A method and system for catheter tip force feedback

Abstract

The application relates to the medical technical field and particularly relates to a feedback method and system based on the force of the end of a micro catheter, which comprises the following steps: obtaining three-dimensional data of a patient's blood vessel through CT scanning, and pre-processing the three-dimensional data; extracting features from the pre-processed three-dimensional data to generate a processed feature data set; collecting a contact force signal S through a force sensor at the end of the catheter; performing wavelet transformation on the collected contact force signal to decompose the contact force signal into approximate coefficients and detail coefficients at different resolutions; performing Kalman filtering on the detail coefficients Dj to obtain filtered estimated signals, and taking the filtered estimated signals as filtered detail coefficients; reconstructing the approximate coefficients A j and the filtered detail coefficients D' j D' into a time domain signal S' based on wavelet inverse transformation, and transmitting the time domain signal to a catheter control system; and fusing CT images, real-time force signals and catheter position information to obtain a fused signal; and the catheter control precision is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of medical technology, in particular to a feedback method and system based on the force of the tip of a microcatheter. BACKGROUND

[0002] In the field of minimally invasive medicine, the application of microcatheters is increasingly widespread. Surgery often requires very high precision and safety, so it is particularly important to accurately control the tip of the microcatheter. Traditional catheter control systems mainly rely on manual operation by the doctor, but this approach is often limited by the doctor's experience and visual feedback during surgery, making it difficult to achieve the desired control accuracy and safety.

[0003] In order to improve the accuracy and safety of catheter control, in recent years, researchers have begun to explore the use of advanced signal processing techniques to optimize catheter control systems. However, existing signal processing methods, such as Fourier transform, while excellent in processing stationary signals, have significant limitations in processing non-stationary signals, especially human body waveforms that change frequently in the time dimension. Fourier transform cannot provide local information in the time domain, so it is difficult to accurately capture the dynamic changes produced by the catheter when moving in the human body.

[0004] In addition, although some studies have attempted to apply filtering techniques to catheter control systems to eliminate noise and interference, these methods are often too simple to effectively deal with complex noise environments and dynamic changes in the human body system. Therefore, in actual surgery, these methods still have difficulty providing stable and reliable catheter control.

[0005] To solve the above problems, we propose a feedback method and system based on the force of the tip of a microcatheter. SUMMARY

[0006] The purpose of the present application is to provide a feedback method and system based on the force of the tip of a microcatheter to solve the problems raised in the background art.

[0007] To achieve the above purpose, the present application provides the following technical solution: a feedback method and system based on the force of the tip of a microcatheter, the method comprising the following steps:

[0008] S1: Obtain three-dimensional data of the patient's blood vessels by CT scanning, and preprocess the three-dimensional data, and generate a processed feature data set by feature extraction from the preprocessed three-dimensional data;

[0009] S2: Collect contact force signals S through the force sensor at the tip of the catheter, and perform wavelet transform on the collected contact force signals to decompose them into approximate coefficients and detail coefficients at different resolutions;

[0010] S3: Perform wavelet reconstruction on the detail coefficients Dj carrying out Kalman filtering to obtain a filtered estimation signal, taking the filtered estimation signal as a filtered detail coefficient;

[0011] S4: reconstructing the approximation coefficient A j and the filtered detail coefficient D j into a time domain signal S', and transmitting the time domain signal to the catheter control system;

[0012] S5: fusing the CT image, the real-time force signal and the catheter position information to obtain a fused signal, and transmitting the fused signal to the catheter control system;

[0013] S6: obtaining the time domain signal and the fused signal, determining an adjustment strategy for the catheter control system based on the time domain signal and the fused signal, and adjusting the parameters of the catheter control system based on a feedback mechanism.

[0014] Preferably, the step of collecting the contact force signal S through the force sensor at the catheter tip and carrying out wavelet transform to decompose the contact force signal into approximation coefficients and detail coefficients at different resolutions comprises: installing a force sensor at the catheter tip or the proximal end, collecting the contact force signal S(t) through the force sensor, wherein t represents time. The signal S(t) contains the force size and direction information when the catheter contacts the blood vessel wall, and the collected analog signal is converted into a digital signal through an analog-to-digital converter (ADC); the collected digital signal is preliminarily filtered, and the signal is normalized to unify the amplitude range; a suitable wavelet basis function is selected according to the signal characteristics, the number of wavelet decomposition layers n is determined according to the signal frequency range and resolution requirements; the preprocessed contact force signal S(t) is subjected to discrete wavelet transform (WT) to decompose it into approximation coefficients and detail coefficients at different resolutions, and the corresponding formula is WT(S) = {Aj, Dj | j = 1, 2, 3…n}, wherein Aj is the approximation coefficient at the jth layer, Dj is the detail coefficient at the jth layer, and each layer of decomposition divides the signal into an approximation coefficient and a detail coefficient, and the approximation coefficient is continuously decomposed until the set number of decomposition layers n is reached.

[0015] Preferably, the step of carrying out wavelet inverse transform on the detail coefficient D jThe step of performing Kalman filtering to obtain a filtered estimation signal comprises: extracting a detail coefficient of a certain layer from the result of wavelet transform; initializing a state vector, a state transition matrix, an observation matrix, a process noise covariance matrix and an observation noise covariance matrix of the Kalman filter according to prior knowledge and signal characteristics; predicting a future state of the signal according to a prior model, and predicting an error covariance matrix of the state estimation; calculating a Kalman gain using the observation noise covariance matrix and the predicted error covariance matrix, and updating the state estimation to obtain the filtered estimation signal according to the detail coefficient of the current layer and the predicted state in combination with the Kalman gain, and taking the state estimation after the Kalman filtering as the filtered detail coefficient.

[0016] Preferably, the step of performing inverse wavelet transform on the approximation coefficients A j and the filtered detail coefficients D j to reconstruct the time-domain signal S' comprises: extracting the approximation coefficients A j and the detail coefficients D j′ that have been processed by Kalman filtering or other denoising methods from the result of wavelet transform, determining a wavelet basis function used for the inverse wavelet transform, obtaining the lowest-frequency approximation coefficients, and using the inverse wavelet transform to reconstruct a rough version of the time-domain signal from the lowest-frequency approximation coefficients, for each detail coefficient layer, using the inverse wavelet transform to combine it with the current reconstructed signal to recover the high-frequency component of the signal layer by layer, and combining the reconstruction results of all layers to obtain the time-domain signal S'.

[0017] Preferably, the step of fusing the CT image, the real-time force signal and the catheter position information to obtain a fusion signal comprises: obtaining the catheter position information by obtaining the position information of the catheter in three-dimensional space, setting initial weights for the CT image, the real-time force signal and the catheter position information at the same time point, and calculating a weighted sum of the CT image, the real-time force signal and the catheter position information to obtain the fusion signal according to the weights for each time point or data point, and the corresponding calculation formula is wherein M i is the i-th modality data, w i is the weight of the i-th modality data, and k is the total number of modalities.

[0018] Preferably, the step of acquiring the time domain signal and the fusion signal, determining the adjustment strategy of the catheter control system based on the time domain signal and the fusion signal, and adjusting the parameters of the catheter control system based on the feedback mechanism comprises: acquiring the time domain signal and the fusion signal, performing feature extraction on the preprocessed time domain signal, determining the adjustment direction of the catheter control system based on the features of the time domain signal; analyzing the fusion signal to extract information directly related to the surgical operation, and formulating a preliminary catheter control strategy based on the fusion signal and the surgical target; combining the analysis results of the time domain signal and the fusion signal to determine the target adjustment strategy of the catheter control system; designing a closed-loop feedback system to compare the output of the catheter control system with the expected target; and generating an error signal according to the comparison result, and adjusting the parameters of the catheter control system according to the size and direction of the error signal.

[0019] A microcatheter tip force feedback system based on the microcatheter tip force feedback method of any one of the preceding claims, comprising:

[0020] A data acquisition module for acquiring three-dimensional data of a patient's blood vessel through CT scanning, and preprocessing the three-dimensional data to generate a processed feature data set by performing feature extraction on the preprocessed three-dimensional data;

[0021] A first processing module for acquiring a contact force signal S through a force sensor at the tip of the catheter, and performing wavelet transform on the acquired contact force signal to decompose it into approximate coefficients and detail coefficients at different resolutions;

[0022] A second processing module for performing Kalman filtering on the detail coefficients D j to obtain filtered estimated signals, and taking the filtered estimated signals as filtered detail coefficients;

[0023] A third processing module for reconstructing the approximate coefficients A j and the filtered detail coefficients D j ′ into a time domain signal S′ based on inverse wavelet transform, and transmitting the time domain signal to a catheter control system;

[0024] A fusion module for fusing CT images, real-time force signals, and catheter position information to obtain a fusion signal, and transmitting the fusion signal to the catheter control system;

[0025] A feedback adjustment module for acquiring the time domain signal and the fusion signal, determining the adjustment strategy of the catheter control system based on the time domain signal and the fusion signal, and adjusting the parameters of the catheter control system based on the feedback mechanism.

[0026] Compared with the prior art, the present application has the following advantages:

[0027] The three-dimensional data of the patient's blood vessels is obtained by CT scanning, and pre-processing and feature extraction are performed, so that a feature data set reflecting the shape and structure of the blood vessels can be generated. At the same time, the contact force signal is collected by using the force sensor at the end of the catheter, and wavelet transform and Kalman filter processing are performed to effectively remove noise and interference and improve the accuracy and reliability of the signal. The processed data provides a more accurate control basis for the catheter control system, thereby improving the accuracy of the catheter control.

[0028] 2. The CT image, real-time force signal and catheter position information are fused to obtain a fusion signal containing rich information. This fusion signal can more comprehensively reflect the actual situation during the operation, and provides more intuitive operation guidance for the doctor. At the same time, the catheter control system adjustment strategy determined based on the time domain signal and the fusion signal can adjust the position and attitude of the catheter in real time, avoiding the operation risk caused by improper operation and enhancing the safety of the operation. BRIEF DESCRIPTION OF DRAWINGS

[0029] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0030] Fig. 1 The method flowchart of the present application is shown in the figure.

[0031] Fig. 2 The system structure block diagram of the present application is shown in the figure. DETAILED DESCRIPTION

[0032] The technical solutions in the embodiments of the present application will be described clearly and completely in the following with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0033] EMBODIMENT

[0034] Please refer to Figs. 1-2 The present application provides a kind of based on micro catheter end force feedback method and system technical scheme: a kind of based on micro catheter end force feedback method and system, comprising the following steps:

[0035] S1: three-dimensional data of the patient's blood vessels is obtained by CT scanning, and the three-dimensional data is pre-processed, and feature extraction is performed from the pre-processed three-dimensional data to generate a processed feature data set;

[0036] Specifically, three-dimensional data of the patient's blood vessels are obtained by CT scanning; Gaussian filtering is used to remove noise in the CT image, and image enhancement techniques are used to improve the clarity of the blood vessel boundaries; a threshold segmentation algorithm is used to extract the blood vessel region and generate a three-dimensional blood vessel model; a convolution kernel is used to extract local features of the blood vessel (such as blood vessel diameter, tortuosity); global features (such as blood vessel branch structure) are extracted by statistical methods; a pre-processed feature dataset D is generated: D = f p (x) for intraoperative navigation and signal analysis;

[0037] Where D contains key anatomical information such as blood vessel location, organ boundary, etc., and irrelevant information that may interfere with subsequent signal analysis is removed. In addition, the preprocessing stage also includes data normalization operations to ensure consistency in spatial resolution and intensity distribution of the CT image. Finally, a feature construction algorithm is used to generate a data representation that matches the subsequent catheter navigation and signal analysis. This process can extract three-dimensional local features using a convolution kernel, extract global features using statistical methods, and finally output a feature dataset D that can be used for real-time processing during surgery;

[0038] S2: Collect the contact force signal S through the force sensor at the end of the catheter, and perform wavelet transform on the collected contact force signal to decompose it into approximate coefficients and detail coefficients at different resolutions;

[0039] The steps of collecting the contact force signal S through the force sensor at the end of the catheter and performing wavelet transform on the collected contact force signal to decompose it into approximate coefficients and detail coefficients at different resolutions include: installing a force sensor at the end or near the end of the catheter, collecting the contact force signal S(t) through the force sensor, where t represents time. The signal S(t) contains the force magnitude and direction information when the catheter contacts the blood vessel wall. The collected analog signal is converted into a digital signal through an analog-to-digital converter (ADC); the collected digital signal is pre-filtered and normalized to unify the amplitude range; a suitable wavelet basis function is selected according to the signal characteristics, and the number of wavelet decomposition layers n is determined according to the signal frequency range and resolution requirements; the pre-processed contact force signal S(t) is subjected to discrete wavelet transform (WT), which is decomposed into approximate coefficients and detail coefficients at different resolutions, and the corresponding formula is WT(S) = {A j ,D j |j = 1, 2, 3…n}, where A j is the approximate coefficient of the jth layer, D j is the detail coefficient of the jth layer, and each layer of decomposition divides the signal into approximate coefficients and detail coefficients, and the approximate coefficients continue to be decomposed until the set number of decomposition layers n is reached.

[0040] It should be noted that the selection criteria for appropriate wavelet basis functions based on signal characteristics are: first, the time-frequency localization characteristics of the wavelet basis functions; and second, the degree of matching between the wavelet basis functions and signal characteristics.

[0041] Specifically, during the procedure, a contact force signal S is acquired using a force sensor at the end of the catheter; the acquired force signal is then subjected to wavelet transform, decomposing it into approximation coefficients (low-frequency components) and detail coefficients (high-frequency components): WT(S)={A j D j |j=1,2,3…n}, where A j The approximation coefficients (low-frequency part) represent the overall trend of the signal, D. j These are detail coefficients (high-frequency components), representing the local variation characteristics of the signal, where n is the number of decomposition levels; approximation coefficient A. j Used to capture the overall trend of force; detail factor D j Used for analyzing the local variation characteristics of force;

[0042] By selecting appropriate wavelet basis functions, a balance can be achieved between the time and frequency domains. The goal at this stage is to lay the foundation for signal denoising and feature extraction while reducing computational complexity. In intraoperative applications, decomposition at different scales can more intuitively demonstrate changes in key components of the signal, aiding in precise catheter localization and lesion analysis.

[0043] Approximation coefficient A j Used to capture the overall trend of contact force signals, reflecting the main changes in the contact force between the catheter and the vessel wall. It can be used to analyze the overall motion state of the catheter (e.g., advancement, rotation).

[0044] Detail factor D j It is used to analyze the local variation characteristics of contact force signals, reflecting minute fluctuations when the catheter comes into contact with the blood vessel wall. It can be used to detect abnormal force signals (such as collisions and friction).

[0045] Noise reduction for high-frequency detail coefficients: j Threshold denoising is performed to remove noise interference. Denoising methods include hard thresholding and soft thresholding.

[0046] Feature extraction: from approximation coefficient A j and the detail coefficient D after noise reduction j Key features to extract: Approximate coefficient features: mean force and trend changes. Detail coefficient features: frequency and amplitude changes of force fluctuations;

[0047] S3: For detail factor D j Kalman filtering is performed to obtain the filtered estimated signal, and the filtered estimated signal is used as the filtered detail coefficients;

[0048] For detail factor Dj The step of performing Kalman filtering to obtain a filtered estimation signal, and taking the filtered estimation signal as the filtered detail coefficient comprises: extracting the detail coefficient of a certain layer from the result of wavelet transform; initializing the state vector, state transition matrix, observation matrix, process noise covariance matrix and observation noise covariance matrix of the Kalman filter according to prior knowledge and signal characteristics; predicting the future state of the signal according to the prior model, and predicting the error covariance matrix of the state estimation; calculating the Kalman gain using the observation noise covariance matrix and the prediction error covariance matrix, and updating the state estimation to obtain the filtered estimation signal according to the detail coefficient of the current layer and the predicted state combined with the Kalman gain, and taking the state estimation after Kalman filtering as the filtered detail coefficient;

[0049] Specifically, let D j be the detail coefficient of a certain layer, and the filtered estimation signal be represented as D j ', D j ' = KF(D j ), wherein KF represents the Kalman filtering function; the future state of the signal is predicted according to the prior model, the prediction result is adjusted according to the actual observation value, the signal is optimally estimated to obtain the filtered estimation signal; the filtered detail coefficient D j ' retains the local characteristics of the signal while reducing noise interference;

[0050] The filtering process consists of two steps: first, the future state of the signal is predicted according to the prior model (i.e. the state transition matrix); second, the prediction result is adjusted according to the actual observation value to obtain a more accurate estimation value. In intraoperative applications, high-frequency components are usually associated with noise, so Kalman filtering is mainly used for optimization of high-frequency signals to reduce the influence of pseudo signals on catheter control;

[0051] S4: reconstructing the approximation coefficient A j and the filtered detail coefficient D j ' into a time-domain signal S' based on wavelet inverse transform, and transmitting the time-domain signal to the catheter control system;

[0052] The step of reconstructing the approximation coefficient A j and the filtered detail coefficient D j ' into a time-domain signal S' based on wavelet inverse transform comprises: extracting the approximation coefficient A j and the detail coefficient D j′where j' represents different decomposition levels, and determines the wavelet basis function for the inverse wavelet transform; the lowest frequency approximation coefficient is obtained, and the wavelet inverse transform is used to reconstruct the lowest frequency approximation coefficient into a rough version of the time domain signal; for each level of detail coefficient, the wavelet inverse transform is used to combine it with the current reconstructed signal to recover the high frequency component of the signal layer by layer; the reconstruction results of all levels are combined to obtain the time domain signal S';

[0053] Specifically, the approximation coefficient A j and the filtered detail coefficient D j ′ are obtained, and the approximation coefficient A j and the filtered detail coefficient D j are reconstructed into the time domain signal S' based on the wavelet inverse transform, and the corresponding signal reconstruction formula is S' = WT -1 (A j ,D j ′), where WT -1 represents the wavelet inverse transform. Generally, the reconstruction process starts from the highest frequency detail coefficient (if there are multiple levels), but in fact, from the perspective of multi-scale decomposition of wavelet transform, the reconstruction should start from the lowest frequency approximation coefficient, and then add the detail coefficients layer by layer. Here, for the sake of logical clarity, the present application is described in the reconstruction order; the wavelet inverse transform is used to reconstruct the lowest frequency approximation coefficient into a rough version of the time domain signal. This step is equivalent to reconstructing the low frequency component of the signal; for each level of detail coefficient (from high frequency to low frequency, or according to the specific decomposition order), the wavelet inverse transform is used to combine it with the current reconstructed signal to recover the high frequency component of the signal layer by layer. This step is achieved by convolving the detail coefficient with the corresponding wavelet function and summing, thereby gradually refining the reconstructed signal; all levels of reconstruction results are combined to obtain the final reconstructed signal S'. Ensure that all frequency components of the signal are accurately recovered;

[0054] The wavelet inverse transform is used to reconstruct the filtered approximation coefficient Aj and the detail coefficient Dj' into the time domain signal S', and the corresponding formula is S' = WT -1 (A j ,D j ′), where WT -1denotes the inverse wavelet transform; the reconstructed signal S' is transmitted to the catheter control system for real-time adjustment of the catheter movement; this process ensures that the denoised signal has high fidelity in the time domain and is suitable for real-time transmission and analysis during surgery. The key to signal reconstruction is to maintain the time characteristics of the original signal while minimizing the destruction of its structure by noise. The reconstructed signal S' is transmitted to the catheter control system to guide the catheter movement and position adjustment, and the catheter control system adjusts the movement and position of the catheter in real time based on the received reconstructed time domain signal, achieving the ultimate purpose of signal processing, i.e., guiding the catheter movement during surgery;

[0055] S5: fuse the CT image, real-time force signal and catheter position information to obtain a fused signal, and transmit the fused signal to the catheter control system;

[0056] The step of fusing the CT image, real-time force signal and catheter position information to obtain a fused signal includes: obtaining the position information of the catheter in the three-dimensional space to obtain the catheter position information, setting initial weights for the CT image, real-time force signal and catheter position information at the same time point, and calculating the weighted sum of the CT image, real-time force signal and catheter position information according to the weights to obtain the fused signal for each time point or data point, and the corresponding calculation formula is wherein M i is the i-th modal data, wi is the weight of the i-th modal data, and k is the total number of modalities;

[0057] Specifically, the weighted fusion process itself helps to reduce the interference of false signals, because the modal data with lower weight has less influence on the fusion result. In addition, false signals can be further eliminated through additional signal processing techniques (such as threshold setting, signal morphology analysis, etc.), and the CT image, real-time force signal and catheter position information can be effectively fused to obtain a fused signal with high credibility and accuracy;

[0058] Specifically, the CT image, real-time force signal and catheter position information are fused; the multi-modal data is weighted and fused to eliminate false signals, and the weights are dynamically adjusted according to the surgical scene to ensure the accuracy and reliability of the fusion result; through the weighted fusion algorithm, the system can integrate the advantages of multiple information sources, eliminate the interference of false signals, and improve the credibility and accuracy of the signal. In addition, the fusion algorithm can dynamically adjust the weights through a learning mechanism to ensure optimal performance in different surgical scenarios;

[0059] S6: obtain the time domain signal and the fused signal, determine the adjustment strategy for the catheter control system based on the time domain signal and the fused signal, and adjust the parameters of the catheter control system based on the feedback mechanism;

[0060] The step of acquiring the time domain signal and the fusion signal, determining the adjustment strategy for the catheter control system based on the time domain signal and the fusion signal, and adjusting the parameters of the catheter control system based on the feedback mechanism comprises: acquiring the time domain signal and the fusion signal, performing feature extraction on the preprocessed time domain signal, determining the adjustment direction for the catheter control system based on the features of the time domain signal; analyzing the fusion signal to extract information directly related to the surgical operation, and formulating a preliminary catheter control strategy based on the fusion signal and the surgical target; combining the analysis results of the time domain signal and the fusion signal to determine the target adjustment strategy of the catheter control system; designing a closed-loop feedback system to compare the output of the catheter control system with the expected target; generating an error signal according to the comparison result, and adjusting the parameters of the catheter control system according to the size and direction of the error signal;

[0061] Specifically, the sensors are used to collect physical quantities such as force, displacement, and velocity during the catheter operation in real time. These physical quantities change over time to form time domain signals. The collected time domain signals are preprocessed, such as filtering and denoising, to improve the accuracy and reliability of the signals. The CT images, real-time force signals, and catheter position information are weighted and fused to obtain a fusion signal. The fusion signal should comprehensively reflect the key information in the catheter operation process, such as anatomical structure, tissue hardness, and catheter position. The preprocessed time domain signals are analyzed to identify abnormal or critical events in the catheter operation process, such as sudden increase in force and sudden change in displacement. According to the characteristics of the time domain signal, the adjustment direction of the catheter control system is preliminarily determined, such as slowing down the speed and increasing the force. The fusion signal is analyzed to extract information directly related to the surgical operation, such as the relative position of the catheter and the surrounding tissue and the hardness distribution of the tissue. In combination with the fusion signal and the surgical target, a detailed catheter control strategy is formulated, such as adjusting the catheter path and avoiding sensitive areas. The analysis results of the time domain signal and the fusion signal are combined to consider factors such as surgical safety, efficiency, and accuracy to determine the final catheter control system adjustment strategy. A closed-loop feedback system is designed to compare the output of the catheter control system (such as catheter position and speed) with the expected target. According to the comparison result, an error signal is generated as the basis for adjusting the parameters of the catheter control system. According to the size and direction of the error signal and the pre-set control algorithm (such as PID control and adaptive control), the parameters of the catheter control system are adjusted. The parameter adjustment should ensure that the catheter can smoothly and accurately track the expected path while avoiding unnecessary damage to the surrounding tissue. The adjustment strategy for the catheter control system can be determined based on the time domain signal and the fusion signal, and the system parameters can be adjusted in real time through the feedback mechanism to achieve accurate control of the catheter operation and safe and efficient surgical process.

[0062] Specifically, real-time feedback signals are acquired through force sensors and position sensors; the rotation, propulsion speed and bending angle of the catheter are adjusted in combination with preset path planning and real-time feedback signals; a safety force threshold is set to detect and warn abnormal force values in real time; flexible contact of the catheter is realized through active force control, resistance compensation and steady-state force maintenance mechanism; according to the reconstruction signals and fusion results, the catheter system needs to be adjusted in real time to adapt to the dynamically changing environment in the operation. The parameters of the catheter are adjusted through the feedback mechanism to optimize the motion mode of the catheter tip. Specifically, the feedback signals are acquired through sensors, and the control algorithm adjusts the rotation, propulsion speed and bending angle of the catheter in combination with the preset path planning and the signal characteristics calculated in real time, so as to ensure the accurate positioning and safety of the catheter.

[0063] A feedback system based on the force at the tip of a microcatheter, applied to the feedback method based on the force at the tip of a microcatheter as claimed in any one of the above, comprising:

[0064] A data acquisition module for acquiring three-dimensional data of a patient's blood vessel through CT scanning, and pre-processing the three-dimensional data to generate a processed feature data set by feature extraction from the pre-processed three-dimensional data;

[0065] A first processing module for acquiring contact force signals S through force sensors at the tip of the catheter, and performing wavelet transform on the acquired contact force signals to decompose them into approximate coefficients and detail coefficients at different resolutions;

[0066] A second processing module for performing Kalman filtering on the detail coefficients D j to obtain filtered estimated signals, and taking the filtered estimated signals as filtered detail coefficients;

[0067] A third processing module for reconstructing the approximate coefficients A j and the filtered detail coefficients D j ′ into time domain signals S′ based on inverse wavelet transform, and transmitting the time domain signals to the catheter control system;

[0068] A fusion module for fusing CT images, real-time force signals and catheter position information to obtain fusion signals, and transmitting the fusion signals to the catheter control system;

[0069] A feedback adjustment module for acquiring time domain signals and fusion signals, determining an adjustment strategy for the catheter control system based on the time domain signals and the fusion signals, and adjusting the parameters of the catheter control system based on a feedback mechanism.

[0070] The CT data of the blood vessel of the patient is collected and three-dimensional reconstruction is completed, the signal is transmitted to the robot main controller through analog-digital conversion and filtering processing by using the improved force feedback system, the purpose of eliminating false signals is achieved through multi-modal information fusion, and a safety force threshold is set.

[0071] Firstly, the CT data of the blood vessel of the patient is collected by the doctor before the operation and three-dimensional reconstruction is completed; in the operation, the force feedback system collects the contact force and direction data through the force sensor at the end or the proximal end of the catheter, and transmits the signal to the robot main controller through analog-digital conversion and filtering processing, so as to ensure the real-time and accuracy of the force information; the force data are fused with the image and position information to eliminate errors and correct false signals, and a safety force threshold is set to detect and warn the abnormal force value in real time, so as to ensure the safety and accuracy of the operation. In the process of eliminating errors, wavelet transform and Kalman filtering method are adopted; the system actively adjusts the catheter movement according to the feedback force, realizes flexible contact through active force control, resistance compensation and steady force keeping mechanism, and transmits the real-time force information to the doctor through tactile or visual feedback.

[0072] Compared with the processing of digital signals by traditional Fourier transform, doctors can process and transmit the changes of human body waveform in time dimension by using wavelet transform, and the Kalman filtering method can help doctors consider the influence of the changes of human body system on the blood vessel. The influence can more scientifically correct the errors caused by information detection in the process of human body dynamic changes.

[0073] In addition, the application provides a new digital signal processing method based on wavelet transform and Kalman filtering algorithm. The use of wavelet transform helps to capture the local characteristics of the signal, especially in the part with high change frequency; the use of Kalman filtering algorithm can optimize the signal estimation by considering the characteristics of process noise and measurement noise, and provide scientific navigation for doctors in actual operation based on physiological activities in the patient's body.

[0074] In the description of the present specification, the description of the terms "one embodiment", "example", "specific example" and the like means that the specific features, structures, materials or characteristics described in combination with the embodiment or example are contained in at least one embodiment or example of the present application. In the present specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0075] While embodiments of the application have been shown and described, it is to be understood that the embodiments described are merely exemplary of the principles and application of the present application. Numerous modifications and adaptions can be effected without departing from the spirit and scope of the present application, which is not limited to the exact construction and arrangement described. It is intended, therefore, to cover all modifications and adaptions that fall within the scope of the claims and their equivalents.

Claims

1. A feedback system based on microcatheter tip force, characterized by, The method comprises the following steps: A data acquisition module is configured to acquire three-dimensional data of a patient's blood vessel through CT scanning, and to preprocess the three-dimensional data, and to extract features from the preprocessed three-dimensional data to generate a processed feature data set; A first processing module is configured to acquire a contact force signal S through a force sensor at the end of a catheter, and to perform wavelet transform on the acquired contact force signal to decompose the contact force signal into approximate coefficients and detail coefficients at different resolutions; The second processing module is used for detail coefficients. Kalman filtering is performed to obtain the filtered estimated signal, and the filtered estimated signal is used as the filtered detail coefficients; a third processing module configured to reconstruct the approximation coefficients and the filtered detail coefficients into a time domain signal based on an inverse wavelet transform and filtered detail coefficients passing the time domain signal to a catheter control system;​ A fusion module is configured to fuse CT images, real-time force signals, and catheter position information to obtain a fusion signal, and to transmit the fusion signal to a catheter control system; A feedback adjustment module is configured to acquire a time domain signal and a fusion signal, to determine an adjustment strategy for the catheter control system based on the time domain signal and the fusion signal, and to adjust parameters of the catheter control system based on a feedback mechanism; The time domain signal and the fusion signal are acquired, features are extracted from the preprocessed time domain signal, the adjustment direction of the catheter control system is determined based on the features of the time domain signal, the fusion signal is analyzed, information directly related to the surgical operation is extracted, a preliminary catheter control strategy is formulated based on the fusion signal and the surgical target, the analysis results of the time domain signal and the fusion signal are combined, and a target adjustment strategy of the catheter control system is determined; A closed-loop feedback system is designed, the output of the catheter control system is compared with the expected target, an error signal is generated according to the comparison result, and the parameters of the catheter control system are adjusted according to the size and direction of the error signal.

2. The microcatheter tip force feedback system of claim 1, wherein: The force sensor at the end of the catheter collects the contact force signal S, and the step of wavelet transform of the collected contact force signal to decompose into approximate coefficients and detail coefficients at different resolutions includes: installing a force sensor at the end or proximal end of the catheter, collecting the contact force signal S(t) through the force sensor, wherein t represents time; the signal S(t) contains the force size and direction information when the catheter contacts the blood vessel wall, and the collected analog signal is converted into a digital signal through an analog-to-digital converter (ADC); the collected digital signal is preliminarily filtered, and the signal is normalized to unify the amplitude range; a suitable wavelet base function is selected according to the signal characteristics, and the number of wavelet decomposition layers is determined according to the signal frequency range and resolution requirement ; the preprocessed contact force signal S(t) is subjected to discrete wavelet transform (WT) to decompose it into approximate coefficients and detail coefficients at different resolutions, and the corresponding formula is , wherein is the approximate coefficient of the jth layer, is the detail coefficient of the jth layer, and each layer of decomposition divides the signal into approximate coefficients and detail coefficients, and the approximate coefficients continue to be decomposed until the set number of decomposition layers n is reached.

3. The microcatheter tip force feedback system of claim 1, wherein: the detail coefficients The step of performing Kalman filtering to obtain the filtered estimation signal, and taking the filtered estimation signal as the filtered detail coefficient comprises: extracting the detail coefficient of a layer from the result of wavelet transform; initializing the state vector, state transition matrix, observation matrix, process noise covariance matrix and observation noise covariance matrix of the Kalman filter according to prior knowledge and signal characteristics; predicting the future state of the signal according to a prior model, and predicting the error covariance matrix of the state estimation; calculating the Kalman gain by using the observation noise covariance matrix and the prediction error covariance matrix, and updating the state estimation to obtain the filtered estimation signal according to the detail coefficient of the current layer and the predicted state in combination with the Kalman gain, and taking the state estimation after the Kalman filtering as the filtered detail coefficient.

4. The microcatheter tip force feedback system of claim 1, wherein: The step of reconstructing the approximation coefficients and the filtered detail coefficients into a time-domain signal includes extracting the approximation coefficients and the detail coefficients processed by Kalman filtering or other denoising methods from the results of wavelet transform, wherein indicates different decomposition levels, and a wavelet basis function used for wavelet inverse transform is determined; the lowest-frequency approximation coefficients are obtained, and the lowest-frequency approximation coefficients are reconstructed into a coarse version of the time-domain signal using wavelet inverse transform; for each level of detail coefficients, the wavelet inverse transform is used to combine the detail coefficients with the current reconstructed signal to recover the high-frequency components of the signal layer by layer; and the reconstructed results of all levels are combined to obtain the time-domain signal .

5. The microcatheter tip force feedback system of claim 1, wherein: The step of fusing the CT image, the real-time force signal and the catheter position information to obtain a fusion signal comprises: obtaining the position information of the catheter in a three-dimensional space to obtain the catheter position information, setting initial weights for the CT image, the real-time force signal and the catheter position information at the same time point, and calculating a weighted sum of the CT image, the real-time force signal and the catheter position information according to the weights to obtain the fusion signal for each time point or data point, and the corresponding calculation formula is wherein, is the i-th modality data, is the weight of the i-th modality data, and k is the total number of modalities.

Images