Adjustable balloon expansion pressure pump and intelligent control method thereof

Abstract

The invention relates to the technical field of minimally invasive vascular interventional operations, in particular to an adjustable balloon expansion pressure pump and an intelligent control method thereof.The method comprises the steps that a real-time image of a connecting pipe is obtained based on perspective imaging equipment, positioning information of a marker in the connecting pipe is recognized, and an initial parameter set is generated; constructing a fluid model according to the initial parameter set, and generating a fluid parameter set by taking distance information between the first identification body and the second identification body as a constraint condition; controlling pump body output pressure according to the fluid parameter set to generate a pre-expansion parameter set; the fluid model is updated based on the pre-expansion parameter set, the pump body is controlled to execute a pressure instruction in the dynamic pressure sequence, a second change value of the distance information is obtained, and a regulation and control data set is generated according to the second change value; and optimizing the regulation and control data set, and controlling the output pressure of the pump body to a target value until the balloon is completely opened. The success rate and safety of balloon dilatation are remarkably improved, and the balloon dilatation process is more controllable and efficient.

Description

Technical Field

[0001] The present invention relates to the technical field of minimally invasive vascular intervention surgery. Specifically, it relates to an adjustable balloon expansion pressure pump and its intelligent control method. Background Art

[0002] In the medical field, balloon expansion technology is applied in vascular intervention therapy, heart valve repair, and other medical scenarios that require local dilation. Existing balloon expansion devices control the inflation and dilation process of the balloon through manual or semi-automatic pressure pumps. Doctors need to rely on experience to adjust the pressure to ensure appropriate dilation effect of the balloon at the target site.

[0003] However, existing pressure pump control methods generally have a core technical drawback, that is, the lack of precise perception and real-time regulation ability for the dynamic changes during the balloon dilation process. This deficiency often makes it difficult for doctors to accurately judge the actual dilation state of the balloon during the operation, which may damage the blood vessel wall due to excessive pressure, or fail to achieve the expected treatment effect due to insufficient pressure. In addition, existing control methods usually adopt a fixed pressure increment mode, ignoring the possible non-linear changes during the dilation process, limiting the adaptability of the device in complex surgical scenarios, not only increasing the surgical risk, but also possibly prolonging the operation time, affecting the treatment effect and postoperative recovery of patients.

[0004] Therefore, there is an urgent need for a new control method that can overcome the limitations of the existing technology in dynamic perception and precise regulation, so as to improve the safety and treatment effect of the balloon expansion device. Summary of the Invention

[0005] The main object of the present invention is to provide an intelligent control method for an adjustable balloon expansion pressure pump, aiming to overcome the technical problem that the existing technology lacks the precise perception and real-time regulation ability for the dynamic changes during the balloon dilation process.

[0006] To solve the above problems of the invention, the present invention proposes an intelligent control method for an adjustable balloon expansion pressure pump, the method comprising: Obtaining a real-time image of the connecting tube based on a fluoroscopic imaging device, identifying the positioning information of the markers in the connecting tube, and generating an initial parameter set, wherein the markers include a first marker body and a second marker body; Constructing a fluid model according to the initial parameter set, using the distance information between the first marker body and the second marker body as a constraint condition of the fluid model, and generating a fluid parameter set; Controlling the pump body to output pressure to an initial value according to the fluid parameter set, calculating a pressure deviation according to a first change value of the distance information, and generating a pre-expansion parameter set; Update the fluid model based on the pre-expansion parameter set and output a dynamic pressure sequence; Control the pump body to execute the pressure commands in the dynamic pressure sequence, obtain a second change value of the distance information, and generate a regulation data set according to the second change value; Optimize the regulation data set, control the pump body to output pressure to the target value until the balloon is fully opened.

[0007] Further, the step of obtaining a real-time image of the connecting tube based on the fluoroscopic imaging device, identifying the positioning information of the markers in the connecting tube, and generating an initial parameter set includes: Preprocess the original image data collected by the fluoroscopic imaging device to obtain a standardized image set; Identify the regions containing the first marker body and the second marker body in the standardized image set and perform segmentation processing to obtain a marker region image; Identify the positioning information of the first marker body and the second marker body in the marker region image and generate an initial parameter set.

[0008] Further, the step of constructing a fluid model according to the initial parameter set, using the distance information between the first marker body and the second marker body as a constraint condition of the fluid model, and generating a fluid parameter set includes: Perform geometric constraint extraction processing on the connecting tube image data according to the distance information between the first marker body and the second marker body in the initial parameter set to obtain a connecting tube geometric feature set; Model the initial state of the fluid according to the distance constraint and inner diameter distribution in the connecting tube geometric feature set to generate a fluid model; Analyze the pressure gradient of the fluid in the connecting tube based on the fluid model to obtain a pressure distribution parameter set; Modify the fluid model according to the pressure distribution parameter set and output to obtain a fluid parameter set.

[0009] Further, the step of controlling the pump body to output pressure to the initial value according to the fluid parameter set, calculating a pressure deviation according to the first change value of the distance information, and generating a pre-expansion parameter set includes: Perform grid decomposition on the fluid parameter set to generate an initial pressure distribution parameter including the pressure values of each grid node; Pre-adjust the pump body drive signal according to the initial pressure distribution parameter to obtain an initial pressure control command; Control the pump body to drive the fluid into the connecting tube according to the initial pressure control command and obtain real-time fluid state data, where the real-time fluid state data includes the first change value of the distance information between the first marker body and the second marker body; Calculate the deformation degree of the balloon according to the first change value, and calculate the actual pressure distribution inside the balloon according to the deformation degree to obtain a first pressure deviation; Optimize the initial pressure control instruction according to the first pressure deviation to obtain a corrected pressure control sequence; Integrate the corrected pressure control sequence with the convergence trend of the first pressure deviation to generate a pre-dilation parameter set including corrected initial pressure, actual flow rate, and deformation characteristics.

[0010] Further, the step of updating the fluid model based on the pre-dilation parameter set and outputting a dynamic pressure sequence includes: Extract the state characteristics of the real-time blood vessel image obtained by the fluoroscopic imaging device to obtain a blood vessel state characteristic set; Perform stress distribution reconstruction processing on the initial fluid model according to the blood vessel state characteristic set to obtain an updated stress distribution model; Predict the flow rate of the stress distribution model to obtain a flow rate prediction sequence; According to the flow rate values and time nodes in the flow rate prediction sequence, perform pressure curve optimization processing on the initial pressure value in the pre-dilation parameter set to obtain a preliminary pressure sequence; Perform time series integration processing on the preliminary pressure sequence for the fluid model to obtain a dynamic pressure sequence.

[0011] Further, the step of predicting the flow rate of the stress distribution model to obtain a flow rate prediction sequence further includes: Obtain the shear stress and radial stress of each calculation unit in the stress distribution model, and calculate the stress distribution characteristics of each unit through a finite element analysis algorithm to obtain a stress distribution characteristic set; Combine the stress distribution characteristic set with the initial flow rate value in the pre-dilation parameter set to simulate the flow rate change trend of the fluid in the connecting tube and blood vessel, and generate a flow rate prediction sequence.

[0012] Further, the step of controlling the pump body to execute the pressure instruction in the dynamic pressure sequence and obtaining a second change value of the distance information, and generating a regulation data set according to the second change value includes: Control the pump body to execute the pressure instruction in the dynamic pressure sequence, and obtain the actual pressure change in the balloon to generate an actual pressure feedback set; Construct a sliding window with a preset number of digits, and perform smoothing processing on the actual pressure feedback set based on the sliding window to obtain a smoothed feedback signal set; Compare the pressure values in the smoothed feedback signal set with the command pressure values in the initial execution instruction set point by point, and calculate the deviation value between the two; Weight the deviation value with the second change value between the first identification body and the second identification body to generate a pressure deviation set including the deviation value, time series, and distance change; Integrate the actual pressure feedback set and the pressure deviation set, and analyze the future short-term pressure change trend through time series prediction technology to generate a real-time regulation data set.

[0013] Further, the step of optimizing the regulation data set and controlling the pump body to output pressure to the target value until the balloon is fully opened includes: Fuse the time series, deviation value, and distance change information in the regulation data set to obtain a comprehensive regulation parameter set; Iteratively optimize the comprehensive regulation parameter set to generate an optimized pressure output instruction sequence; Control the pump body to output pressure according to the optimized pressure output instruction sequence until the actual pressure inside the balloon reaches the preset target pressure value; Monitor the third change value of the distance between the first identification body and the second identification body. When the third change value reaches the preset stable threshold, it is determined that the balloon is fully opened; Stop the pressure output of the pump body and output a status signal indicating that the balloon is fully opened.

[0014] This application also discloses an adjustable balloon expansion pressure pump applied to the intelligent control method described in any one of the above. The adjustable balloon expansion pressure pump includes: A pump body; A connection cavity, which is communicated with the pump body; An outer tube, one end of which extends into the connection cavity and is connected to the output port of the pump body, and the other end is connected to the balloon through a connection tube to transmit the pressure generated by the pump body to the balloon. A first identifier and a second identifier are arranged in the connection tube; A fluid storage tank, which is connected to the connection cavity and is used to store a fluid medium for the pump body to extract and pressurize and output.

[0015] Further, the first identifier and the second identifier are arranged at one end of the connection tube far from the connection cavity. The first identifier is provided with a reflective groove, and an elastic member is arranged on the side of the first identifier close to the outer tube, and the elastic member is arranged in the connection tube.

[0016] Advantageous effects: An intelligent control method for an adjustable balloon expansion pressure pump proposed in this application uses a fluoroscopic imaging device to obtain real-time images of the connecting tube, generates an initial parameter set by identifying the positioning information of the first marker and the second marker, and captures minute changes during the balloon expansion process in real time, thereby greatly improving the sensing accuracy. By constructing a fluid model with the distance information between the markers as a constraint condition and generating a fluid parameter set, it realizes the accurate simulation of fluid behavior in a complex physiological environment, overcomes the drawback of the fixed pressure increment mode in the prior art that ignores non-linear changes, and enables the pressure control to better adapt to individual patient differences and dynamic changes in the vascular environment. Calculate the pressure deviation according to the first change value of the distance information and generate a pre-expansion parameter set. Combining the output of the dynamic pressure sequence and the optimization of the regulation data set, continuously update the model and finely adjust the pressure output during the expansion process until the balloon is fully opened. This closed-loop feedback mechanism ensures the real-time and accuracy of pressure regulation, effectively avoiding problems such as vascular damage caused by excessive pressure or incomplete expansion caused by insufficient pressure. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 FIG. is a schematic diagram of the overall steps of an intelligent control method for an adjustable balloon expansion pressure pump in an embodiment of the present invention; Figure 2 FIG. is a schematic diagram of the overall structure of an adjustable balloon expansion pressure pump in an embodiment of the present invention; Figure 3 FIG. is a schematic cross-sectional side view structure diagram of an adjustable balloon expansion pressure pump in an embodiment of the present invention; Figure 4 is an embodiment of the present invention Figure 3 The enlarged structure diagram at position A in.

[0018] Wherein, the reference numerals are: 1. Pump body; 2. Connecting cavity; 3. Outer tube; 4. Connecting tube; 5. First marker; 51. Reflective groove; 6. Second marker; 7. Elastic member; 8. Fluid storage tank.

[0019] The realization, functional features and advantages of the object of the present invention will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] In order to make the object, technical solution and advantages of this application clearer, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not used to limit this application.

[0021] Those skilled in the art can understand that, unless specifically stated otherwise, the singular forms "a", "an", "the above" and "the" used herein may also include the plural forms. It should be further understood that the term "comprising" used in the specification of the present invention means the presence of features, integers, steps, operations, elements, modules and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, modules, components and / or their groups. It should be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to other elements, or there may also be intermediate elements. In addition, the "connection" or "coupling" used herein may include wireless connection or wireless coupling. The phrase "and / or" used herein includes all or any one of the modules and all combinations of one or more related listed items.

[0022] Those skilled in the art can understand that, unless otherwise defined, all terms used herein (including technical terms and scientific terms) have the same meaning as the general understanding of those of ordinary skill in the field to which the present invention belongs. It should also be understood that terms such as those defined in a general dictionary should be understood to have a meaning consistent with the meaning in the context of the prior art, and will not be interpreted with an idealized or overly formal meaning unless specifically defined as here.

[0023] Referring to Figure 1 , an embodiment of the present invention provides an intelligent control method for an adjustable balloon inflation pressure pump, and the method includes: S1: Obtain a real-time image of a connecting tube based on a fluoroscopic imaging device, identify the positioning information of the markers in the connecting tube, and generate an initial parameter set, wherein the markers include a first marker body and a second marker body; In step S1, a dynamic image of the blood vessel and the connecting tube is captured in real time by an X-ray fluoroscopy imaging device to generate high-resolution two-dimensional or three-dimensional image data. These fluoroscopic images are subjected to deep learning analysis and processing. Specifically, a convolutional neural network (CNN) algorithm is used to segment and extract features from the images. The connecting tube is embedded with a first identification body and a second identification body, and these two identification bodies can be cylindrical structures with specific optical features. Through layer-by-layer convolution and pooling operations on the images, the convolutional neural network can segment the contours of these identification bodies and extract the optical features of their through holes, such as the contrast at the edges or the uniqueness of the geometric shape. After identifying the identification bodies, the spatial positions and orientations of the first identification body and the second identification body are calculated to generate positioning data in a three-dimensional coordinate system. For example, the X-ray device captures multiple frames of images from different angles, and the coordinates of the identification bodies in the three-dimensional space are determined by the triangulation method. Assuming that the first identification body is located at coordinates (x1, y1, z1) and the second identification body is located at (x2, y2, z2), these coordinate data can directly reflect the relative positions and distances between the identification bodies, and the geometric features between the identification bodies are analyzed, such as the gap size between the first identification body and the second identification body, the relative distance between the connecting tube and the blood vessel wall is estimated, and the resistance coefficient of the fluid passing through the connecting tube is calculated in combination with the principle of fluid mechanics. All the analysis results are integrated to generate an initial parameter set containing information such as the position coordinates of the first identification body and the second identification body, the blood vessel diameter, and the resistance coefficient.

[0024] S2: Construct a fluid model based on the initial parameter set, use the distance information between the first identification body and the second identification body as a constraint condition of the fluid model, and generate a fluid parameter set; In step S2, based on the initial parameter set, a hydrodynamic model of the connecting tube and the balloon is constructed within the control unit. The construction process of this model uses the basic principles of hydrodynamics, such as the Navier-Stokes equations, which describe the velocity distribution and pressure gradient of fluids in complex geometric environments. In the connecting tube and balloon system, the fluid is a liquid (such as normal saline), and its movement is jointly affected by the geometric shape of the tube wall, fluid viscosity, and external pressure. Through the Navier-Stokes equations, the velocity distribution of the fluid in the connecting tube can be analyzed, and then the pressure gradient can be deduced. When constructing the fluid model, the distance information between the first identification body and the second identification body is used as a constraint condition. By incorporating the distance information into the fluid model, the boundary conditions can be dynamically adjusted in the simulation, thus more realistically reflecting the behavior of the fluid in the connecting tube and the balloon. Specifically, the distance information can be used to correct the geometric parameters in the model, such as the equivalent length or cross-sectional area of the connecting tube, and then optimize the calculation results of the fluid flow rate. During the calculation of the fluid parameters, the characteristics of the blood vessel are combined to further improve the model. Specifically, the elastic parameters of the blood vessel wall can be estimated through the blood vessel diameter, and the boundary conditions can be dynamically adjusted in the fluid model, thus more accurately simulating the mechanical response of the blood vessel. This process can be combined with the finite element analysis method to calculate the deformation amount of the blood vessel wall under different pressures through numerical simulation, and then deduce the interaction force between the fluid and the blood vessel wall. In another embodiment, the fluid flow in the connecting tube and balloon system is not completely laminar, especially turbulence may occur at the through-hole. The finite element analysis method is used to numerically simulate the turbulence characteristics at the through-hole. The geometric model of the connecting tube and the balloon is discretized into a finite number of mesh elements, and the velocity and pressure distributions of the fluid are solved in each element. By analyzing the turbulence characteristics, a local resistance correction factor can be generated to correct the resistance parameters in the fluid model and improve the accuracy of the model. After completing the above modeling and analysis, a fluid parameter set including predicted pressure, flow rate, and resistance correction factor is generated.

[0025] S3: Control the pump body to output pressure to the initial value according to the fluid parameter set, calculate the pressure deviation according to the first change value of the distance information, and generate a pre-expansion parameter set; In step S3, according to the predicted pressure in the fluid parameter set, the output pressure of the pump body is adjusted to the initial value. The initial value is the starting point of the balloon dilation process and determines the initial driving force for injecting fluid into the balloon. The fluid (such as normal saline) is injected into the balloon in a controllable manner through the pump body. After the initial pressure is set and the fluid injection starts, the distance change between the first marker body and the second marker body during the balloon dilation process is monitored. The distance change reflects the interaction between the internal pressure of the balloon and the deformation of the blood vessel wall. By analyzing the relative position change of the marker bodies in the image, the geometric deformation of the balloon during the dilation process is deduced. Based on the gap deformation characteristics, the actual pressure inside the balloon is calculated. The deformation characteristics are combined with the mechanical parameters in the fluid model to calculate the difference between the actual pressure and the initially set pressure. A pressure deviation is generated by comparing the actual pressure with the initial pressure, and the output pressure of the pump body is dynamically adjusted using the proportional-integral-derivative (PID) control algorithm. The PID algorithm is a classic feedback control method that can quickly respond to deviations through the proportional term, eliminate steady-state errors through the integral term, and predict the change trend of deviations through the derivative term to achieve precise control of the pump body output. Specifically, the proportional term directly adjusts the output pressure according to the magnitude of the pressure deviation, the integral term accumulates the deviation to eliminate long-term errors, and the derivative term adjusts the pressure in advance according to the change rate of the deviation to prevent system overshoot or oscillation. Through the PID algorithm, the initial pressure can be dynamically corrected according to the real-time deviation to generate a corrected initial pressure that better meets the actual requirements. At the same time, the change of the actual flow rate is recorded to form a pre-dilation parameter set containing the corrected initial pressure and the actual flow rate.

[0026] S4: Update the fluid model based on the pre-dilation parameter set and output a dynamic pressure sequence; In step S4, the image data of the blood vessel and the connecting tube obtained by the real-time fluoroscopic imaging device is used to further analyze the deformation characteristics of the blood vessel. Specifically, the fluoroscopic image can capture the geometric changes of the blood vessel under the action of the balloon pressure, such as the radial expansion or local bending of the blood vessel wall. Through image processing techniques, the local stress distribution and elastic modulus of the blood vessel wall can be identified. The local stress distribution reflects the pressure intensity received by the blood vessel wall in different regions, and the elastic modulus characterizes the material properties of the blood vessel wall. When updating the fluid mechanics model, the corrected initial pressure and the actual flow rate in the pre-dilation parameter set are used as inputs, and the distance change between the first identification body and the second identification body is used as a constraint condition. By monitoring the distance change in real time, the boundary conditions in the model can be dynamically adjusted, thereby generating new predicted values of pressure and flow rate, while taking into account the influence of the local stress distribution and elastic modulus of the blood vessel wall. Specifically, the Deep Deterministic Policy Gradient (DDPG) reinforcement learning algorithm is applied. This algorithm aims at stress equilibrium and learns the pressure curves under different blood vessel states through multiple rounds of iteration. The DDPG algorithm takes the current state of the blood vessel (such as stress distribution, elastic modulus, distance change, etc.) as input, predicts the next pressure adjustment action, and evaluates the effect of this action according to the blood vessel deformation feedback. In each round of iteration, the algorithm calibrates the pressure value according to the change in the gap between the cylinder and the small cylinder feedback by the real-time fluoroscopic image. For example, if the gap change indicates that the balloon expansion speed is too fast, the algorithm may reduce the pressure increment to slow down the expansion; conversely, if the gap change is small, the pressure is increased to accelerate the expansion. Through multiple rounds of learning, a dynamic pressure sequence containing multiple time nodes is gradually constructed, and each node corresponds to a specific pressure command.

[0027] S5: Control the pump body to execute the pressure command in the dynamic pressure sequence, obtain the second change value of the distance information, and generate a regulation data set according to the second change value; In step S5, the pump body is controlled to execute the pressure commands in the dynamic pressure sequence to guide the balloon to gradually expand to the target state. Specifically, through high-frequency sampling technology, commands are sent to the pump body at extremely short time intervals to ensure the real-time and accuracy of pressure output. The change in the gap between the first marker and the second marker (i.e., the second change value) is continuously monitored by a real-time fluoroscopic imaging device to sense the possible non-linear response during the balloon expansion process and accurately calculate the actual pressure of the balloon. The Kalman filter algorithm can be used to suppress the noise of the actual pressure signal. By establishing a state space model and combining the previous pressure estimate and the current measurement data, a smooth pressure feedback signal is generated. The smooth pressure feedback signal is compared with the commanded pressure in the dynamic pressure sequence, and the deviation between the two is calculated. The deviation reflects the gap between the current pressure and the target pressure. Based on the magnitude of the pressure deviation, the drive signal of the pump body is adjusted in real time, such as increasing or decreasing the fluid injection rate, so that the actual pressure gradually approaches the commanded pressure. At the same time, the output frequency of the fluid storage tank is optimized to ensure the stability of the flow rate. The stability of the flow rate is crucial for balloon expansion because the fluctuation of the flow rate may lead to uneven pressure distribution and even uneven balloon expansion. The trend of short-term future pressure fluctuations can be predicted through time series analysis. The historical pressure data and the current system state can be used to construct a prediction model to estimate the possible pressure changes during the balloon expansion process, so as to adjust and correct the pressure sequence in advance. The prediction result is combined with the real-time pressure deviation analysis to form a real-time regulation data set including the corrected pressure sequence, flow rate sequence and adjustment data.

[0028] S6: Optimize the regulation data set, and control the pump body to output pressure to the target value until the balloon is fully opened.

[0029] In step S6, the regulatory data set is analyzed to extract key dynamic response features. By combining the real-time images obtained by the fluoroscopic imaging device, the geometric changes after vasodilation are analyzed, such as the increment of blood vessel diameter, the change in the gap between the balloon and the blood vessel wall, and the uniformity of stress distribution. The resistance correction factor and pressure regulation parameters in the hydrodynamic model are iteratively adjusted through the Bayesian optimization algorithm. The Bayesian optimization algorithm is a global optimization method based on a probability model. In this embodiment, the goal is to find a set of pressure and flow rate values such that the balloon can be fully opened in an optimal manner in the current blood vessel environment, while ensuring uniform pressure distribution and avoiding excessive stress on the blood vessel wall. Specifically, using the corrected pressure sequence and flow rate sequence in the regulatory data set as the initial input, combined with the dynamic response features of vasodilation (such as the diameter change rate, local peaks of stress distribution, etc.), a probability model is constructed to predict the balloon dilation effect under different combinations of pressure and flow rate. Through multiple iterations, the algorithm will gradually converge to a set of optimal parameter values, which can maximize the stability and uniformity of balloon opening while minimizing potential damage to the blood vessel. After generating the optimized parameters, these parameters are converted into specific control instructions for the pump body, that is, the optimized pressure and flow rate values. These values are used to drive the fluid injection into the balloon, thereby promoting further balloon dilation until it is fully opened. It should be noted that being fully opened here does not only mean that the balloon reaches the expected geometric shape, but also includes that its fit with the blood vessel wall reaches an ideal state, and the pressure distribution remains uniform throughout the contact area. To achieve this goal, the interaction state between the balloon and the blood vessel is monitored in real time, and the distance change value between the first marker and the second marker is obtained. If the gap change shows that the dilation speed of a certain part of the balloon is too fast or too slow, the output pressure or flow rate of the pump body is fine-tuned to correct the local pressure distribution, so as to ensure the smooth and uniform opening process of the whole balloon, generating a balloon opening parameter set including optimized pressure, flow rate, and blood vessel response features. In actual execution, the pump body will continue to operate according to the optimized pressure and flow rate values until the monitoring data indicates that the balloon has been fully opened. At this time, the state of the balloon is verified through the fluoroscopic imaging device, such as confirming whether it is fully attached to the blood vessel wall and whether there is local stress concentration. If deviations are found, a new round of fine-tuning will be triggered until all indicators meet the requirements.

[0030] In one embodiment, the step of obtaining the real-time image of the connecting tube by the fluoroscopic imaging device and identifying the positioning information of the marker in the connecting tube to generate the initial parameter set includes: Preprocess the original image data collected by the fluoroscopic imaging device to obtain a standardized image set; Identify the area containing the first marker and the second marker in the standardized image set and perform segmentation processing to obtain the marker area image; Identify the first marker and the second marker location information in the marker area image and generate an initial parameter set.

[0031] In this embodiment, the original image sequence containing the connecting tube and its internal markers is captured in real time by X-ray or other imaging technology. The brightness and contrast of the original image can be adjusted by image grayscale normalization technology, so that the grayscale value distribution of different frame images tends to be consistent, and a standardized image set is obtained. The area containing the first marker and the second marker is identified and segmented from the standardized image set, and the standardized image set is subjected to layer-by-layer feature extraction to capture the unique visual features of the first marker and the second marker. For example, the first marker is a structure with a through hole, and its elliptical shape and boundary contour have a high degree of recognition, while the second marker may have a specific surface texture or landmark. During the segmentation process, the network will automatically identify and mark the area containing the marker according to the trained model parameters, and remove irrelevant background information at the same time, and obtain a marker area image containing only the first marker and the second marker. The marker area image is subjected to fine feature extraction and spatial positioning analysis. Taking the first marker as an example, the first marker has a through hole structure, and its optical features can be analyzed by a deep learning feature detection algorithm, for example, by detecting the elliptical shape and boundary contour of the through hole, its center point position and radius information are determined. For the second marker, extract its unique surface texture or marker point features, for example, by analyzing its grayscale distribution or local contrast, determine the position of the marker point. After feature extraction, generate a feature set containing the through-hole geometry, center coordinates, and the position of the second marker point. Utilize the multi-view geometry principle of perspective imaging and combine the timing information of the image sequence for three-dimensional reconstruction. Specifically, by analyzing the position changes of the center coordinates of the through-hole of the first marker and the marker point of the second marker in multiple frames of images, the triangulation method can be applied to reconstruct the position of the marker in three-dimensional space, obtain a three-dimensional coordinate set based on the center line of the connecting tube, and calculate the relative distance and angle information between the first marker and the second marker, thereby forming an initial parameter set. Specifically, the straight-line distance between the center of the through hole of the first marker and the marking point of the second marker is determined by the Euclidean distance calculation method, and the distance is corrected in combination with the curvature information of the connecting pipe to reflect the actual geometric relationship. At the same time, the dynamic contour changes of the connecting pipe in the perspective image sequence can be combined, and its inner diameter, wall thickness and other parameters can be estimated through the timing analysis method, and the resistance coefficient of the fluid passing through the through hole of the first marker is calculated. The generated initial parameter set includes information such as the spatial coordinates of the marker, the through hole resistance coefficient, the inner diameter of the connecting pipe and the relative distance.

[0032] In one embodiment, the step of constructing a fluid model according to the initial parameter set, taking the distance information between the first identifier and the second identifier as a constraint condition of the fluid model, and generating a fluid parameter set comprises: According to the distance information between the first identification body and the second identification body in the initial parameter set, perform geometric constraint extraction processing on the connecting pipe image data to obtain a connecting pipe geometric feature set; According to the distance constraint and inner diameter distribution in the connecting pipe geometric feature set, perform modeling processing on the initial state of the fluid to generate a fluid model; Based on the fluid model, analyze the pressure gradient of the fluid in the connecting pipe to obtain a pressure distribution parameter set; According to the pressure distribution parameter set, correct the fluid model and output a fluid parameter set.

[0033] In the above embodiment, the three-dimensional coordinates of the first identification body and the second identification body are used to calculate the distance between them, and vector analysis is performed on the axis curvature of the connecting pipe to generate a geometric feature set including distance constraint, inner diameter distribution, and curvature characteristics. Using the information in the geometric feature set and combining the basic principles of fluid mechanics, such as mass conservation and momentum conservation, the initial state of the fluid in the connecting pipe is simulated. The specific modeling process includes determining the radial constraint of the fluid according to the inner diameter distribution and calculating the axial flow characteristics of the fluid between the first identification body and the second identification body in combination with the distance constraint. By iteratively solving the continuity equation, the velocity distribution and initial pressure field of the fluid can be preliminarily determined, thereby generating a preliminary model including fluid dynamics characteristics. After generating the preliminary fluid model, based on the preliminary fluid model, analyze the pressure transfer characteristics of the fluid in the connecting pipe. Since the velocity distribution of the fluid is closely related to the pressure gradient, use the correlation between the velocity gradient and shear stress to calculate the pressure change trend of each section in the connecting pipe. For example, in the region between the first identification body and the second identification body, the distance constraint may cause local changes in the fluid velocity, thereby triggering pressure fluctuations. To ensure the accuracy of the analysis, combine the curvature information in the geometric feature set to correct the abnormal values of the local pressure distribution. Through this series of calculations, the system generates a pressure distribution parameter set including axial pressure distribution, local pressure peak, and pressure gradient coefficient. Using the information in the pressure distribution parameter set, analyze the influence of the pressure gradient on the fluid model. For example, if the pressure distribution shows a significant pressure peak in a certain section of the connecting pipe, adjust the velocity distribution or boundary conditions in this region to reduce the deviation between the model and the actual physical phenomenon. In addition, combine the inner diameter and curvature information in the connecting pipe geometric feature set to comprehensively evaluate the dynamic response of the fluid in the model. Through an iterative optimization algorithm, calibrate the parameters of the fluid model, including flow rate, resistance coefficient, and local turbulence characteristics, and finally generate a fluid parameter set including predicted pressure, flow rate, and resistance correction factor.

[0034] In one embodiment, the step of controlling the pump body to output pressure to the initial value according to the fluid parameter set and calculating the pressure deviation according to the first change value of the distance information to generate a pre-expansion parameter set includes: Perform a grid decomposition on the set of fluid parameters to generate initial pressure distribution parameters including the pressure values of each grid node; Perform a pre - adjustment process on the pump body driving signal according to the initial pressure distribution parameters to obtain an initial pressure control instruction; Control the pump body to drive the fluid into the connecting pipe according to the initial pressure control instruction, and obtain real - time fluid state data, where the real - time fluid state data includes a first change value of the distance information between the first identification body and the second identification body; Calculate the deformation degree of the balloon according to the first change value, and calculate the actual pressure distribution inside the balloon according to the deformation degree to obtain a first pressure deviation; Optimize the initial pressure control instruction according to the first pressure deviation to obtain a corrected pressure control sequence; Integrate the corrected pressure control sequence with the convergence trend of the first pressure deviation to generate a pre - expansion parameter set including corrected initial pressure, actual flow rate, and deformation characteristics.

[0035] In the above embodiments, the fluid parameter set includes information such as fluid density, viscosity, and geometric characteristics of the connecting tube and the balloon. Using these data, the pressure distribution of the fluid in the connecting tube and the balloon is analyzed by the finite element analysis method. The internal spaces of the connecting tube and the balloon are divided into multiple small grid units, and the fluid state of each grid node is simulated and calculated. During the calculation process, in combination with the distance constraint between the first identification body and the second identification body, the pressure distribution characteristics of the fluid under static conditions are simulated, and an initial pressure distribution parameter containing the pressure values of each grid node is generated. According to the pressure values of the grid nodes in the initial pressure distribution parameter, the predicted pressure value at the balloon inlet is extracted. The predicted pressure value reflects the theoretical pressure required for the fluid to enter the balloon. Combining the resistance coefficient and the pipe length in the fluid model, the initial driving force required for the pump body is calculated. The resistance coefficient reflects the frictional loss when the fluid flows in the pipe, and the pipe length affects the pressure attenuation of fluid transmission. By comprehensively considering these factors, control parameters corresponding to the pump body motor frequency and the pulse width modulation signal are generated. These parameters are converted into executable drive signals through the signal conversion module to form an initial pressure control instruction. The pump body is controlled to operate according to the initial pressure control instruction, and the fluid is injected into the connecting tube and the balloon. The flow rate of the fluid and the dynamic pressure data at the balloon inlet are collected in real time through the flow sensor and the pressure sensor. Combining the real-time image obtained by the fluoroscopic imaging device, the distance change characteristics between the first identification body and the second identification body are extracted to generate real-time fluid state data including the flow rate, pressure, and distance change. The image processing algorithm is used to analyze the distance change value to calculate the deformation degree of the balloon. The deformation degree can be quantified by the relative displacement between the identification bodies, which reflects the expansion state of the balloon under the action of fluid pressure. Combining the flow rate and pressure in the real-time fluid state data, the control unit further calculates the actual pressure distribution inside the balloon. The actual pressure distribution is compared with the predicted pressure in the initial pressure distribution parameter to generate a first pressure deviation. The first pressure deviation is the difference between the actual pressure and the predicted pressure, which reflects the deviation degree of the initial pressure control instruction. In the optimization stage, based on the magnitude and direction of the first pressure deviation, an adaptive control algorithm is used to adjust the pump body drive signal in real time to generate a series of pressure control signals that change with time, forming a corrected pressure control sequence. The time series analysis of the corrected pressure control sequence is performed to extract the pressure values, flow rates, and corresponding distance change characteristics at each time point in the sequence. Combining the convergence trend of the first pressure deviation, the effect of the pressure adjustment is evaluated to determine whether it tends to be stable. Through the data fusion algorithm, these parameters are weighted and integrated to form a pre-expansion parameter set, which includes the corrected initial pressure, the actual flow rate, and the deformation characteristics of the balloon, and can comprehensively guide the pressure regulation in the balloon pre-expansion stage.

[0036] In one embodiment, the step of updating the fluid model based on the pre-expansion parameter set and outputting a dynamic pressure sequence includes: Extract the state features of the real-time vascular image obtained by the perspective imaging device to obtain a vascular state feature set; Perform stress distribution reconstruction processing on the initial fluid model according to the vascular state feature set to obtain an updated stress distribution model; Predict the flow rate of the stress distribution model to obtain a flow rate prediction sequence; According to the flow rate values and time nodes in the flow rate prediction sequence, perform pressure curve optimization processing on the initial pressure value in the pre-dilation parameter set to obtain a preliminary pressure sequence; Perform time series integration processing on the fluid model with the preliminary pressure sequence to obtain a dynamic pressure sequence.

[0037] In the above embodiments, edge detection techniques (such as the Canny algorithm or Sobel operator) can be used to locate the boundary contour of blood vessels. The boundary points are fitted into a three-dimensional surface through geometric modeling, so as to be able to describe the local curvature and dilation trend of blood vessels. Combining the dynamic distance change between the first identification body and the second identification body provided in the pre-dilation parameter set, the deformation rate of the blood vessel wall is calculated. The deformation rate is obtained by analyzing the time series of the distance change, which reflects the dynamic behavior of the blood vessel during the dilation process. Integrate information such as blood vessel diameter, wall thickness estimation value, and deformation rate to form a blood vessel state feature set. Using the geometric parameters (such as diameter and wall thickness) and deformation rate in the blood vessel state feature set, combined with the principle of fluid mechanics, recalculate the shear stress and radial stress distributions under the interaction between the fluid and the blood vessel wall. Specifically, the finite element analysis method can be used to discretize the blood vessel wall into multiple calculation units, and each unit corresponds to the mechanical properties of a local area. Based on the deformation rate provided in the feature set, the stress increment of each unit is derived, and at the same time, a wall thickness non-uniformity correction factor is introduced to reflect the influence of the change of blood vessel wall thickness at different parts on the mechanical behavior. The stress distribution is continuously updated through iterative calculation until the stress field of the entire model converges to a stable state. Extract the boundary conditions in the stress distribution model, such as the stress distribution of the blood vessel wall and the inlet velocity of the fluid, and then calculate the fluid velocity field through numerical integration methods (such as finite difference or finite volume method). During the calculation process, the geometric constraints of the pipeline (such as the change of blood vessel diameter) and the viscous resistance characteristics of the fluid are comprehensively considered to predict the flow rate values at different time points. To ensure the accuracy of the prediction, a multi-step time advancement strategy can be adopted to verify the coupling relationship between the flow rate and the stress distribution point by point, and eliminate possible numerical errors through iterative optimization. The generated flow rate prediction sequence contains multiple time nodes, and each node corresponds to a flow rate value. Adopt a reinforcement learning framework, take flow rate balance and stress stability as the objective function, and gradually optimize the pressure parameters through multiple rounds of trial and error. In each iteration, the algorithm records the deviation between the flow rate and the expected value, and updates the pressure adjustment strategy according to the deviation, so as to generate a pressure curve containing time nodes. Adopt time series analysis methods to decompose the preliminary pressure sequence into multiple time segments, and verify its matching degree with the flow rate and stress distribution segment by segment. Within each time segment, the algorithm dynamically adjusts model parameters, such as time step or pressure increment, to generate a dynamic pressure sequence containing multiple time nodes. This sequence can not only describe the dynamic behavior of the fluid in the blood vessel.

[0038] In another embodiment, the calculation expression of the deformation rate in the above embodiment is: ; where the deformation rate is used to quantify the dynamic deformation behavior of the blood vessel wall during dilation, and reflects the change speed of the geometric shape (such as diameter or boundary contour) of the blood vessel over time under the action of fluid pressure or other external forces; It represents the dynamic distance change, which is the change in the distance between the first identification body and the second identification body at time t, reflecting the local geometric deformation of the blood vessel wall during the process of dilation or contraction. For example, when the blood vessel diameter increases, the distance between the identification bodies increases. It is a positive value. It is the initial reference distance, which represents the distance between the first identification body and the second identification body when the blood vessel is in the initial state (without deformation), and is measured by imaging techniques (such as ultrasound or CT) when the blood vessel is in an unloaded state. It is the time interval, which represents the time interval between two measurements of the distance between the identification bodies and is the time step in time series analysis. It is the local curvature, which represents the local curvature of the blood vessel wall at time t and is calculated from the boundary points through geometric modeling (such as three-dimensional surface fitting). It is the relative deformation amount, which represents the ratio of the dynamic distance change to the initial reference distance, quantifying the relative deformation degree of the blood vessel wall. It is the time derivative factor, which represents the derivative factor of the deformation amount with respect to time and is used to convert the relative deformation amount into a rate. It is the curvature correction factor, without a correction factor based on the local curvature The correction factor is used to adjust the deformation rate to reflect the influence of the blood vessel geometry. When the curvature is large, the correction factor reduces the deformation rate, reflecting the constraint effect of the curvature on the deformation; when the curvature approaches 0 (straight pipe segment), the correction factor approaches 1 and the deformation rate approaches the uncorrected value.

[0039] In one embodiment, the step of predicting the flow rate of the stress distribution model to obtain the flow rate prediction sequence further includes: Obtain the shear stress and radial stress of each calculation unit in the stress distribution model, calculate the stress distribution characteristics of each unit through the finite element analysis algorithm, and obtain the stress distribution characteristic set. Combine the stress distribution characteristic set with the initial flow rate value in the pre-dilation parameter set, simulate the flow rate change trend of the fluid in the connecting pipe and the blood vessel, and generate the flow rate prediction sequence.

[0040] In this embodiment, recurrent neural network architectures such as long short-term memory networks (LSTMs) or gated recurrent units (GRUs) can be adopted. The stress distribution feature set is used as the input feature, and the initial flow rate value and its historical changes are used as the training labels. Through the training process, the model can learn the flow rate change law of the fluid in the blood vessel. After training, the model can predict the future flow rate change trend based on the new stress distribution feature set and generate a more accurate flow rate prediction sequence. In addition, to enhance the generalization ability of the model, data from different individuals, different blood vessel sites, and different dilation conditions can be collected as training samples to ensure that the model can maintain a high prediction accuracy when dealing with various complex situations.

[0041] In one embodiment, the step of controlling the pump body to execute the pressure instruction in the dynamic pressure sequence and obtaining the second change value of the distance information, and generating a regulation data set according to the second change value includes: Controlling the pump body to execute the pressure instruction in the dynamic pressure sequence, and obtaining the actual pressure change in the balloon to generate an actual pressure feedback set; Constructing a sliding window with a preset number of digits, and smoothing the actual pressure feedback set based on the sliding window to obtain a smoothed feedback signal set; Comparing the pressure values in the smoothed feedback signal set with the command pressure values in the initial execution instruction set point by point, and calculating the deviation value between the two; Weighting and correcting the deviation value with the second change value between the first identification body and the second identification body to generate a pressure deviation set including the deviation value, time series, and distance change; Integrating the actual pressure feedback set and the pressure deviation set, and analyzing the future short-term pressure change trend through time series prediction technology to generate a real-time regulation data set.

[0042] In the above embodiments, each pressure command in the dynamic pressure sequence is sent to the pump body in timestamp order to drive the pump body to output pressure at a specific frequency and intensity. Using the known relationship between the output pressure and flow rate of the pump body and combining the commands of the dynamic pressure sequence with the operating parameters of the pump body, the actual pressure change delivered to the balloon can be directly reflected, generating an actual pressure feedback set containing multi-dimensional data. A sliding window with a preset number of bits is constructed to smooth the pressure data. The number of bits of the sliding window can be determined according to the sampling frequency of the system and the expected smoothing degree, such as a 10-point or 20-point window. The data within the window is processed by methods such as weighted average or low-pass filtering to reduce the influence of high-frequency noise. By comparing the actual pressure values in the smoothed feedback signal set with these command pressure values one by one, the deviation value at each time point can be calculated. The calculation of the deviation value can use a simple difference method, that is, the actual pressure value minus the command pressure value, or more complex metrics can be introduced according to specific applications, such as relative error or weighted error. To improve the accuracy of deviation analysis, the second change value between the first identification body and the second identification body is introduced to weight and correct the deviation value. Here, the second change value can be understood as the dynamic change amount of the distance between the identification bodies within a certain time period (the second time), reflecting the physical state of balloon expansion. By using the distance change as a weighting factor, the influence of the geometric change of the balloon on the pressure demand can be comprehensively considered, generating a pressure deviation set containing deviation values, time series, and distance changes. The actual pressure feedback set and the pressure deviation set are integrated. The integration process includes operations such as data alignment, normalization, and feature extraction, and time series prediction techniques, such as autoregressive integrated moving average (ARIMA), long short-term memory network (LSTM), or other machine learning algorithms, are used to analyze the short-term trend of pressure change. These prediction models will infer the pressure demand within the next few seconds or milliseconds based on historical data (including deviation values, actual pressure values, and distance changes), and combine the flow rate constraints in the fluid model and the dynamic characteristics of balloon expansion to correct the prediction results, generating a real-time regulation data set. This data set contains the corrected pressure sequence, as well as the flow rate sequence, deviation data, and prediction trend.

[0043] In one embodiment, the step of optimizing the regulation data set and controlling the pump body to output pressure to the target value until the balloon is fully opened includes: Fusing the time series, deviation values, and distance change information in the regulation data set to obtain a comprehensive regulation parameter set; Iteratively optimizing the comprehensive regulation parameter set to generate an optimized pressure output command sequence; Controlling the pump body to output pressure according to the optimized pressure output command sequence until the actual pressure inside the balloon reaches the preset target pressure value; Monitor the third change value of the distance between the first identification body and the second identification body. When the third change value reaches the preset stable threshold, it is determined that the balloon is fully opened; Stop the pressure output of the pump body and output a status signal indicating that the balloon is fully opened.

[0044] In the above embodiment, the time series, deviation value, and distance change information in the regulation data set are fused and processed, and these information are integrated through a data fusion algorithm to generate a data set containing comprehensive regulation parameters. The comprehensive regulation parameter set contains pressure information and incorporates the geometric characteristics and dynamic behavior of balloon expansion. Iterative optimization is performed on the comprehensive regulation parameter set, with the goal of generating a pressure output instruction sequence that can precisely control the pressure output of the pump body. The iterative optimization process can use gradient-based optimization algorithms such as conjugate gradient method or quasi-Newton method, or heuristic optimization algorithms such as genetic algorithm or particle swarm algorithm. In each iteration, the error between the output pressure and the target pressure is calculated based on the current parameter set, and the parameter set is updated according to the magnitude and direction of the error until the error converges within an acceptable range. Control the pump body to output pressure according to the optimized pressure output instruction sequence. This process requires precise control of the operating parameters of the pump body, such as motor frequency and pulse width modulation signal, to ensure that the pump body can output pressure according to the requirements of the instruction sequence. At the same time, by real-time monitoring the actual pressure inside the balloon, it can be judged whether the balloon has reached the preset target pressure value and maintain this pressure value for a period of time to ensure the full expansion of the balloon. Monitoring the third change value of the distance between the first identification body and the second identification body is to judge whether the balloon has been fully opened. When the third change value reaches the preset stable threshold, it can be considered that the balloon has been fully opened. At this time, the pressure output of the pump body can be stopped and a status signal indicating that the balloon is fully opened is output. The status signal can be a signal sent by the control system when the third change value reaches the preset stable threshold.

[0045] Reference Figures 2 to 4, the present application also discloses an adjustable balloon inflation pressure pump, which is applied to the intelligent control method described in any one of the above, and includes: a pump body 1; a connection cavity 2, the connection cavity 2 is communicated with the pump body 1; an outer tube 3, one end of the outer tube 3 extends into the connection cavity 2 and is connected to the output port of the pump body 1, and the other end is connected to the balloon through a connection tube 4 to transmit the pressure generated by the pump body 1 to the balloon. A first marker 5 and a second marker 6 are arranged in the connection tube 4; a fluid storage tank 8, the fluid storage tank 8 is connected to the connection cavity 2 and is used for storing a fluid medium for the pump body 1 to extract and pressurize and output; the first marker 5 and the second marker 6 are arranged at one end of the connection tube 4 away from the connection cavity 2, a reflective groove 51 is arranged on the first marker 5, and an elastic member 7 is arranged on the side of the first marker 5 close to the outer tube 3, and the elastic member 7 is arranged in the connection tube 4.

[0046] In this embodiment, the pump body 1 is the power source of the entire pressure pump, responsible for extracting the fluid medium and pressurizing and outputting it. The pump body 1 communicates with the connection cavity 2 through its output port. The connection cavity 2 is a transition area for fluid transfer, connecting the pump body 1 with the external pipeline and also connected to the fluid storage tank 8. The function of the fluid storage tank 8 is to store the fluid medium, such as physiological saline or contrast agent, for the pump body 1 to continuously extract during operation. One end of the outer tube 3 extends into the connection cavity 2 and is directly connected to the output port of the pump body 1, so as to receive the pressurized fluid generated by the pump body 1; the other end is connected to the balloon through the connecting tube 4 to transmit the pressure to the balloon and achieve the expansion of the balloon. The first marker 5 and the second marker 6 are arranged in the connecting tube 4. These two markers are located at one end of the connecting tube 4 away from the connection cavity 2 and close to the balloon, so as to more directly sense the changes when the balloon is pressurized. The reflective groove 51 arranged on the first marker 5 can enhance the recognition ability of the marker by the fluoroscopic imaging device. The reflective groove 51 reflects light, making the marker present a higher contrast in the imaging, so as to facilitate the real-time image acquisition system to accurately capture its position information. An elastic member 7 is arranged on the side of the first marker 5 close to the outer tube 3, and this elastic member 7 is located inside the connecting tube 4. The function of the elastic member 7 is to realize the dynamic change measurement of the distance between the first marker 5 and the second marker 6. Specifically, the second marker 6 is relatively fixed in the connecting tube 4, while the first marker 5 is connected to the inner wall of the connecting tube 4 through the elastic member 7. When the pressurized fluid output by the pump body 1 flows through the connecting tube 4 to the balloon, the pressure of the fluid will push the first marker 5 closer to the second marker 6, thereby compressing the elastic member 7 and reducing the distance between the two. The elastic characteristics of the elastic member 7 ensure that when the pressure decreases, the first marker 5 can be pulled back to the original position, thus restoring the initial distance. Through the real-time image captured by the fluoroscopic imaging device, the system can identify the positioning information of the first marker 5 and the second marker 6, and calculate the pressure deviation based on the distance change value to generate a pre-expansion parameter set and a dynamic pressure sequence. Further, in the control method, the pump body 1 needs to generate a regulation data set according to the second change value of the distance information and optimize the output pressure to the target value. The presence of the elastic member 7 enables the first marker 5 to be sensitive to minute pressure changes, thus providing high-precision data support for the generation of the regulation data set. For example, when the balloon gradually expands, the fine adjustment of the fluid pressure will cause a slight change in the distance between the first marker 5 and the second marker 6. The elastic member 7 ensures that these changes can be accurately captured through its stretching and rebounding characteristics and feedback to the control system, ultimately achieving precise control of the complete opening of the balloon.

[0047] In summary, through the pump body 1, the connection cavity 2, the outer tube 3, the fluid storage tank 8, and the connecting pipe 4 with the first marker 5, the second marker 6, and the elastic member 7, this embodiment constructs an efficient and sensitive pressure transmission and monitoring system. The pump body 1 and the fluid storage tank 8 ensure the stable supply of the fluid. The outer tube 3 and the connecting pipe 4 achieve the precise transmission of pressure. The first marker 5 and the second marker 6, in cooperation with the elastic member 7, dynamically sense the pressure change and can effectively achieve the precise control of balloon expansion in practical applications.

[0048] It should be noted that the first marker 5, the second marker 6, and the elastic member 7 are arranged inside the connecting pipe 4, and a gap is left on its inner peripheral side for the liquid to flow through. The first marker 5 and the second marker 6 can be of a hollow structure to reduce the resistance to fluid flow and avoid hindering fluid transmission.

[0049] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention accordingly. Any equivalent structural or equivalent process transformations made by using the content of the specification and drawings of the present invention, or directly or indirectly applied to other related technical fields, are similarly included in the patent protection scope of the present invention.

Claims

1. An intelligent control method for an adjustable balloon inflation pressure pump, characterized in that: The method comprises: Based on the perspective imaging device, a real-time image of the connecting tube is acquired, positioning information of a marker in the connecting tube is identified, and an initial parameter set is generated, wherein the marker includes a first marker and a second marker; Constructing a fluid model according to the initial parameter set, taking the distance information between the first identifier and the second identifier as a constraint condition of the fluid model, and generating a fluid parameter set; Control the pump output pressure to an initial value according to the fluid parameter set, calculate the pressure deviation according to the first change value of the distance information, and generate a pre-expansion parameter set; updating the fluid model based on the pre-expansion parameter set and outputting a dynamic pressure sequence; Controlling the pump body to execute the pressure instruction in the dynamic pressure sequence, and obtaining a second change value of the distance information, and generating a control data set according to the second change value; The control data set is optimized to control the pump output pressure to a target value until the balloon is fully opened.

2. The intelligent control method of the adjustable balloon inflation pressure pump according to claim 1, characterized in that: The step of acquiring a real-time image of the connecting tube based on a fluoroscopic imaging device, identifying the positioning information of the marker in the connecting tube, and generating an initial parameter set includes: Preprocessing the raw image data collected by the fluoroscopic imaging device to obtain a standardized image set; Identifying the region containing the first marker and the second marker in the standardized image set and performing segmentation processing to obtain a marker region image; Identify the first marker and the second marker location information in the marker area image and generate an initial parameter set.

3. The intelligent control method of the adjustable balloon inflation pressure pump according to claim 1, characterized in that: The step of constructing a fluid model according to the initial parameter set, taking the distance information between the first identifier and the second identifier as a constraint condition of the fluid model, and generating a fluid parameter set comprises: According to the distance information between the first identifier and the second identifier in the initial parameter set, geometric constraint extraction processing is performed on the image data of the connection pipe to obtain a geometric feature set of the connection pipe; Modeling the initial state of the fluid according to the distance constraint and the inner diameter distribution in the connecting pipe geometric feature set to generate a fluid model; Analyze the pressure gradient of the fluid in the connecting pipe based on the fluid model to obtain a pressure distribution parameter set; The fluid model is modified according to the pressure distribution parameter set, and the fluid parameter set is output.

4. The intelligent control method of the adjustable balloon inflation pressure pump according to claim 1, characterized in that: The step of controlling the pump output pressure to an initial value according to the fluid parameter set, calculating the pressure deviation according to the first change value of the distance information, and generating a pre-expansion parameter set comprises: Decomposing the fluid parameter set into a grid to generate initial pressure distribution parameters including pressure values ​​of each grid node; Pre-adjusting the pump drive signal according to the initial pressure distribution parameters to obtain an initial pressure control instruction; Controlling the pump body to drive the fluid into the connecting pipe according to the initial pressure control instruction, and obtaining real-time fluid state data, wherein the real-time fluid state data includes a first change value of the distance information between the first identifier and the second identifier; Calculating the deformation degree of the balloon according to the first change value, and calculating the actual pressure distribution inside the balloon according to the deformation degree to obtain a first pressure deviation; Optimizing the initial pressure control instruction according to the first pressure deviation to obtain a modified pressure control sequence; The modified pressure control sequence is integrated with the convergence trend of the first pressure deviation to generate a pre-expansion parameter set including a modified initial pressure, an actual flow rate and a deformation characteristic.

5. The intelligent control method of the adjustable balloon inflation pressure pump according to claim 1, characterized in that: The step of updating the fluid model based on the pre-expansion parameter set and outputting a dynamic pressure sequence comprises: Extracting state features of the real-time blood vessel image acquired by the fluoroscopic imaging device to obtain a blood vessel state feature set; Performing stress distribution reconstruction processing on the initial fluid model according to the blood vessel state feature set to obtain an updated stress distribution model; Predicting the flow rate of the stress distribution model to obtain a flow rate prediction sequence; According to the flow rate values ​​and time nodes in the flow rate prediction sequence, the initial pressure values ​​in the pre-expansion parameter set are subjected to pressure curve optimization processing to obtain a preliminary pressure sequence; The preliminary pressure sequence is subjected to time series integration processing on the fluid model to obtain a dynamic pressure sequence.

6. The intelligent control method of the adjustable balloon inflation pressure pump according to claim 5, characterized in that: The step of predicting the flow rate of the stress distribution model to obtain a flow rate prediction sequence further includes: Obtaining the shear stress and radial stress of each calculation unit in the stress distribution model, calculating the stress distribution characteristics of each unit by a finite element analysis algorithm, and obtaining a stress distribution characteristic set; The stress distribution feature set is combined with the initial flow rate value in the pre-expansion parameter set to simulate the flow rate change trend of the fluid in the connecting tube and the blood vessel, and generate a flow rate prediction sequence.

7. The intelligent control method of the adjustable balloon inflation pressure pump according to claim 1, characterized in that: The step of controlling the pump body to execute the pressure instruction in the dynamic pressure sequence, obtaining the second change value of the distance information, and generating a control data set according to the second change value includes: Controlling the pump body to execute the pressure instructions in the dynamic pressure sequence, and obtaining the actual pressure change in the balloon to generate an actual pressure feedback set; Constructing a sliding window with a preset number of bits, and performing smoothing processing on the actual pressure feedback set based on the sliding window to obtain a smoothed feedback signal set; Compare the pressure value in the smoothed feedback signal set with the command pressure value in the initial execution command set point by point, and calculate the deviation value between the two; Performing weighted correction on the deviation value by using a second change value between the first identifier and the second identifier to generate a pressure deviation set including the deviation value, time series and distance change; The actual pressure feedback set and the pressure deviation set are integrated, and the future short-term pressure change trend is analyzed through time series prediction technology to generate a real-time control data set.

8. The intelligent control method of the adjustable balloon inflation pressure pump according to claim 1, characterized in that: The step of optimizing the control data set and controlling the pump output pressure to a target value until the balloon is fully opened includes: The time series, deviation value and distance change information in the control data set are fused to obtain a comprehensive control parameter set; Iteratively optimizing the comprehensive control parameter set to generate an optimized pressure output instruction sequence; Controlling the pump body to output pressure according to the optimized pressure output instruction sequence until the actual pressure inside the balloon reaches a preset target pressure value; Monitoring a third change value of the distance between the first marker and the second marker, and determining that the balloon is fully opened when the third change value reaches a preset stability threshold; Stop the pressure output of the pump body and output a status signal that the balloon is fully opened.

9. An adjustable balloon inflation pressure pump, applied to the intelligent control method according to any one of claims 1 to 8, characterized in that: The adjustable balloon inflation pressure pump comprises: Pump body (1); A connecting chamber (2), the connecting chamber (2) being in communication with the pump body (1); an outer tube (3), one end of the outer tube (3) extending into the connecting cavity (2) and connected to the output port of the pump body (1), and the other end of the outer tube (3) being connected to the balloon via a connecting tube (4) so ​​as to transmit the pressure generated by the pump body (1) to the balloon, wherein a first marker (5) and a second marker (6) are arranged in the connecting tube (4); A fluid storage tank (8), the fluid storage tank (8) being connected to the connecting chamber (2) and being used to store fluid medium for the pump body (1) to extract and output under pressure.

10. The adjustable balloon inflation pressure pump according to claim 9, characterized in that: The first marker (5) and the second marker (6) are arranged at one end of the connecting tube (4) away from the connecting cavity (2); a reflective groove (51) is arranged on the first marker (5); an elastic member (7) is arranged on a side of the first marker (5) close to the outer tube (3); and the elastic member (7) is arranged inside the connecting tube (4).

Images