thickness and area. Therefore, the structural feature vector is: ; After injection of ICG fluorescent agent, the fluorescence signal of the blood vessel wall was recorded at different time points, and the following key parameters were extracted: fluorescence intensity ratio : ; This indicates the fluorescence intensity 60 seconds after ICG fluorescent agent injection. This represents the initial maximum fluorescence intensity, expressed in AU. elution rate : ; Fluorescence change slope : The rate of change calculated from the fluorescence time curve; The fluorescence feature vector is: ; S1.2 Feature fusion; To balance the recognition value of both structural feature vectors and fluorescence feature vectors, a weighted fusion strategy is adopted: in, Indicates feature concatenation; weights , These represent the contribution weights of structural and fluorescence characteristics, respectively. The feature vector obtained after fusion ; S1.3, Classification; Fusion features The input is fed into a nonlinear support vector machine model to determine whether a given region is a patch; the model is defined as follows: in: For the first The true class labels of each training sample correspond to the plaque region and the normal endothelial region, respectively; For Lagrange multipliers, It is a radial basis function (RBF) kernel. Indicates the first training set The fusion features of each sample For the bias term, output This represents the classification result output by the model at the end; S1.4 Result Interpretation and Output Control; The classification results are transmitted to the laser ablation system. If a patchy area is identified, the ablation wavelength and energy are adjusted, and targeted ablation is performed based on the fluorescence type. Medium-power, medium-wavelength lasers were selected for lipid plaques; High-energy, short-wavelength lasers are used to treat calcification lesions. Low-energy broadband lasers were used to treat the fiber patches. When the plaque structure is eroded to the point where the mechanical feedback is less than the minimum preset value of 0.3N, the conduit enters the next area and restarts the detection-identification-erosion cycle.

[0013] Preferably, in the feature fusion, the feature weights are allocated as follows: Structural feature contribution weight: 60%; Fluorescence feature contribution weight: 40%.

[0014] Preferably, the method for the propulsion module to control the motion state of the sensing module includes: S2.1 Obtain the current thrust of the power unit; During propulsion, the current of the power unit is acquired in real time. With propulsion speed The current thrust is estimated based on the following linear model: in, The current propulsion thrust is expressed in N. The current operating current of the current power unit is expressed in amperes (A) and ranges from [0,10]. The linear propulsion speed of the current power unit is in mm / s, ranging from [0, 0.5]. The coefficient 0.5 is the thrust constant (N / A), and the coefficient 0.1 is the viscous damping coefficient (N·s / mm). The negative correlation between F(t) and v(t) reflects the constant power drive characteristics of the motor under variable load conditions. When the propulsion resistance increases, the motor speed decreases and the current increases, which together lead to an increase in the estimated thrust. S2.2, Advance control process; S2.21, Start-up Initialization: Power unit returns to zero position; guide wire tension sensor calibrated to... ; S2.22. Real-time monitoring of the power unit's status: reading the current. and speed Substitute the values ​​into the inference model to calculate the instantaneous thrust. ; S2.23, Thrust Judgment and Feedback Control: If If the vascular resistance is considered low, continue proceeding; if If a patch is detected, the process pauses and patch recognition is activated; where F min =0.3N, F max =0.5N; S2.24, Advance / Pause / Laser Co-operation Switching: If advancement is allowed, continue linear advancement; if advancement is not allowed, initiate co-laser confocal endoscope recognition, perform laser ablation, and resume advancement after the mechanical feedback decreases again.

[0015] Preferably, the control method of the laser control module includes: S3.1 Low-power scanning: The endothelial region marked by laser confocal endoscopy is surface scanned using a 532nm low-power laser to excite autofluorescence to verify endothelial integrity; S3.2, Ablation Restricted Area Setting: When the ICG fluorescence intensity is within the normal endothelial range, laser emission in that area is automatically blocked. The normal endothelial range is F... IR Values ​​range from 0.8 to 1.2; S3.3, Patch-specific erosion removal: Lipid plaques: preferentially activated by the 2940nm erbium-doped yttrium aluminum garnet laser therapy device, energy density formula: E=40-0.1⋅F IR (mJ / cm) 2 ) When F IR Enabled when >2.0; where 40 is the reference energy density and 0.1 is F. IR Adjustment coefficient; Calcified plaques: The 308nm excimer laser is activated only when a WR > 80% is detected. Energy density formula: E=100+0.2⋅WR(mJ / cm 2 ) When F IR Enabled when <0.5; where 100 is the reference energy density and 0.2 is the WR adjustment coefficient; WR (Washout Ratio) is the fluorescence elution rate, obtained according to the calculation method of elution rate in S1.1, reflecting the rate at which the ICG fluorescent agent is eluted from the plaque area, with a value range of 0 to 100%; Fiber patch: A Z-type fiber combiner is used to achieve a dual-band laser coupling efficiency of more than 97%, with alternating output and a laser switching time of less than 1ms.

[0016] Preferably, the method for the dynamic ablation planning module to plan the laser ablation sequence and path includes: S4.1 Triggering conditions and graph construction; When the catheter tip encounters resistance (≥0.5N), confocal laser endoscopy and ICG fluorescence mapping are triggered to acquire a plaque map of a segment of the vessel. Candidate regions for erosion are defined based on fluorescence thresholds. ; S4.2, Reinforcement Learning Model Construction; S4.21, Environment Definition; state space The current catheter tip is shown in the image. The position above ,coordinate , and local curvature Local plaque characteristics I t Information such as the previous action, etc. t-1 : Action space Laser direction selection dir i Laser power level j Should we proceed with the move? k : ; S4.22, Reward Function Design; Define reward function Used to guide path selection to achieve the optimal strategy: in, This indicates the erosion efficiency of the current patch area per unit time; Indicates the safety factor; η represents the laser energy consumption per unit time. t σ t ε t All values ​​are dimensionless values ​​after normalization (range 0-1), η t σ is calculated based on the rate of change in fluorescence intensity before and after erosion. t Based on the mechanical feedback guidewire tension data, ε t The reward signal for reinforcement learning is calculated based on the product of laser power and irradiation time; each component is normalized and then weighted and summed. S4.3 Reinforcement Learning Algorithm Selection; Deep Q-Learning is used as the control core, and the network structure is as follows: Input layer: State vector Hidden layer: 2 fully connected layers, 128 nodes each; Output layer: Q-value for each action. Strategy Update: ; in, For learning rate, Discount factor; S4.4 Decision performance indicators; After the erosion area is planned, the laser control module automatically adjusts the following parameters: laser wavelength, laser power, and duration. If the fluorescence intensity of a certain area to be etched is continuously irradiated with laser, then the fluorescence intensity of that area is... Once the area is deemed to have completed erosion, the propulsion module is reactivated.

[0017] Preferably, the control method for the thermal management and safety control module includes: S5.1 Cooling control trigger mechanism: The real-time temperature monitoring point is located 1mm to 2mm in front of the catheter tip; when the detected temperature... Exceeding the critical temperature The cooling injection mechanism is activated; a temperature rise rate-driven injection model is adopted, and the injection intensity is controlled according to the heat accumulation rate. S5.2, Jet Flow Model: Jet velocity of coolant With current temperature Temperature difference exceeding the threshold The flow rate is directly proportional to the temperature, and increases linearly with increasing temperature. ; Where 0.05 is in mL / (s·℃), it means that when the temperature exceeds 60℃, the flow rate increases by 0.05 mL / s for every 1℃ increase; the value range of Q [0, 0.2] mL / s represents the physical feasible domain of the jet flow rate; when T > 64℃, the jet flow rate has reached the maximum cooling capacity of the system. At this time, the emergency withdrawal mechanism of S5.3 should be activated as a supplementary safety measure, rather than continuing to rely on cooling jets for temperature control; S5.3 Emergency withdrawal mechanism; S5.31, Triggering conditions and logical judgments; This mechanism is based on two core triggering conditions: Mechanical contact threshold: contact force ; Temperature safety threshold: ; The emergency rollback mechanism can be triggered if any one of the conditions is met. S5.32, Withdrawal Control Formula; Logical function for judgment: ; in, Real-time temperature; For real-time contact force; For triggering flag variables; Withdrawal control function: when The power unit switches from propulsion mode to retraction mode, controlling the duct to operate at a fixed speed. Reverse motion; Power unit position control commands: , Indicates the distance traveled; S5.33, Cooling Co-control; During the retraction, the cooling mechanism continues to operate to rapidly reduce tissue temperature: .

[0018] Compared with the prior art, the present invention has the following advantages and technical effects: The present invention provides a robot and control system for intravascular plaque laser ablation. The catheter is inserted into the blood vessel and stops advancing upon encountering mechanical feedback. Laser confocal endoscopic scanning then begins. Using fluorescence combined with indocyanine green tracing, the intravascular plaque is stained, and a plaque distribution spectrum is mapped. By analyzing differentially stained areas, laser ablation targets are identified, and the ablation wavelength and energy are adjusted to ablate the plaque area. Once the plaque has been ablated to a certain extent, the mechanical feedback disappears, and the catheter automatically advances, repeating the cycle. This invention can significantly reduce the risk of vascular perforation, and the time for a single ablation procedure can be shortened to approximately 20 minutes. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the structure of the intravascular plaque laser ablation robot of the present invention; Figure 2 This is a schematic cross-sectional view of the inside of the catheter of the present invention; Figure 3 This is a schematic diagram showing the connection between the inner and outer annular hinges and the transmission assembly of the present invention. Figure 4 This is a cross-sectional schematic diagram of the hemispherical head end of the present invention; Figure 5 This is a schematic diagram of the inner and outer annular hinges of the present invention; Figure 6 This is a schematic diagram of the power device of the present invention; Figure 7 This is a top view of the power unit of the present invention; In the diagram: 1. Laser confocal endoscope; 2. Laser ablation fiber; 3. Hemispherical tip; 4. Inner annular hinge; 5. Inner drive shaft; 6. Outer annular hinge; 7. Outer drive shaft; 8. Conduit; 9. Mechanical feedback guidewire; 10. Conduit housing; 11. Coolant inlet; 12. Housing clip; 13. Thruster; 131. Thruster housing; 132. Telescopic cylinder; 14. Conduit interface; 15. Housing slot; 16. Power unit; 161. Power unit housing; 162. First motor; 163. Mounting shaft; 164. Inner shaft connector; 165. Outer shaft connector; 166. Gear ring; 167. Second motor; 168. Gear. Detailed Implementation

[0021] It should be noted that, unless otherwise specified, the embodiments and features described in this invention can be combined with each other. The described embodiments are merely some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of this invention. The invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0022] like Figures 1 to 5 As shown, the present invention provides a laser ablation robot for intravascular plaques, including a catheter 8, with multiple mechanical feedback guidewires 9 embedded inside the catheter 8; a hemispherical head end 3 is provided at one end of the catheter 8, and an elastic connecting ring is provided between the hemispherical head end 3 and the catheter 8, and the hemispherical head end 3 is fixedly installed on the catheter 8 through the elastic connecting ring. A laser confocal endoscope 1 and a laser ablation fiber 2 are installed on the hemispherical end 3. The laser ablation fiber 2 is arranged at equal intervals along the circumference on the outside of the laser confocal endoscope 1. The end of the laser ablation fiber 2 is connected to a laser ablation system. Both the laser confocal endoscope 1 and the laser ablation fiber 2 are inserted into the catheter 8. The laser confocal endoscope 1 is used to identify the plaque type in the plaque area in the blood vessel, and the laser ablation fiber 2 is used for laser transmission and ablation of the plaque. The outer side of the conduit 8 is fitted with a conduit housing 10, and the conduit housing 10 is provided with a coolant inlet 11 for injecting coolant. A pusher 13 is installed at the end of the conduit shell 10 away from the hemispherical head end 3. The pusher 13 is equipped with a controller and a power unit 16. The conduit 8 is detachably installed on the power unit 16. A transmission assembly is installed inside the conduit 8. The hemispherical head end 3 is connected to the power unit 16 through the transmission assembly. The mechanical feedback guide wire 9, the pusher 13 and the power unit 16 are all electrically connected to the controller.

[0023] The mechanical feedback guide wire 9 can transmit the force exerted on the guide tube 8 when it is bent to the controller, thereby facilitating the drive control of the thruster 13 and the power unit 16.

[0024] Through the coordinated operation of catheter 8, mechanical feedback guidewire 9, hemispherical tip 3, laser confocal endoscope 1, laser ablation fiber optic 2, catheter housing 10, pusher 13, controller and power unit 16, precise identification and targeted ablation of plaques in blood vessels can be achieved. At the same time, the mechanical feedback guidewire 9 can sense the force on the tip of catheter 8 in real time, providing a data basis for subsequent advancement control and safety protection, thereby effectively avoiding the risk of vascular perforation caused by blind advancement.

[0025] The scheme is further optimized. The transmission assembly includes an inner drive shaft 5, one end of which is connected to the power unit 16, and the other end of which is fixedly connected to an inner annular hinge 4. An outer drive shaft 7 is sleeved on the outside of the inner drive shaft 5. One end of the outer drive shaft 7 is connected to the power unit 16, and the other end of the outer drive shaft 7 is fixedly connected to an outer annular hinge 6. The inner annular hinge 4 and the outer annular hinge 6 are rotatably engaged. Both the outer annular hinge 6 and the outer drive shaft 7 are located inside the guide tube 8. Both the inner ring hinge 4 and the outer ring hinge 6 have two first arc-shaped protrusions on the side near the hemispherical head end 3. The central angle between the two first arc-shaped protrusions on the same ring hinge is 135°, and at most only one of the first arc-shaped protrusions on the inner ring hinge 4 and the first arc-shaped protrusion on the outer ring hinge 6 is opposite to each other. Four second arc-shaped protrusions are equally spaced on the side of the hemispherical head end 3 near the first arc-shaped protrusions. The inner ring hinge 4 and the outer ring hinge 6 adjust the orientation of the hemispherical head end 3 through the first arc-shaped protrusions and the second arc-shaped protrusions.

[0026] Through the cooperation of the inner drive shaft 5, outer drive shaft 7, inner annular hinge 4, outer annular hinge 6, and the first and second arc-shaped protrusions, the orientation of the hemispherical head end 3 can be precisely adjusted, thereby enabling the laser ablation fiber 2 to be aligned with patches in different directions for ablation, improving the accuracy and flexibility of ablation. At the same time, the rotational cooperation between the inner annular hinge 4 and the outer annular hinge 6, as well as the setting of the protrusion structure, makes the steering control of the head end more stable and reliable.

[0027] In a further optimized design, a housing buckle 12 is provided at the end of the conduit housing 10 away from the hemispherical head end 3, and a housing groove 15 is provided at the head end of the pusher 13. The conduit housing 10 and the pusher 13 are connected through the housing buckle 12 and the housing groove 15. The head end of the power unit 16 is provided with multiple conduit interfaces 14. The conduit 8 is detachably installed on the power unit 16 through the conduit interfaces 14, and the mechanical feedback guide wire 9 is electrically connected to the controller through the conduit interfaces 14.

[0028] The outer shell buckle 12 and outer shell slot 15 work together to achieve a quick and stable connection between the catheter outer shell 10 and the pusher 13, which is convenient for installation before surgery and disassembly after surgery. The catheter interface 14 can be set to enable quick and detachable installation of the catheter 8 and the power device 16, while ensuring the electrical connection between the mechanical feedback guidewire 9 and the controller, which is convenient for signal transmission.

[0029] The design is further optimized so that the thruster 13 includes a thruster housing 131, a telescopic cylinder 132 is fixedly installed inside the thruster housing 131, and a power unit 16 is fixedly installed at the movable end of the telescopic cylinder 132. The extension and retraction of the telescopic cylinder 132 is used to push the guide tube 8 and the power unit 16 forward or backward.

[0030] The telescopic cylinder provides a stable and controllable driving force for the advancement and retraction of catheter 8, thereby enabling precise advancement and retraction of catheter 8 within the blood vessel and improving the automation and safety of the surgical procedure.

[0031] In a further optimized design, the power unit 16 includes a power unit housing 161. A first motor 162 is fixedly installed in the middle of the inner side of the power unit housing 161. A mounting shaft 163 is fixedly connected to the output shaft of the first motor 162. An inner shaft insertion plate 164 is fixedly sleeved on the mounting shaft 163. An inner drive shaft 5 is detachably installed on the inner shaft insertion plate 164. An outer shaft insertion plate 165 is rotatably sleeved on the outer side of the inner shaft insertion plate 164. A gear ring 166 is fixedly connected to the bottom of the outer shaft insertion plate 165. A second motor 167 is fixedly connected inside the power unit housing 161. A gear 168 is fixedly connected to the output shaft of the second motor 167. The gear 168 meshes with the gear ring 166. An outer drive shaft 7 is detachably installed on the outer shaft insertion plate 165.

[0032] Through the cooperation of the first motor 162, the inner shaft connector 164, the second motor 167, the gear 168, the gear ring 166, and the outer shaft connector 165, independent driving of the inner drive shaft 5 and the outer drive shaft 7 can be achieved, thereby providing two different power inputs for the steering of the hemispherical head end 3, further improving the flexibility and precision of steering control.

[0033] In use, forward drive is achieved by the propeller 13 working in conjunction with the power unit 16. The outer casing buckle 12 and the conduit casing 10 remain fixed and stationary relative to the propeller 13. The propeller 13 pushes the conduit 8 and the power unit 16 forward through an internal drive assembly, which can be an electric telescopic cylinder. Steering drive refers to the turning of the hemispherical head end 3 at the head of the conduit 8. The conduit 8 is connected to the power unit 16 through the conduit interface 14. The conduit casing 10 remains fixed. The power unit 16 drives the inner drive shaft 5 and the outer drive shaft 7 respectively, which in turn drive the inner annular hinge 4 and the outer annular hinge 6, further driving the rotation of the hemispherical head end 3, causing the laser confocal endoscope 1 and the laser ablation fiber 2 to turn at different angles. Reverse drive is the opposite of forward drive.

[0034] A robotic control system for intravascular plaque laser ablation, based on the aforementioned intravascular plaque laser ablation robot, includes: The sensing module includes a laser confocal endoscope 1 and a laser ablation fiber 2, which are used to identify the plaque type in the plaque area within the blood vessel and adjust the laser wavelength to ablate the plaque according to the plaque type. The propulsion module includes a thruster 13 and a power unit 16, which is used to control the motion state of the sensing module according to the magnitude of the provided thrust. The laser control module includes a laser ablation system, which is used to adjust the laser parameters in the sensing module according to the identified patch type and determine the irradiation range of the laser. The dynamic ablation planning module includes a controller and a mechanical feedback guide wire 9, which is used to dynamically plan the laser ablation sequence and path of the patch area based on the guide wire tension, patch distribution and location. The thermal management and safety control module is used to monitor the temperature of the eroded area in real time and ensure that the temperature is always within a safe range.

[0035] Through the systematic setup of sensing modules, propulsion modules, laser control modules, dynamic ablation planning modules, and thermal management and safety control modules, comprehensive intelligent control of the intravascular plaque ablation process can be achieved. From plaque identification, laser parameter adjustment, ablation path planning to safety monitoring, a complete closed-loop control system is formed, which greatly improves the success rate and safety of ablation surgery.

[0036] Further optimization of the scheme, including methods for the perception module to identify patch types, includes: S1.1 Feature extraction; High-resolution images of the vascular intima were obtained using laser confocal endoscopy 1, and the following structural indices were extracted: Endothelial continuity index : Measures whether there is tearing or interruption of the endothelium, with a range of values. Endothelial fluorescence signal is continuously represented by a value of 0, while discontinuous endothelial fluorescence signal indicates tearing or interruption and is represented by a value of 1; Endothelial continuity index The Sobel operator is used to obtain the image edge contours, and the continuity of the fluorescence signal is determined based on whether the contours are continuous. Surface roughness index : Calculated from image texture entropy, it is used to quantify the irregularity of the vascular intima surface. The higher the value, the rougher the intima surface, and the greater the risk of inflammatory erosion or plaque rupture; surface roughness index Calculate the surface roughness index based on the edge contour extracted by the Sobel operator. ,include: The grayscale image of the endometrium acquired by laser confocal endoscopy is quantized (quantization level N=64) to obtain a grayscale image with quantization level N. A grayscale co-occurrence matrix P is constructed. Let the image grayscale levels be i and j. Under the direction θ and pixel distance d, the elements of the normalized grayscale co-occurrence matrix are p(i,j). The direction θ takes 0°, 45°, 90°, and 135° as directions, and the pixel distance d takes 1 pixel. The texture entropy of the four directions is calculated respectively, and the arithmetic mean is taken as the final H. Surface roughness index The texture entropy is calculated using the following formula: in, , To avoid extremely small constants whose logarithmic arguments are zero (such as...) ); To eliminate the influence of differences in image acquisition conditions on the absolute entropy value, the texture entropy is normalized to the [0,1] interval, resulting in: in, This is the theoretical maximum entropy value; Reflectivity distribution characteristics By analyzing the reflectance changes in images obtained by laser confocal endoscopy 1, the lipid core and calcified particles can be identified. Morphological statistical characteristics This includes patch thickness and area parameters; calculation of morphological statistical characteristics. ,include: Divide the lumen region With plaque area And calibrate the physical side length s of a single pixel; Draw a ray at angle θ from the centroid O of the lumen to obtain the intersection points with the lumen boundary and the outer boundary of the patch. , Calculate radial thickness ; Take the maximum radial thickness and divided by the equivalent vessel diameter Obtain the normalized thickness ratio ; Calculate the patch area load ratio ; where N lumen N is the lumen area. plaque The area of ​​the patch; Calculate the Sigmoid mapping for the radial thickness variation coefficient ,in E idx The relative volatility after eliminating the influence of thickness dimensions (units); μ d σ is the arithmetic mean of the radial thickness data. d The overall standard deviation of the radial thickness data is used to reflect the absolute dispersion of the thickness at each measurement point relative to the mean. Constructing feature vectors .

[0037] Therefore, the structural feature vector is: ; After injection of ICG fluorescent agent, the fluorescence signal of the blood vessel wall was recorded at different time points, and the following key parameters were extracted: fluorescence intensity ratio : ; This indicates the fluorescence intensity 60 seconds after ICG fluorescent agent injection. This represents the initial maximum fluorescence intensity, expressed in AU. elution rate : ; Fluorescence change slope : The rate of change calculated from the fluorescence time curve; The fluorescence feature vector is: ; S1.2 Feature fusion; To balance the recognition value of both structural feature vectors and fluorescence feature vectors, a weighted fusion strategy is adopted: in, Indicates feature concatenation; weights , These represent the contribution weights of structural and fluorescence characteristics, respectively. The feature vector obtained after fusion ; S1.3, Classification; Fusion features The input is fed into a nonlinear support vector machine model to determine whether a given region is a patch; the model is defined as follows: in: The sum of the contributions of the support vectors of N samples. For the first The true class labels of each training sample correspond to the plaque region and the normal endothelial region, respectively; For Lagrange multipliers, It is a radial basis function (RBF) kernel. Indicates the first training set The fusion features of each sample For the bias term, output This represents the classification result output by the model at the end; S1.4 Result Interpretation and Output Control; The classification results are transmitted to the laser ablation system. If a patchy area is identified, the ablation wavelength and energy are adjusted, and targeted ablation is performed based on the fluorescence type. Medium-power, medium-wavelength lasers were selected for lipid plaques; High-energy, short-wavelength lasers are used to treat calcification lesions. Low-energy broadband lasers were used to treat the fiber patches. When the plaque structure is eroded to the point where the mechanical feedback is less than the minimum preset value of 0.3N, the conduit 8 enters the next region and restarts the detection-identification-erosion cycle.

[0038] By employing feature extraction, feature fusion, and classification steps, accurate identification of plaque types can be achieved, avoiding errors associated with single-image recognition. Simultaneously, through result interpretation and output control, optimal laser ablation parameters can be automatically matched according to different plaque types, realizing personalized and precise treatment.

[0039] To further optimize the scheme, the feature weights are allocated during feature fusion as follows: Structural feature contribution weight: 60%; Fluorescence feature contribution weight: 40%.

[0040] By allocating appropriate weights, the fundamental role of structural features in patch identification is preserved, while the advantages of fluorescence features in distinguishing patch activity and type are fully utilized, thereby improving the overall accuracy of identification.

[0041] Further optimization of the scheme and methods to advance the motion state control of the module sensing module include: S2.1 Obtain the current thrust of the power unit 16; During propulsion, the current of the power unit 16 is acquired in real time. With propulsion speed The current thrust is estimated based on the following linear model: in, The current propulsion thrust is expressed in N. The operating current of the current power unit 16 is in A, ranging from [0,10]. The linear propulsion speed of the current power unit 16 is in mm / s, ranging from [0, 0.5]. The coefficient 0.5 is the thrust constant (N / A), and the coefficient 0.1 is the viscous damping coefficient (N·s / mm). The negative correlation between F(t) and v(t) reflects the constant power drive characteristics of the motor under variable load conditions. When the propulsion resistance increases, the motor speed decreases and the current increases, which together lead to an increase in the estimated thrust. S2.2, Advance control process; S2.21, Start-up Initialization: Power unit 16 returns to zero position; guide wire tension sensor calibrated to... ; S2.22, Real-time detection of the status of power unit 16: Reading current and speed Substitute the values ​​into the inference model to calculate the instantaneous thrust. ; S2.23, Thrust Judgment and Feedback Control: If If the vascular resistance is considered low, continue proceeding; if If a patch is detected, the process pauses and patch recognition is activated; where F min =0.3N, F max =0.5N; S2.24, Advance / Pause / Laser Co-operation Switching: If advancement is allowed, continue linear advancement; if advancement is not allowed, after the co-laser confocal endoscope 1 is activated for identification, laser ablation is performed, and advancement resumes after the mechanical feedback decreases again.

[0042] By acquiring the current thrust of the power unit 16 and combining it with a preset threshold, the resistance of the catheter 8 in the blood vessel can be sensed in real time, thereby accurately determining whether plaque or vascular bend is encountered; through coordinated switching of advance / pause / laser, the automatic alternation of advance and ablation is realized, improving the automation and safety of the operation.

[0043] Further optimization of the scheme, including the control method of the laser control module, includes: S3.1 Low-power scanning: The endothelial region marked by laser confocal endoscope 1 is surface scanned using a 532nm low-power laser to excite autofluorescence to verify endothelial integrity; S3.2, Ablation Restricted Area Setting: When the ICG fluorescence intensity is within the normal endothelial range, laser emission in that area is automatically blocked. The normal endothelial range is F... IR Values ​​range from 0.8 to 1.2; S3.3, Patch-specific erosion removal: Lipid plaques: preferentially activated by the 2940nm erbium-doped yttrium aluminum garnet laser therapy device, energy density formula: E=40-0.1⋅F IR (mJ / cm) 2 ) When F IR Enabled when >2.0; where 40 is the reference energy density and 0.1 is F. IR Adjustment coefficient; Calcified plaques: The 308nm excimer laser is activated only when a WR > 80% is detected. Energy density formula: E=100+0.2⋅WR(mJ / cm 2 ) When F IR Enabled when <0.5; where 100 is the reference energy density and 0.2 is the WR adjustment coefficient; WR (Washout Ratio) is the fluorescence elution rate, obtained according to the calculation method of elution rate in S1.1, reflecting the rate at which the ICG fluorescent agent is eluted from the plaque area, with a value range of 0 to 100%; Fiber patch: A Z-type fiber combiner is used to achieve a dual-band laser coupling efficiency of more than 97%, with alternating output and a laser switching time of less than 1ms.

[0044] By using low-power scanning and setting ablation exclusion zones, accidental ablation of normal endothelial tissue can be effectively avoided, improving the safety of the procedure. Through plaque-specific ablation, the optimal laser wavelength and energy can be selected according to different plaque types, improving ablation efficiency and effectiveness.

[0045] Further optimization of the scheme includes methods for the dynamic ablation planning module to plan the laser ablation sequence and path, including: S4.1 Triggering conditions and graph construction; When the advancement of catheter tip 8 encounters resistance ≥0.5N, confocal laser endoscopy and ICG fluorescence mapping are triggered to acquire a plaque atlas of a segment of the intravascular region. Candidate regions for erosion are defined based on fluorescence thresholds. ; S4.2, Reinforcement Learning Model Construction; S4.21, Environment Definition; state space The current catheter tip is shown in the image. The position above ,coordinate , and local curvature Local plaque characteristics I t Information such as the previous action, etc. t-1 : Action space Laser direction selection dir i Laser power level j Should we proceed with the move? k : ; S4.22, Reward Function Design; Define reward function Used to guide path selection to achieve the optimal strategy: in, This indicates the erosion efficiency of the current patch area per unit time; Indicates the safety factor; η represents the laser energy consumption per unit time. t σ t ε t All values ​​are dimensionless values ​​after normalization (range 0-1), ηt The fluorescence intensity change rate before and after ablation is calculated as (fluorescence intensity before ablation - fluorescence intensity after ablation) / fluorescence intensity before ablation, σ. t Based on the mechanical feedback guidewire tension data, ε t The reward signal for reinforcement learning is calculated based on the product of laser power and irradiation time; each component is normalized and then weighted and summed. S4.3 Reinforcement Learning Algorithm Selection; Deep Q-Learning is used as the control core, and the network structure is as follows: Input layer: State vector Hidden layer: 2 fully connected layers, 128 nodes each; Output layer: Q-value for each action. Strategy Update: ; in, For learning rate, Discount factor; S4.4 Decision performance indicators; After the erosion area is planned, the laser control module automatically adjusts the following parameters: laser wavelength, laser power, and duration. If the fluorescence intensity of a certain area to be etched is continuously irradiated with laser, then the fluorescence intensity of that area is... Once the area is deemed to have completed erosion, the propulsion module is reactivated.

[0046] By using trigger conditions and map construction, scanning and planning procedures can be automatically initiated when patches are encountered; by constructing and applying reinforcement learning models, intelligent optimization of erosion paths and sequences can be achieved, avoiding the blindness and inaccuracy of manual planning; by setting decision performance indicators, the efficiency and effectiveness of the erosion process can be ensured.

[0047] Further optimization of the scheme, including the control methods for the thermal management and safety control module, includes: S5.1 Cooling control trigger mechanism: The real-time temperature monitoring point is located 1mm to 2mm in front of the catheter tip to measure tissue temperature, not the actual temperature at the contact point; when the detected temperature... Exceeding the critical temperature The cooling injection mechanism is activated; a temperature rise rate-driven injection model is adopted, and the injection intensity is controlled according to the heat accumulation rate. S5.2, Jet Flow Model: Jet velocity of coolant With current temperature Temperature difference exceeding the threshold The flow rate is directly proportional to the temperature, and increases linearly with increasing temperature. ; Where 0.05 is in mL / (s·℃), it means that when the temperature exceeds 60℃, the flow rate increases by 0.05 mL / s for every 1℃ increase; the value range of Q [0, 0.2] mL / s represents the physical feasible domain of the jet flow rate; when T > 64℃, the jet flow rate has reached the maximum cooling capacity of the system. At this time, the emergency withdrawal mechanism of S5.3 should be activated as a supplementary safety measure, rather than continuing to rely on cooling jets for temperature control; S5.3 Emergency withdrawal mechanism; S5.31, Triggering conditions and logical judgments; This mechanism is based on two core triggering conditions: Mechanical contact threshold: contact force ; Temperature safety threshold: ; The emergency rollback mechanism can be triggered if any one of the conditions is met. S5.32, Withdrawal Control Formula; Logical function for judgment: ; in, Real-time temperature; For real-time contact force; For triggering flag variables; Withdrawal control function: when The power unit 16 switches from propulsion mode to retraction mode, controlling the duct 8 to operate at a fixed speed. Reverse motion; Power unit 16 position control command: , Indicates the distance traveled; S5.33, Cooling Co-control; During the retraction, the cooling mechanism continues to operate to rapidly reduce tissue temperature: .

[0048] By using a cooling control trigger mechanism and a jet flow model, precise temperature control of the ablation zone can be achieved, avoiding thermal damage. With the emergency retraction mechanism in place, the catheter 8 can be automatically retracted in the event of abnormal mechanical contact or excessively high temperature, further improving the safety of the procedure and providing multiple safety guarantees.

[0049] The vascular plaque laser ablation robot and control system provided by this invention have the following overall implementation process: First, a catheter 8 is inserted into the patient's blood vessel and driven forward by a thruster 13 and a power unit 16. During the forward movement, a mechanical feedback guidewire 9 senses the force on the tip of the catheter 8 in real time. When encountering plaque that increases resistance, the controller controls the thruster 13 to pause the forward movement. Subsequently, a laser confocal endoscope 1 is activated to scan the area in front and performs fluorescence imaging using an ICG fluorescent agent to obtain a plaque map. The sensing module analyzes the image and fluorescence data to identify the plaque type. The dynamic ablation planning module... Based on the plaque distribution, the optimal ablation sequence and path are planned. The laser control module automatically adjusts the laser wavelength and energy according to the plaque type, and the plaque is precisely ablated through the laser ablation fiber 2. During the ablation process, the thermal management and safety control module monitors the temperature in real time and sprays coolant through the coolant inlet 11 to ensure that the temperature is within a safe range. When the plaque is ablated to a certain extent, the resistance detected by the mechanical feedback guidewire 9 decreases, and the controller controls the pusher 13 to continue advancing the catheter 8 into the next segment of the blood vessel, repeating the above detection-identification-ablation cycle until the plaque ablation treatment of the entire blood vessel segment is completed. Through the above process, this invention achieves automation, intelligence, and precision in intravascular plaque ablation, which can significantly reduce the risk of vascular perforation, shorten the time of a single ablation operation to about 20 minutes, and significantly improve the success rate and safety of the operation.

[0050] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A laser ablation robot for intravascular plaques, characterized in that, The device includes a catheter, the interior of which is embedded multiple mechanical feedback guidewires; one end of the catheter is provided with a hemispherical tip, and an elastic connecting ring is provided between the hemispherical tip and the catheter, the hemispherical tip being fixedly mounted on the catheter through the elastic connecting ring; A laser confocal endoscope and a laser ablation fiber are mounted on the hemispherical tip. The laser ablation fiber is arranged at equal intervals along the circumference outside the laser confocal endoscope. The end of the laser ablation fiber is connected to a laser ablation system. Both the laser confocal endoscope and the laser ablation fiber are inserted into the catheter. The laser confocal endoscope is used to identify the type of plaque in the plaque area within the blood vessel, and the laser ablation fiber is used for laser transmission and ablation of the plaque. The outer side of the conduit is fitted with a conduit shell, and the conduit shell is provided with a coolant injection port for injecting coolant. A pusher is installed at the end of the conduit shell away from the hemispherical head end. The pusher is equipped with a controller and a power device. The conduit is detachably installed on the power device. A transmission assembly is installed inside the conduit. The hemispherical head end is connected to the power device through the transmission assembly. The mechanical feedback guide wire, the pusher, and the power device are all electrically connected to the controller.

2. The intravascular plaque laser ablation robot according to claim 1, characterized in that, The transmission assembly includes an inner transmission shaft, one end of which is connected to the power device, and the other end of which is fixedly connected to an inner annular hinge; an outer transmission shaft is sleeved on the outside of the inner transmission shaft, one end of which is connected to the power device, and the other end of which is fixedly connected to an outer annular hinge, with the inner annular hinge and the outer annular hinge rotatably engaged; both the outer annular hinge and the outer transmission shaft are disposed within the guide tube; Both the inner and outer annular hinges have two first arc-shaped protrusions on the side near the hemispherical head end. The central angle between the two first arc-shaped protrusions on the same annular hinge is 135°, and at most one of the first arc-shaped protrusions on the inner annular hinge and the first arc-shaped protrusion on the outer annular hinge is opposite each other. Four second arc-shaped protrusions are equally spaced on the side of the hemispherical head end near the first arc-shaped protrusions. The inner and outer annular hinges adjust the orientation of the hemispherical head end through the first arc-shaped protrusions and the second arc-shaped protrusions.

3. The intravascular plaque laser ablation robot according to claim 1, characterized in that, The end of the conduit shell away from the hemispherical head end is provided with a shell buckle, and the head end of the pusher is provided with a shell groove. The conduit shell and the pusher are connected through the shell buckle and the shell groove. The head end of the power unit is provided with multiple conduit interfaces. The conduits are detachably installed on the power unit through the conduit interfaces, and the mechanical feedback guidewire is electrically connected to the controller through the conduit interfaces.

4. A control system for a laser ablation robot for intravascular plaques, based on the laser ablation robot for intravascular plaques according to any one of claims 1-3, characterized in that, include: The sensing module is used to identify the plaque type in the plaque area within the blood vessel and adjust the laser wavelength according to the plaque type to ablate the plaque; A propulsion module is used to control the motion state of the sensing module according to the magnitude of the provided thrust. The laser control module is used to adjust the laser parameters in the sensing module according to the identified patch type and to determine the laser irradiation range. The dynamic ablation planning module is used to dynamically plan the laser ablation sequence and path of the patch area based on the guide wire tension, patch distribution and location. The thermal management and safety control module is used to monitor the temperature of the eroded area in real time and ensure that the temperature is always within a safe range.

5. The robotic control system for intravascular plaque laser ablation according to claim 4, characterized in that, The method for the sensing module to identify patch types includes: S1.1 Feature extraction; High-resolution images of the vascular intima were obtained using laser confocal endoscopy, and the following structural indicators were extracted: Endothelial continuity index : Measures whether there is tearing or interruption of the endothelium, with a range of values. The endothelial fluorescence signal is continuously 0, and the discontinuous endothelial fluorescence signal is 1, indicating tearing or interruption. Surface roughness index : Calculated from image texture entropy; Reflectivity distribution characteristics By analyzing the reflectance changes in images obtained by laser confocal endoscopy, the lipid core and calcified particles can be identified. Morphological statistical characteristics This includes parameters such as patch thickness and area. Therefore, the structural feature vector is: ; After injection of fluorescein sodium (ICG) fluorescent agent, the fluorescence signal of the blood vessel wall was recorded at different time points, and the following key parameters were extracted: fluorescence intensity ratio : ; This indicates the fluorescence intensity 60 seconds after ICG fluorescent agent injection. This represents the initial maximum fluorescence intensity, expressed in AU. elution rate : ; Fluorescence change slope : The rate of change calculated from the fluorescence time curve; The fluorescence feature vector is: ; S1.2 Feature fusion; To balance the recognition value of both structural feature vectors and fluorescence feature vectors, a weighted fusion strategy is adopted: in, Indicates feature concatenation; weights , These represent the contribution weights of structural and fluorescence characteristics, respectively. The feature vector obtained after fusion ; S1.3, Classification; Fusion features The input is fed into a nonlinear support vector machine model to determine whether a given region is a patch; the model is defined as follows: in: For the first The true class labels of each training sample correspond to the plaque region and the normal endothelial region, respectively; For Lagrange multipliers, It is the radial basis function kernel. Indicates the first training set The fusion features of each sample For the bias term, output This represents the classification result output by the model at the end; S1.4 Result Interpretation and Output Control; The classification results are transmitted to the laser ablation system. If a patchy area is identified, the ablation wavelength and energy are adjusted, and targeted ablation is performed based on the fluorescence type. Medium-power, medium-wavelength lasers were selected for lipid plaques; High-energy, short-wavelength lasers are used to treat calcification lesions. Low-energy broadband lasers were used to treat the fiber patches. When the plaque structure is eroded to the point where the mechanical feedback is less than the minimum preset value, the conduit enters the next region and restarts the detection-identification-erosion cycle.

6. The robotic control system for laser ablation of intravascular plaques according to claim 5, characterized in that, In the feature fusion process, the feature weights are assigned as follows: Structural feature contribution weight: 60%; Fluorescence feature contribution weight: 40%.

7. The robotic control system for intravascular plaque laser ablation according to claim 4, characterized in that, The method for the propulsion module to control the motion state of the sensing module includes: S2.1 Obtain the current thrust of the power unit; During propulsion, the current of the power unit is acquired in real time. With propulsion speed The current thrust is estimated based on the following linear model: in, The current propulsion thrust is expressed in N. The current operating current of the current power unit is expressed in amperes (A) and ranges from [0,10]. The linear propulsion speed of the current power unit is in mm / s, ranging from [0, 0.5]. The coefficient 0.5 is the thrust constant (N / A), and the coefficient 0.1 is the viscous damping coefficient (N·s / mm). The negative correlation between F(t) and v(t) reflects the constant power drive characteristics of the motor under variable load conditions. When the propulsion resistance increases, the motor speed decreases and the current increases, which together lead to an increase in the estimated thrust. S2.2, Advance control process; S2.21, Start-up Initialization: Power unit returns to zero position; guide wire tension sensor calibrated to... ; S2.

22. Real-time monitoring of the power unit's status: reading the current. and speed Substitute into the inference model to calculate the instantaneous thrust. ; S2.23, Thrust Judgment and Feedback Control: If If the vascular resistance is considered low, continue proceeding; if If a patch is detected, the process pauses and patch recognition is activated; where F min =0.3N, F max =0.5N; S2.24, Advance / Pause / Laser Co-operation Switching: If advancement is allowed, continue linear advancement; if advancement is not allowed, initiate co-laser confocal endoscope recognition, perform laser ablation, and resume advancement after the mechanical feedback decreases again.

8. The robotic control system for laser ablation of intravascular plaques according to claim 4, characterized in that, The control method of the laser control module includes: S3.1 Low-power scanning: The endothelial region marked by laser confocal endoscopy is surface scanned using a 532nm low-power laser to excite autofluorescence to verify endothelial integrity; S3.2, Ablation Restricted Area Setting: When the ICG fluorescence intensity is within the normal endothelial range, laser emission in that area is automatically blocked. The normal endothelial range is F... IR Values ​​range from 0.8 to 1.2; S3.3, Patch-specific erosion removal: Lipid plaques: preferentially activated by the 2940nm erbium-doped yttrium aluminum garnet laser therapy device, energy density formula: E=40-0.1⋅F IR (mJ / cm 2 ) When F IR Enabled when >2.0; where 40 is the reference energy density and 0.1 is F. IR Adjustment coefficient; Calcified plaques: The 308nm excimer laser is activated only when a WR > 80% is detected. Energy density formula: E=100+0.2⋅WR(mJ / cm 2 ) When F IR Enabled when <0.5; where 100 is the reference energy density, 0.2 is the WR adjustment coefficient; WR is the fluorescence elution rate, obtained according to the calculation method of elution rate in S1.1, reflecting the rate of elution of ICG fluorescent agent from the plaque area, with a value range of 0 to 100%; Fiber patch: A Z-type fiber combiner is used to achieve a dual-band laser coupling efficiency of more than 97%, with alternating output and a laser switching time of less than 1ms.

9. The robotic control system for intravascular plaque laser ablation according to claim 4, characterized in that, The method for planning the laser ablation sequence and path by the dynamic ablation planning module includes: S4.1 Triggering conditions and graph construction; When the catheter tip encounters resistance (≥0.5N), confocal laser endoscopy and ICG fluorescence mapping are triggered to acquire a plaque map of a segment of the vessel. Candidate regions for erosion are defined based on fluorescence thresholds. ; S4.2, Reinforcement Learning Model Construction; S4.21, Environment Definition; state space The current catheter tip is shown in the image. The position above ,coordinate , and local curvature Local plaque characteristics I t Information such as the previous action, etc. t-1 : Action space Laser direction selection dir i Laser power level j Should we proceed with the move? k : ； S4.22, Reward Function Design; Define reward function Used to guide path selection to achieve the optimal strategy: in, This indicates the erosion efficiency of the current patch area per unit time; Indicates the safety factor; η represents the laser energy consumption per unit time. t σ t ε t All values ​​are dimensionless values ​​after normalization, η t σ is calculated based on the rate of change in fluorescence intensity before and after erosion. t Based on the mechanical feedback guidewire tension data, ε t The reward signal for reinforcement learning is calculated based on the product of laser power and irradiation time; each component is normalized and then weighted and summed. S4.3 Reinforcement Learning Algorithm Selection; Deep Q-Learning is used as the control core, and the network structure is as follows: Input layer: State vector Hidden layer: 2 fully connected layers, 128 nodes each; Output layer: Q-value for each action. Strategy Update: ; in, For learning rate, Discount factor; S4.4 Decision performance indicators; After the erosion area is planned, the laser control module automatically adjusts the following parameters: laser wavelength, laser power, and duration. If the fluorescence intensity of a certain area to be etched is continuously irradiated with laser, then the fluorescence intensity of that area is... Once the area is deemed to have completed erosion, the propulsion module is reactivated.

10. The robotic control system for intravascular plaque laser ablation according to claim 4, characterized in that, The control method of the thermal management and safety control module includes: S5.1 Cooling control trigger mechanism: The real-time temperature monitoring point is located 1mm to 2mm in front of the catheter tip; when the detected temperature... Exceeding the critical temperature The cooling injection mechanism is activated; a temperature rise rate-driven injection model is adopted, and the injection intensity is controlled according to the heat accumulation rate. S5.2, Jet Flow Model: Jet velocity of coolant With current temperature Temperature difference exceeding the threshold The flow rate is directly proportional to the temperature, and increases linearly with increasing temperature. ； Where 0.05 is in mL / (s·℃), it means that when the temperature exceeds 60℃, the flow rate increases by 0.05 mL / s for every 1℃ increase; the value range of Q [0, 0.2] mL / s represents the physical feasible domain of the jet flow rate; when T > 64℃, the jet flow rate has reached the maximum cooling capacity of the system. At this time, the emergency withdrawal mechanism of S5.3 should be activated as a supplementary safety measure, rather than continuing to rely on cooling jets for temperature control; S5.3 Emergency withdrawal mechanism; S5.31, Triggering conditions and logical judgments; This mechanism is based on two core triggering conditions: Mechanical contact threshold: contact force ; Temperature safety threshold: ; The emergency rollback mechanism can be triggered if any one of the conditions is met. S5.32, Withdrawal Control Formula; Logical function for judgment: ; in, Real-time temperature; For real-time contact force; For triggering flag variables; Withdrawal control function: when The power unit switches from propulsion mode to retraction mode, controlling the duct to operate at a fixed speed. Reverse motion; Power unit position control commands: , Indicates the distance traveled; S5.33, Cooling Co-control; During the retraction, the cooling mechanism continues to operate to rapidly reduce tissue temperature: 。

Images